Win

Gui
Larry
Marti
Jason
Ralph
Review
Calder,
Yousef
Mizono
Brian S

Summa
practice
Categor
types of
applicat
libraries,
Categor
Applies
Services
Source:
E-book
422 pag

ndow
idanc
Franks,
nez, Va
n Roth, M
h Squilla
wers: Thia
Janice Ch
Khalidi, C
ov, Sasha
wan, Tina
ry: Window
s and recom
ies cover p
f Windows A
ions. It also
, and stacks
ry: Guide
to: Window
, and Enter
MSDN Libr
publicatio
es
ws Az
ce
Sidney
lery Miz
Mark Sim
ace, Patr
go Almeid
hoi, Brian
hristian M
Nosov, Ja
Stewart,
ws Azure Pr
mmendatio
lanning, de
Azure appli
o provides g
s with Wind
ws Azure , W
prise Integr
rary (link to
on date: Ma

zure
Higa, S
zonov, W
mms, Pa
rick Wic
da, Jaime
Goldfarb,
Martinez, P
ames Podg
Michael T
escriptive G
ns for work
esigning, de
cations, inc
guidance fo
dows Azure
Windows A
ration Appl
o source con
ay 2012
Presc
Suren M
Walter M
aolo Sa
ckline, T
Alva Brav
, Sidney H
Paulette M
gorski, Vla
Thomassy,
Guidance pr
king with th
eveloping, a
cluding mo
or develope
.
Azure SQL D
ications
ntent)
cript
Machiraju
Myers II
lvatori,
Trace Yo
vo, Monile
Higa, Mich
McKay, Sho
adRomane
, Steve W
rovides you
he Windows
and managi
bile, enterp
ers using no
Database, W
tive
u, Christ
II, Rama
Adam S
oung
ee Atkinso
hael Hyatt,
ont Miller
enko, Ralp
ilkins, Ste
u with some
s Azure plat
ing a variety
prise, and co
on-.NET app
Windows Az
tian
a Raman
Skewga
on, Brad
,
r, Valery
ph Squilla
ve Young
e of the bes
tform.
y of differe
onsumer-ba
plications,
ure Cloud

ni,
r,
ace,

st
nt
ased

Copyright 2012 by Microsoft Corporation

All rights reserved. No part of the contents of this book may be reproduced or transmitted in any form or by any means
without the written permission of the publisher.


Microsoft and the trademarks listed at
http://www.microsoft.com/about/legal/en/us/IntellectualProperty/Trademarks/EN-US.aspx are trademarks of the
Microsoft group of companies. All other marks are property of their respective owners.

The example companies, organizations, products, domain names, email addresses, logos, people, places, and events
depicted herein are fictitious. No association with any real company, organization, product, domain name, email address,
logo, person, place, or event is intended or should be inferred.

This book expresses the authors views and opinions. The information contained in this book is provided without any
express, statutory, or implied warranties. Neither the authors, Microsoft Corporation, nor its resellers, or distributors will
be held liable for any damages caused or alleged to be caused either directly or indirectly by this book.


1
Contents
Welcome to the Windows Azure Prescriptive Guidance ................................................................. 2
Planning and Designing Windows Azure Applications .................................................................... 3
Is Your Application a Good Fit for Windows Azure? .................................................................... 4
Designing Multitenant Applications on Windows Azure ............................................................. 12
Packaging and Deploying an Application to Windows Azure .................................................... 30
Best Practices for Running Unmanaged Code in Windows Azure Applications ........................ 34
Other Third Party Software on Windows Azure ......................................................................... 43
Leveraging Node.js libuv in Windows Azure .......................................................................... 43
Business Continuity for Windows Azure .................................................................................... 49
Messaging ..................................................................................................................................... 57
Capacity Planning for Service Bus Queues and Topics ............................................................ 58
Best Practices for Handling Large Messages with Windows Azure Queues ............................. 61
Best Practices for Maximizing Scalability and Cost Effectiveness of Queue-Based Messaging
Solutions on Windows Azure .................................................................................................. 98
How to Integrate a BizTalk Server Application with Service Bus Queues and Topics ............ 127
Service Bus Queues ............................................................................................................. 128
Service Bus Topics ............................................................................................................... 130
BrokeredMessage ................................................................................................................. 132
NetMessagingBinding ........................................................................................................... 135
BizTalk WCF Adapters ......................................................................................................... 139
Scenarios .............................................................................................................................. 142
Solution ................................................................................................................................. 150
Testing the Solution .............................................................................................................. 274
Implementing a Message Fan-Out Scenario ........................................................................ 276
Conclusion ............................................................................................................................ 277
Managing and Testing Topics, Queues and Relay Services with the Service Bus Explorer Tool
.............................................................................................................................................. 278
Testing ......................................................................................................................................... 337
Using Visual Studio Load Tests in Windows Azure Roles ....................................................... 338
Visual Studio Load Test in Windows Azure Overview .......................................................... 339
Windows Azure Load Test Prerequisites and Setup ............................................................ 344
Provisioning Windows Azure For a Load Test ...................................................................... 346
Publishing the Load Test To Windows Azure ....................................................................... 352
Running Load Tests In Mixed Environments ........................................................................ 357
Guidance on Efficiently Testing Azure Solutions ..................................................................... 367
Database ..................................................................................................................................... 381
Data Migration to SQL Azure: Tools and Techniques .............................................................. 382
Using the ReportViewer ASP.NET Control in Windows Azure ................................................ 405


2
Welcome to the Windows Azure Prescriptive
Guidance
Windows Azure Prescriptive Guidance provides you with some of the best practices and
recommendations for working with the Windows Azure platform. Categories cover planning,
designing, developing, and managing a variety of different types of Windows Azure applications,
including mobile, enterprise and consumer-based applications. It also provides guidance for
developers using non-.NET applications, libraries, and stacks with Windows Azure.
The scenarios described here are based on direct customer input to the Windows Azure
Customer Advisory Team (CAT) and the developer documentation team. These teams will
continue to collect new guidance over the coming weeks and months. The most current topics,
including any updates to the topics in this book, are located at the Windows Azure Development
Guidance. If there is a scenario or topic that you need best practices for, please contact us at
[email protected].


3
Planning and Designing Windows Azure
Applications
This section contains articles that provide guidance about planning and designing either Windows
Azure applications themselves or their provisioning and management.



4
Is Your Application a Good Fit for Windows
Azure?
Author: Jason Roth
Reviewers: Paulette McKay, Ralph Squillace, Sidney Higa, Brian Swan
If you're considering using Windows Azure to host an application, you might wonder if your
application or business requirements are best served by the platform. This topic attempts to
answer this question by:
Looking at the benefits Windows Azure provides to your application
Applying the strengths of the platform to common scenarios
Rejecting scenarios that do not leverage the strengths of the platform
Examining some common architecture and development considerations
The intent is to provide a framework for thinking about your application and how it relates to the
capabilities of Windows Azure. In many cases, links to additional resources are provided to
improve your ability to analyze your application and make a decision on how to move to the
cloud.
Understand the Benefits of Windows Azure
Before you can determine if your application is well-suited for Windows Azure, you must first
understand some of the main benefits of the platform. A complete list of benefits can be found in
the Windows Azure documentation and many articles and videos about Windows Azure. One
excellent paper on this subject is Cloud Optimization Expanding Capabilities, while Aligning
Computing and Business Needs.
There are several benefits to having hardware and infrastructure resources managed for you.
Let's look at a few of these benefits at a high level before we discuss scenarios that take
advantage of these features.
Resource Management
When you deploy your application and services to the cloud, Windows Azure provides the
necessary virtual machines, network bandwidth, and other infrastructure resources. If machines
go down for hardware updates or due to unexpected failures, new virtual machines are
automatically located for your application.
Because you only pay for what you use, you can start off with a smaller investment rather than
incurring the typical upfront costs required for an on-premises deployment. This can be especially
useful for small companies. In an on-premises scenario, these organizations might not have the
data center space, IT skills, or hardware skills necessary to successfully deploy their applications.
The automatic infrastructure services provided by Windows Azure offer a low barrier of entry for
application deployment and management.

5
Dynamic Scaling
Dynamic scaling refers to the capability to both scale out and scale back your application
depending on resource requirements. This is also referred to as elastic scale. Before describing
how this works, you should understand the basic architecture of a Windows Azure application. In
Windows Azure, you create roles that work together to implement your application logic. For
example, one web role could host the ASP.NET front-end of your application, and one or more
worker roles could perform necessary background tasks. Each role is hosted on one or more
virtual machines, called role instances, in the Windows Azure data center. Requests are load
balanced across these instances. For more information about roles, see the paper The Windows
Azure Programming Model.
If resource demands increase, new role instances running your application code can be
provisioned to handle the load. When demand decreases, these instances can be removed so
that you don't have to pay for unnecessary computing power. This is much different from an on-
premises deployment where hardware must be over-provisioned to anticipate peak demands.
This scaling does not happen automatically, but it is easily achieved through either the web portal
or the Service Management API. The paper Dynamically Scaling an Application demonstrates
one way to automatically scale Windows Azure applications. There is also an Autoscaling
Application Block created by the Microsoft Patterns and Practices team.
If your application requires fluctuating or unpredictable demands for computing resources,
Windows Azure allows you to easily adjust your resource utilization to match the load.
High Availability and Durability
Windows Azure provides a platform for highly available applications that can reliably store and
access backend data through storage services or SQL Azure.
First Windows Azure ensures high availability of your compute resources when you have multiple
instances of each role. Role instances are automatically monitored, so it is able to respond
quickly to hardware restarts or failures by automatically deploying a role to a new instance.
Second, Windows Azure ensures high availability and durability for data stored through one of the
storage services. Windows Azure storage services replicate all data to at least three different
servers. Similarly, SQL Azure replicates all data to guarantee availability and durability.
Other Windows Azure services provide similar high availability guarantees. For more information,
see the Windows Azure SLA.
Target Scenarios that Leverage the Strengths of Windows Azure
With an understanding of the strengths of the Windows Azure platform, you can begin to look at
the scenarios that are best suited for the cloud. The following sections discuss several of these
patterns and how Windows Azure is ideally suited for certain workloads and goals. The video
Windows Azure Design Patterns explains many of the scenarios below and provides a good
overview of the Windows Azure platform.
Although there is a focus on application scenarios here, understand that you can choose
to use individual services of Windows Azure. For example, if you find that using blob
Tip

6
storage solves an application problem, it is possible that the rest of your application
remains outside of the Cloud. This is called a hybrid application and is discussed later in
this topic.
Highly Available Services
Windows Azure is well-suited to hosting highly available services. Consider an online store
deployed in Windows Azure. Because an online store is a revenue generator, it is critical that it
stay running. This is accomplished by the service monitoring and automatic instance
management performed in the Windows Azure data center. The online store must also stay
responsive to customer demand. This is accomplished by the elastic scaling ability of Windows
Azure. During peak shopping times, new instances can come online to handle the increased
usage. In addition, the online store must not lose orders or fail to completely process placed
orders. Windows Azure storage and SQL Azure both provide highly available and durable storage
options to hold the order details and state throughout the order lifecycle.
Periodic Workloads
Another good fit for Windows Azure is some form of an "on and off" workload. Some applications
do not need to run continuously. One simple example of this is a demo or utility application that
you want to make available only for several days or weeks. Windows Azure allows you to easily
create, deploy, and share that application with the world. But once its purpose is accomplished,
you can remove the application and you are only charged for the time it was deployed.
Note: You must remove the deployment, not just suspend the application, to avoid
charges for compute time.
Also consider a large company that runs complex data analysis of sales numbers at the end of
each month. Although processing-intensive, the total time required to complete the analysis is at
most two days. In an on-premises scenario, the servers required for this work would be
underutilized for the majority of the month. In Windows Azure, the business would only pay for the
time that the analysis application is running in the cloud. And assuming the architecture of the
application is designed for parallel processing, the scale out features of Windows Azure could
enable the company to create large numbers of worker role instances to complete more complex
work in less time. In this example, you should use code or scripting to automatically deploy the
application at the appropriate time each month.
Unpredictable Growth
All businesses have a goal of rapid and sustainable growth. But growth is very hard to handle in
the traditional on-premises model. If the expected growth does not materialize, you've spent
money maintaining underutilized hardware and infrastructure. But if growth happens more quickly
than expected, you might be unable to handle the load, resulting in lost business and poor
customer experience. For small companies, there might not even be enough initial capital to
prepare for or keep up with rapid growth.
Note

7
Windows Azure is ideal for handling this situation. Consider a small sports news web site that
makes money from advertising. The amount of revenue is directly proportional to the amount of
traffic that the site generates. In this example, initial capital for the venture is limited, and they do
not have the money required to setup and run their own data center. By designing the web site to
run on Windows Azure, they can easily deploy their solution as an ASP.NET application that uses
a backend SQL Azure database for relational data and blob storage for pictures and videos. If the
popularity of the web site grows dramatically, they can increase the number of web role instances
for their front-end or increase the size of the SQL Azure database. The blob storage has built-in
scalability features within Windows Azure. If business decreases, they can remove any
unnecessary instances. Because their revenue is proportional to the traffic on the site, Windows
Azure helps them to start small, grow fast, and reduce risk.
With Windows Azure, you have complete control to determine how aggressively to manage your
computing costs. You could decide to use the Service Management API or the Autoscaling
Application Block to create an automatic scaling engine that creates and removes instances
based on custom rules. You could choose to vary the number of instances based on a
predetermined amount, such as four instances during business hours versus two instances
during non-business hours. Or you could keep the number of instances constant and only
increase them manually through the web portal as demand increases over time. Windows Azure
gives you the flexibility to make the decisions that are right for your business.
Workload Spikes
This is another workload pattern that requires elastic scale. Consider the previous example of a
sports news web site. Even as their business is steadily growing, there is still the possibility of
temporary spikes or bursts of activity. For example, if they are referenced by another popular
news outlet, the numbers of visitors to their site could dramatically increase in a single day. In a
more predictable scenario, major sporting events and sports championships will result in more
activity on their site.
An alternative example is a service that processes daily reports at the end of the day. When the
business day closes, each office sends in a report which is processed at the company
headquarters. Since the process needs to be active only a few hours each day, it is also a
candidate for elastic scaling and deployment.
Windows Azure is well suited for temporarily scaling out an application to handle spikes in load
and then scaling back again after the event has passed.
Infrastructure Offloading
As demonstrated in the previous examples, many of the most common cloud scenarios take
advantage of the elastic scale of Windows Azure. However, even applications with steady
workload patterns can realize cost savings in Windows Azure. It is expensive to manage your
own data center, especially when you consider the cost of energy, people-skills, hardware,
software licensing, and facilities. It is also hard to understand how costs are tied to individual
applications. In Windows Azure, the goal is to reduce total costs as well as to make those costs
more transparent. The paper, Cloud Optimization Expanding Capabilities, while Aligning
Computing and Business Needs, does a great job of explaining typical on-premises hosting costs

8
and how these can be reduced with Windows Azure. Windows Azure also provides a pricing
calculator for understanding specific costs and a TCO (Total Cost of Ownership) calculator for
estimating the overall cost reduction that could occur by adopting Windows Azure. For links to
these calculator tools and other pricing information, see the Windows Azure web site.
Scenarios that Do Not Require the Capabilities of Windows
Azure
Not all applications should be moved to the cloud. Only applications that benefit from Windows
Azure features should be moved to the cloud.
A good example of this would be a personal blog website intended for friends and family. A site
like this might contain articles and photographs. Although you could use Windows Azure for this
project, there are several reasons why Windows Azure is not the best choice. First, even though
the site might only receive a few hits a day, one role instance would have to be continuously
running to handles those few requests (note that two instances would be required to achieve the
Windows Azure SLA for compute). In Windows Azure, the cost is based on the amount of time
that each role instance has been deployed (this is known in Windows Azure nomenclature as
compute time); suspending an application does not suspend the consumption of (and charge for)
compute time. Even if this site responded to only one hit during the day, it would still be charged
for 24 hours of compute time. In a sense, this is rented space on the virtual machine that is
running your code. So, at the time of writing this topic, even one extra small instance of a web
role would cost $30 a month. And if 20 GB of pictures were stored in blob storage, that storage
plus transactions and bandwidth could add another $6 to the cost. The monthly cost of hosting
this type of site on Windows Azure is higher than the cost of a simple web hosting solution from a
third party. Most importantly, this type of web site does not require resource management,
dynamic scaling, high availability, and durability.
Windows Azure allows you to choose only the options that are suited to your business needs.
For example, you might find instances in which certain data cannot be hosted in the cloud for
legal or regulatory reasons. In these cases, you might consider a hybrid solution, where only
certain data or specific parts of your application that are not as sensitive and need to be highly
available are hosted in Windows Azure.
There are other scenarios that are not well-suited to Windows Azure. By understanding the
strengths of Windows Azure, you can recognize applications or parts of an application that will not
leverage these strengths. You can then more successfully develop the overall solution that most
effectively utilizes Windows Azure capabilities.
Evaluate Architecture and Development
Of course, evaluating a move to Windows Azure involves more than just knowing that your
application or business goals are well-suited for the cloud. It is also important to evaluate
architectural and development characteristics of your existing or new application. A quick way to
start this analysis is to use the Microsoft Assessment Tool (MAT) for Windows Azure. This tool
asks questions to determine the types of issues you might have in moving to Windows Azure.
Next to each question is a link called "See full consideration", which provides additional
information about that specific area in Windows Azure. These questions and the additional

9
information can help to identify potential changes to the design of your existing or new application
in the cloud.
In addition to the MAT tool, you should have a solid understanding of the basics of the Windows
Azure platform. This includes an understanding of common design patterns for the platform. Start
by reviewing the Windows Azure videos or reading some of the introductory white papers, such
as The Windows Azure Programming Model. Then review the different services available in
Windows Azure and consider how they could factor into your solution. For an overview of the
Windows Azure services, see the MSDN documentation.
It is beyond the scope of this paper to cover all of the possible considerations and mitigations for
Windows Azure solutions. However, the following table lists four common design considerations
along with links to additional resources.

Area Description
Hybrid Solutions It can be difficult to move complex legacy applications to Windows
Azure. There are also sometimes regulatory concerns with storing
certain types of data in the cloud. However, it is possible to create
hybrid solutions that connect services hosted by Windows Azure
with on-premises applications and data.
There are multiple Windows Azure technologies that support this
capability, including Service Bus, Access Control Service, and
Windows Azure Connect. For a good video on this subject from
October 2010, see Connecting Cloud & On-Premises Apps with
the Windows Azure Platform. For hybrid architecture guidance
based on real-world customer implementations, see Hybrid
Reference Implementation Using BizTalk Server, Windows Azure,
Service Bus and SQL Azure.
State Management If you are moving an existing application to Windows Azure, one
of the biggest considerations is state management. Many on-
premises applications store state locally on the hard drive. Other
features, such as the default ASP.NET session state, use the
memory of the local machine for state management. Although
your roles have access to their virtual machine's local drive space
and memory, Windows Azure load balances all requests across
all role instances. In addition, your role instance could be taken
down and moved at any time (for example, when the machine
running the role instance requires an update).
This dynamic management of running role instances is important
for the scalability and availability features of Windows Azure.
Consequently, application code in the cloud must be designed to
store data and state remotely using services such as Windows
Azure storage or SQL Azure. For more information about storage
options, see the resources in the Store and Access Data section
of the Windows Azure web site.

10
Area Description
Storage Requirements SQL Azure is the relational database solution in Windows Azure.
If you currently use SQL Server, the transition to SQL Azure
should be easier. If you are migrating from another type of
database system, there are SQL Server Migration Assistants that
can help with this process. For more information on migrating
data to SQL Azure, see Data Migration to SQL Azure: Tools and
Techniques.
Also consider Windows Azure storage for durable, highly
available, and scalable data storage. One good design pattern is
to effectively combine the use of SQL Azure and Windows Azure
storage tables, queues, and blobs. A common example is to use
SQL Azure to store a pointer to a blob in Windows Azure storage
rather than storing the large binary object in the database itself.
This is both efficient and cost-effective. For a discussion of
storage options, see the article on Data Storage Offerings on the
Windows Azure Platform.
Interoperability The easiest application to design or port to Windows Azure is a
.NET application. The Windows Azure SDK and tools for Visual
Studio greatly simplify the process of creating Windows Azure
applications.
But what if you are using open source software or third-party
development languages and tools? The Windows Azure SDK
uses a REST API that is interoperable with many other
languages. Of course, there are some challenges to address
depending on your technology. For some technologies, you can
choose to use a stub .NET project in Visual Studio and overload
the Run method for your role. Microsoft provides Windows Azure
SDKs for Java and Node.js that you can use to develop and
deploy applications. There are also community-created SDKs that
interact with Windows Azure. A great resource in this area is the
Ineroperability Bridges and Labs Center.
Deploying projects that use open source software can also be a
challenge. For example, the following blog post discusses options
for deploying Ruby applications on Windows Azure:
http://blogs.msdn.com/b/silverlining/archive/2011/08/29/deploying-
ruby-java-python-and-node-js-applications-to-windows-
azure.aspx.
The important point is that Windows Azure is accessible from a
variety of languages, so you should look into the options for your
particular language of choice before determining whether the
application is a good candidate for Windows Azure.


11
Beyond these issues, you can learn a lot about potential development challenges and solutions
by reviewing content on migrating applications to Windows Azure. The Patterns and Practices
group at Microsoft published the following guidance on migration: Moving Applications to the
Cloud on the Microsoft Windows Azure Platform. You can find additional resources on migration
from the Windows Azure web site: Migrate Services and Data.
Summary
Windows Azure offers a platform for creating and managing highly scalable and available
services. You pay only for the resources that you require and then scale them up and down at
any time. And you don't have to own the hardware or supporting infrastructure to do this. If your
business can leverage the platform to increase agility, lower costs, or lower risk, then Windows
Azure is a good fit for your application. After making this determination, you can then look at
specific architecture and development options for using the platform. This includes decisions
about new development, migration, or hybrid scenarios. At the end of this analysis, you should
have the necessary information to make an informed decision about how to most effectively use
Windows Azure to reach your business goals.

12
Designing Multitenant Applications on Windows
Azure
Authors: Suren Machiraju and Ralph Squillace
Reviewers: Christian Martinez, James Podgorski, Valery Mizonov, and Michael Thomassy
This paper describes the approaches necessary to design multitenant applications (typically ISV
applications that provide services for other organizations) for Windows Azure that are more
efficient that is, either less expensive to run or build, and/or more performant, robust, or
scalable. The paper first describes the general principles of multitenant applications, the structure
and behavior of the Windows Azure platform that is relevant to building and running multitenant
applications, and then focuses on specific areas of multitenancy, especially as they apply to
Windows Azure: application and data resources, as well as the provisioning and management of
those resources. In each area of focus, this paper describes the security, cost, and scalability
issues that you must balance to create a successful multitenant application on Windows Azure.


What is a Multitenant Application?
Wikipedia defines multitenancy in the following way.
If you already know what were talking about in this article, go ahead and start reading
the Windows Azure Multitenant Service Overview to get straight to the guidance. You can
return to this section later if theres something you didnt quite understand.
Wikipedia.org
What does this mean, simply put? Well, an example of this might be a restaurant reservation
application. Restaurants themselves do not have the IT infrastructure to build a full-fledged
internet reservation site that connects with the front desk in nearly real time to make or cancel
reservations. (In most cases, restaurants hire other companies to build their static websites!) So
you might create a single application that enables individual restaurants to sign up, create an
account, customize the look of the web pages with their logos and colors, upload their menu,
connect the web application with the computer at the front desk, handle reservation requests from
other reservation applications, and so on.
In this example, each restaurant sees an application that they use, and that once logged in is
only theirs. Their customers see only the web site of the restaurant, and for all they know, the
restaurant built everything itself. (Its this added value to the hungry customer that the restaurant
is paying the ISV for!)This might look something like the following diagram, in which the
Reservations Application supports multiple tenants in this case, the Contoso and Fabrikiam
Restaurants. In this diagram, Bob, Priyanka, Jasmine, and other customers are pondering
whether to make a reservation or not, perhaps by looking at the menu.
It might look like the following diagram.
Important

13
The Reservations Application diagram.


Architecturally speaking, it is possible to build such an application by hosting a brand new
instance of it for each customer with special customizations for them; perhaps each customer
gets its own virtual machine (VM). This is common enough, but it has many problems for the ISV,
almost all of which appear once your application becomes reasonably popular (well have a bit
more to say on this later, but suffice it to say that this article does not focus on this approach,
which is often referred to as a single-tenant, multi-instance design).
If you want your application to become popular among restaurants and perhaps eventually
among any company needing generic reservation support youll want to build one application
that shares resources among tenants (in this case, restaurants) such that your maintenance costs
and resource costs are dramatically reduced, resulting in better performance and lower cost as
you increase workloads. Building the application this way is what Wikipedia refers to as a
multitenant, single-instance application.
The Application Instance
It is worth noting that instance in this sense refers to the entire logical application, and not a VM,
or an .exe, a process, or anything like that. In fact, the Reservations application diagrammed
above is composed of many different external and internal services components that
communicate through web services.
Such web applications are typically built to work with web farming in which groups of servers
handle multiple instances of important components of the application by storing any application
state data in remote storage while executing, so that if any one instance of an application
vanishes for any reason, another instance can either recover or complete the process. In short,
modern, hugely scalable web applications are decomposed into process components and storage
components that, along with communication and security components, together make up a single,
outwardly facing instance of the overall application.
The Rise of Multitenant Services
Multitenancy on Windows Azure means decomposing applications into stateless services and
storage and managing data flow by identity where necessary: State is only introduced in those
services that must deal directly with the tenants identities. This follows the general service
oriented application principle of Service Statelessness. For example, top-level services like web
pages and public facing web services, or the security services themselves, of necessity must
handle security claims and tokens directly. Other components and services that are part of the

14
larger application, however, should strive to be reusable by any other portion of the application
acting on behalf of any tenant.
Tenants and Their Customers Do Not Care About Multitenancy
This section is here to emphasize one thing: Tenant companies and their customers do not care
how your application is built. They care ONLY about how the application works: its robustness,
performance, feature set, and cost to them. If they care about other things, theyre just nosy
geeks. (We shall concede there are indeed cases in which you may have to demonstrate through
an audit of some kind that you really ARE building and running things correctly for security and
especially legal or regulatory reasons. But these are only the exceptions that prove the rule; those
audits are for compliance, not architecture per se.)
Windows Azure Multitenant Service Overview
The key problem designing a multitenant system is deciding upon the right balance between
providing tenant isolation (on the one hand) and the cost of providing such dedicated resources.
Said a different way, you must figure out just how much of the resources can be shared amongst
your tenants so that your solution works properly but is as cost efficient as possible. Typically, the
more a resource can be shared amongst many tenants, the more cost efficient that solution is
so long as it still satisfies the performance and security requirements.
If at any time youre not clear about what multitenancy is or why its important for
Windows Azure applications, read What is a Multitenant Application?to get some context;
then you can return here.
Within that larger problem, the major considerations center around:
Security - Isolating stored data, authentication & authorization mechanisms
Scaling - Auto-scaling platform compute, scaling platform storage
Management and monitoring
Provisioning tenant resources
Metering and billing
Observe that devising a multitenant architecture is an involved process, with many moving parts
and this article does not attempt to address every detail, but rather put together a perspective on
the major considerations with as many commonly helpful tips and pointers to the details as
possible. With those points in mind, lets start to explore the architecture from a high level.
High Level Architecture
One approach to supporting multiple distinct customers is to reject the multitenant route
completely and allocate resources for each customer on demand. When you use this approach
called single tenant -- you are automating the provisioning of dedicated resources for customers.
If you imagine a spectrum of distributed application architectures, this approach represents
completely tenant-dedicated resources on one side, with completely shared resources on the
other. This paper attempts to describe how to design distributed applications that are closer to the
Note
comp
multite
When
isolati
applic
exclus
propo
and m
resou
emph
canno
flexibi
Whic
The fi
some
resou
and d

In this
layer
Blobs
often
patter
depen
letely shared
enant) applic
n you build a m
ion you want
cation and dat
sively to custo
osition when it
most likely will
urces are excl
hasizing this a
ot be and to
ility. Use it.
ch Resourc
igure below il
of the decisio
urces in the di
define where i
s diagram, Wi
resources, as
s, and SQL Az
serves as tem
rns, and as a
nding upon th
portion of thi
ation and dat
multitenant ap
to provide ea
ta resources
omers, but it's
t comes to re
l -- mix and m
usive to indiv
again: You ca
o what extent
ces Can B
lustrates the W
ons behind ea
agram (and m
t sits within th
indows Azure
s are relay en
zure are clear
mporary stora
result can be
heir usage.
s design spec
ta resources.
pplication, yo
ach customer.
are shared be
s important to
source sharin
match which re
vidual custom
n decide whic
they are shar
e Shared?
Windows Azu
ach of these a
multitenant ar
he application
e compute an
ndpoints, whic
rly data layer
age that enab
e seen as both
ctrum: distribu
u need to we
. Its easy to t
etween custo
o realize that
ng. In a servic
esources are
ers dependin
ch application
red. This give
?
ure resources
as motivation
rchitectures in
n or data tier.
d the Access
ch serve to co
resources. C
les certain ap
h storage or a
uted applicati
igh the costs
think that you
omers and wh
being multiten
ce-oriented ap
shared betwe
g on your req
n services can
es you a treme
s under consid
n. One helpful
n general) is t

Control Serv
onnect compu
Caching, howe
pplication des
as application
ions with sha
of resources
will need to c
hich are alloca
nant is not an
pplication, yo
een custome
quirements. It
n be shared, a
endous amou
deration, and
way to look a
o take each c
vice are clearl
uting instance
ever, along w
signs and com
n layer resour

1
red (or
against the
consider whic
ated
n all-or-nothin
u are able to
rs and which
s worth
and which
unt of design
well explore
at all the
component

y application
es. Tables,
ith queues,
mmunication
rces,
15
ch
g

e

16
Furthermore, exactly which approach you can take in any one case depends upon how many
features you intend to expose to your tenant. For example, if you want to allow your tenant to
offer a tenant-specific identity to its clients, you may want to have a general and multitenant-
safe username/password security token service (STS). But you might also enable your tenant
to implement one, and federate with Access Control Service instead. Another example is whether
you intend to allow your tenants customers to upload and execute code in the tenants web or
worker role instances. Doing so may well require a particular multitenant solution that mixes and
matches multitenant and single-tenant role instances.
With these ideas in mind, lets consider how to successfully share the resource. Although well
discuss in this and other topics the general approaches you can take to supporting multiple
tenants in one resource from the point of view of an application, well also discuss how to
provision and manage those resources, and how to use Azure to do so. Provisioning and
managing resources that are shared does not pose great problems, but in some cases its worth
a closer look.
Application Resources
Designing a multitenant application means making sure that to the extent possible compute, data,
and other application components handle multiple tenant identities and potentially each tenants
customer identities as well. This section discusses the issues involved in various approaches
using Windows Azure and suggests the best approaches to handling them in your application
design.
Compute
Lets begin by looking at the application resources provided by Windows Azure web and worker
roles, and how they should be used in a multitenant solution to host web services, web sites, and
other general processing.
The clearest use of Windows Azure web and worker roles in a multitenant application is to
provide additional (or to reduce!) compute resources on demand in the service of the tenants,
with the notion that a single role instance serves multiple tenants.
Segment Web site Tenants by Host Headers
In the most coarse-grained solution, the approach is to configure each web role instance to
support multiple web sites and web applications, perhaps one for each tenant. This can be
accomplished by modifying the Sites element in the Service Definition. Each website can either
be bound to the same endpoint and use the HTTP/1.1 Host header to isolate them or they can be
bound to their own endpoints. Note, however, that binding each web site to its own endpoint
doesnt scale per tenant since Windows Azure limits you to 5 input endpoints per role instance).
If you use HTTP/1.1 Host headers to support multiple tenants, the tenants (and their customers)
simply visit a different a URL (such as www.fabrikam.com for one and www.contoso.com for the
other tenant) and get the website for the desired tenant, even though they are in fact to talking to
a single endpoint within a single web role instance. This approach allows you to isolate different
web site files by tenant with minimal effort. As an additional benefit, this approach enables you to
isolate tenants by application pool because by default each web site gets its own application pool.

17
Note, however, that because the URLs that provide the HTTP/1.1 Host headers are defined in the
domain name system (DNS), you will need to configure the DNS with your domain provider and
point it to your *.cloudapp.net address; remember that it may take 24 hours for your new tenant
URL to be fully discoverable on the internet. For a great example of how one instance can host
multiple websites differentiated by host headers, check out the Windows Azure Accelerator for
Web Roles.

SSL Communication
One cautionary note to using the HTTP/1.1 Host headers approach is that Windows Azure does
not enforce a security boundary between these websites; this means that SSL certificates used
by one website are useable from the other websites. To complicate matters, IIS 7 does not
support having multiple websites with their own certificate if only their host headers are different.
It is possible to have multiple SSL Web sites that use unique server certificates if each
Web site uses a separate IP address or port for its HTTPS binding. As in IIS 6.0, IIS 7.0
does not support having multiple Web sites with their own server certificates if the HTTPS
bindings for those Web sites are on the same IP address/port and differentiate only by
using host headers. This is because the host header information is not available when
the SSL negotiation occurs. Because of this, the only way to have multiple Web sites on
the same IP address that use host headers is to configure all of those Web sites to use
the same SSL certificate with a wildcard CN. For more information, see
http://www.microsoft.com/technet/prodtechnol/WindowsServer2003/Library/IIS/596b9108
-b1a7-494d-885d-f8941b07554c.mspx?mfr=true.
Therefore, if you are using SSL certificates, you may need to take a different approach. One is to
use a single certificate with a wildcard CN (of the form *.yourdomain.com) or a Unified
Communications Certificate (where you list all the subdomains explicitly). These types of
certificates are available from certificate authorities like GoDaddy and Thawte and are
straightforward to create. Each tenant would then access resources over SSL using URLs like
https://fabrikam.yourdomain.com or https://contoso.yourdomain.com.
The alternative is to have each website live on a dedicated web role and only the one tenant
certificate. As this is not a multitenant approach, it does not scale as well and therefore absorbs
more compute resources as well as management overhead. It nevertheless must sometimes be
done.
Segment Website Tenants by Query Parameters
Another approach is to use a single website to serve resources for multiple tenants. If your
website is built using ASP.NET MVC, you could create a separate MVC area for each tenant.
Alternatively, you share the default area between them and instead have the controllers render
differently based on query parameters (such as a tenant id). If you want to provide process
isolation by tenant, then you need to define virtual applications within the website. Effectively,
each tenant gets a virtual application underneath the website. This is also accomplished by
editing the Sites element within the Service Definition, but instead of adding a new Site for each
tenant, within your main Site element you add one VirtualApplication element for each tenant.
Important

18
Web Services in Worker Roles
For fully multitenant solutions, worker roles hosting web services behave the same way as their
relatives hosted in a web roleall tenants access the same service endpoints.
Worker Roles are also used to provide background processing. In the fully multitenant scenario,
the approach you take is to have a worker role whose RoleEntryPoint.Run method loop
processes work items agnostic of which tenant created it -- hence you are sharing the compute
resources across all tenants.
Storage
Multitenant applications must handle how the data they store and use is isolated and yet performs
well. This section discusses data storage technologies in Windows Azure, including SQL Azure
databases; Windows Azure blobs, tables, and queues; Service Bus queues and topics, and the
Caching service in a multitenant application.
SQL Azure storage is clearly an important location in acting as the storage for an application, but
there use in a multitenant architecture extends to acting as storage for provisioning new tenants
and for storing management data.
Using SQL Azure for Application Resources
In this scenario application data for multiple tenants are stored within a single logical database.
While there are obvious cost benefits to delivering a multitenant database (realized by splitting
the cost across all tenants), these come with the tradeoff of increased complexity in dealing with
various requirements specific to this scenario:
Adequately isolating data and preventing one tenant from inadvertently or maliciously
accessing another tenants data.
Maintaining reasonable query performance for all tenants, in addition to limiting the effect of a
single, especially resource-hungry tenant.
Handling the rapidly increasing size of a shared database, which may exceed the maximum
50 GB database size that SQL Azure currently supports.
Conceptually, there is a performance benefit gained when distributing tenant data and query
processing across multiple databasesthe question becomes, what architectures allow you to
leverage this for your solution? Before we answer that question, its is helpful to look at the
architectures on a spectrum. On one end of this spectrum is the traditional scale-up, in which you
simply throw bigger hardware at the problem. As with the single-tenant architectural approach, we
wont focus on scaling-up except where it helps understanding, because it is too limiting for
multitenant solutions. We focus on the various scale-out approaches using sharding techniques:
Linear Shard
Expanded Shard
Compressed Shard and
Hybrid Linear/Expanded Shard architectures.

All of
scale-
to as
define
the sa
persp
Retur
mode
multip
shard
spect
expan
The o
there
Simila
made
where
partic
words
shard
Librar
with fa
shard
imple
steps
Secu
We be
the da
amou
object
per-te
schem
sec
the preceding
-out, and for e
a data domai
ed using certa
ake of simplic
pective that th
rning to the sp
el. In this appr
ple tenants pe
d architecture,
rum, each ten
nded shard ap
other two appr
are fewer SQ
arly, scenario
e (such as a re
e each region
cular tenants
s, the architec
ding approach
ry, that knows
anning out qu
d.SQL Azure F
mentation of
you can take
ring Multiten
egin by lookin
atabase level
unting to creat
ts (schemas,
enant SQL Se
ma objects us
urity applied t
g four useful s
each there is
in or data fed
ain key values
city in this artic
e data domai
pectrum diagr
roach all tena
er database. I
, in which eac
nant may get
pproach.
roaches addr
QL Azure data
s where there
egion and ten
n gets a partic
data is sprea
cture scales o
hes necessitat
s how to navig
ueries to all sh
Federations c
sharding arch
e to implemen
nant Data in S
ng at how to s
for SQL Azu
ting security p
tables, views
erver logins, D
sing these sec
to schemas re
sharding tech
some logical
eration) that d
s (such as a c
cle, and in the
in is the tenan
ram, on one e
nts share the
n the middle
ch tenant gets
multiple SQL
ress in-betwee
abases than t
e are multiple
nant), then we
cular expande
d across mult
out linearly by
te a need for
gate the shar
hards and gu
coming out the
hitectures imp
nt these archit
SQL Azure
secure your te
re is almost id
principals and
s, stored proc
Database Use
curity principa
estrict access
hniques add d
separation b
defines the sh
customer id) o
e context of m
nt.
end of the spe
e same SQL A
of the spectru
s its own SQL
L Azure datab
en scenarios.
enants, but m
data domain
e have Hybrid
ed shard each
tiple database
y region, but in
a layer of abs
ding approac
iding data mo
e second half
plemented on
tectures toda
enants data. I
dentical to tha
d using them t
cedures, and s
ers, and Data
als. One appr
s to just that t
databases to t
etween the d
hard. These d
or an attribute
multitenant arc
ectrum we ha
Azure databas
um diagram,
L Azure datab
bases, which i
. In a compre
more than one
ns with which
d Linear/Expa
h shard having
es in that exp
n an expande
straction, like
ch implemente
odifications op
f of 2011 prom
n SQL Azure,
y.
In its core for
at of an on-pr
to control acc
so on). Then
base Roles, a
roach is to def
enant.
the architectu
ata stored (of
data domains
e (such as the
chitectures, w
ave the traditio
se. In other w
above, we ha
base. On the f
is referred to
ssed shard a
e SQL Azure d
sharding dec
anded Shard a
g multiple dat
panded shard
ed fashion by
e the Enzo SQ
ed, and in par
perations to th
mises to simp
but lets cons
m, managing
remise SQL S
cess to specif
you create th
and apply per
fine one sche

1

ure to support
ften referred
can be
e year). For
we take the
onal scale-Up
words we have
ave the linear
far end of the
as the
rchitecture
database.
cisions are
architecture
tabases, and
. In other
tenant.These
QL Shard
rticular assists
he appropriat
plify the
sider what
security at
Server,
fic schema
he appropriate
rmissions to
ema per tenan
19
t
p
e
e
a
e
s
te
e
nt

20
Another approach is to store all tenant data within the same set of tables, separating them within
the tables by some key (such as a tenant id column), thereby creating a logical separation. In this
approach logic in the business tier is responsible for correctly filtering data accesses by tenant
(albeit SQL Azure Federations can also perform this task). The crux of this approach is enforcing
that all requests for data must go through this logic in order to limit access to the correct tenants;
OData (using WCF Data Services that use Query and Change Interceptors) is a great technology
with which to implement this approach. There are even some frameworks, like Enzo (discussed
later) that relieve you from having to implement this logic.
With this in mind, lets apply this to the scale-out architectures previously described. Security in
the Linear Shard architecture can be pretty simple: each tenant gets its own database, and as
such gets its own dedicated logins and users specifically for that shard. In all of the other
sharding architectures, security needs to be more granular and relies on a combination of
permissions defined at the schema object level and enforcement at the business tier.
Performance Considerations
Next, lets turn our focus to performance. SQL Azure maintains three replicas of any SQL Azure
database. While data for each database is replicated between one primary and two secondary
replicas, all reads and writes take place on the primary replica- the secondary replicas only help
out with query processing in a fail over scenario. The SQL Azure fabric determines the location of
the primary and two secondary replicas. If a machine hosting your primary replica is under heavy
load (due to your own tenants or other tenants), the SQL Azure fabric may switch the primary for
a secondary on another machine that has a lighter work load. This switch happens quickly, but
does result in the disconnection of active sessionsa situation which your multitenant application
must be prepared to handle.
With this in mind, if per-tenant performance is a top priority, then you should consider a sharding
architecture that allocates one or more databases per tenant (such as the Linear, Expanded or
Hybrid architectures). However, if you have a lot of tenants for which you allocate a database this
way, you run into a few obstacles. Obviously, each database you create has a cost, but the
obstacle to be aware of is the limit of 149 databases per SQL Azure Server, which can be raised
if you call Cloud Support Services (CSS). For multitenant scenarios with lots of databases, your
solution also needs to be able to allocate SQL Azure Servers as this limit is approached.
Using Windows Azure Tables for Application Resources
Windows Azure Tables are secured using a storage account name and key, which can only be
allocated globally for all storage resources. This means that to provide isolated storage of per-
tenant data the service provider will have to create a storage account for each tenant. This can be
performed through the Windows Azure Service Management API. There is no additional cost to
allocating new storage accounts and there is a performance benefit to be gained by doing so
because different resources can be applied to different accounts. That said, there are quotas
governing the number of storage accounts a subscription can have, which can be adjusted by
contacting Cloud Services Support (CSS).
If your application needs to store multiple tenants worth of data within a single account, you have
to create your own abstraction to filter the data returned to a given tenant. One approach is to
create one table per tenant, and authorize access at a per table grain. This approach is beneficial
in that you get to leverage the full partition key and row key to quickly access the data, with a

21
fairly simple logic for routing queries to the correct table. The downside is the extra complexity in
that you have to route queries to the appropriate table based on the tenant. Another approach is
to partition horizontally by tenant within the table, where your data access layer only returns data
having a certain partition key to the client. This eliminates the need to juggle routing queries to
the appropriate table, but limits you to a single partition key value, which in turn can affect your
performance on very large per-tenant datasets.
Using Windows Azure Blobs for Application Resources
When it comes to the storage needs of the application, Windows Azure Blobs offer a great deal
for multitenant applications. You should create public blob containers for read-only access to
shared data like website images, graphics and tenant logos. Private blob containers (only
accessible by the tenants or the tenants and the system) would be used for application data like
documents or per-tenant configuration files. Outside of anonymous access, they key approach is
to be sure to specify a container-level access policy for containers or blobs secured using Shared
Access Signatures for various actions such as blob read/write, block list, properties, or metadata,
deleting a blob, leasing a blob and enumerating blobs within a container. By specifying a
container level access policy, you can the ability to adjust permissions without having to issue
new URLs for the resources protected with shared access signatures.
Windows Azure Queues for Application Resources
Windows Azure queues are commonly used to drive processing on behalf of tenants, but may
also be used to distribute work required for provisioning or management.
Here, the consideration is whether a given queue manages items belonging to multiple tenants,
or if each queue is dedicated to a single tenant. The Windows Azure Storage model works at the
storage account level only, therefore if queues need to be isolated by tenant, they need to be
created under separate storage accounts just as for tables. Moreover, in the case of a single
queue, its not possible to define permissions restricting use to only put or only receive
messagesonce a tenant has access with the storage account the tenant can do everything
allowed by the Azure Queue API, even delete the shared queue! In short, the queues should not
be exposed to the tenant, but rather managed by the service automatically.
Service Bus Queues for Application Resources
For tenant specific application functionality that pushes work to a shared a service, you can use a
single queue where each tenant sender only has permissions (as derived from claims issued from
ACS) to push to that queue, while only the receivers from the service have permission to pull from
the queue the data coming from multiple tenants. The reverse direction of this is possible with AF
Queues, you can have system services push messages destined for specific tenant receivers
across a single queue. You can also multicast these messages by using the Subscription
functionality of AF Queues. Finally, you can use Session Messages to drive customer specific
work to customer-specific receivers, albeit only pushing through a single, shared queue.
While these approaches minimize the number of queues you have to manage (which to a certain
extent reduces your overall system complexity), it can limit the throughput of enqueue or dequeue
operations because the load is not partitioned amongst more resources (as would be the case
when using multiple queues). Also, in the case of multiple tenant senders, you must be careful
not to be at the mercy of a single tenant who floods the queue and effectively starves other

22
tenants of downstream processing. With AF Queues you can defer messages, so it is also
possible to detect such a flood coming from a single tenant, defer those messages temporarily to
process messages from other tenants and later return to processing them.
Cache Service for Application Resources
Within a multitenant application, the Cache is traditionally used for frequently accessed tenant-
specific application data. As it does not provide the ability to set distinct permissions within a
single cache, the pattern is to provision a distinct cache per tenant. This implies provisioning a
new AppFabric namespace for the tenant and creating the cache within that. At present, there is
no management API to automate Cache creation and this still needs to be done by human
operator using the Windows Azure Portal.
Connection and Security Services
Windows Azure multitenant applications use other middleware services for connectivity and
security: Service Bus relay services and the Windows Azure Access Control Service (ACS).
Using ACS for Application Resources
Windows Azure Access Control Service
read about it here
Whether you are building a multitenant system or not, ACS is the way to secure access to the
application itself. Using ACS enables your application to be built using one authentication API
(the ACS API) that can process security tokens from any identity provider. This means you dont
write separate code each for Facebook, Google, Yahoo, Windows Live, a different third-party
identity provider, or Active Directory Federation Server (ADFS). Instead, you only write code for
ACS, and let ACS and the identity providers do the heavy lifting for you.
Provisioning Identities in ACS
The details of configuring ACS or implementing your own STS are beyond the scope of this
document, but you can get started thinking about such things in Authorization in Claims-Aware
Web Applications and Services. To get the fundamentals of the approach used to outsource
security from your application via federated security, see this great guide. To get a start on
building your own, you might want to get the background on how to build one using Windows
Identity Foundation (WIF) from here, but you will likely benefit by starting with a fully fleshed out
STS like Thinktectures Starter STS, and customizing it to suite your needs.
Within a multitenant application that uses ACS, identities are usually provisioned with the
following high-level steps:
Creating certificates (for example, per tenant)
ACS namespace provisioning (if isolating by namespace), including the configuration of
tenants application that is to be secured -- called a relying party (RP) -- and claims
transformation rules. This is done using ACS management APIs or the ACS Portal.
Creating root administrator account for tenant to login with (using API of an identity providing
security token service).


23
Using Service Bus Relay for Application Resources
The services that are exposed as endpoints may belong to the tenant (for example, hosted
outside of the system, such as on-premise), or they may be services provisioned specifically for
the tenant (because sensitive, tenant-specific data travels across them). In either scenario,
handling multiple tenants is really not the issue; enforcing tenant-specific usage is. Access to
these endpoints is secured using Access Control Service (ACS), where clients must present an
Issuer Name and Key, a SAML token, or a Simple Web Token. This can be configured
programmatically using the Service Bus and ACS management APIs.
Remember that Service Bus queues and topics and subscriptions are discussed as
storage, though they often serve to handle application data flow in a particular way.
Provisioning Resources
This section discusses the provisioning of resources in a way that supports multitenant
applications. As with supporting more than one tenant in application design, designing multitenant
provisioning also has several decision points, depending upon the features you intend to enable
and what Windows Azure services your application uses.
As with the design of multitenant applications, well discuss how compute, storage, and
connectivity and security services are used to provision multitenant applications.
Provisioning and Managing Resources Using Azure Roles
A dedicated worker role in multitenant solution is typically used to provision and de-provision per
tenant resources (such as when a new tenant signs-up or cancels), collecting metrics for
metering use, and managing scale by following a certain schedule or in response to the crossing
of thresholds of key performance indicators. This same role may also be used to push out
updates and upgrades to the solution.
A web role is typically dedicated to enable the service provider to monitor and manage system
resources, view logs, performance counters, and provision manually, and so on.
Using ACS for Provisioning Resources
When provisioning tenant resources protected by ACS, you will need to use the ACS
management APIs or portal to create the initial admin" account for newly provisioned tenants.
Using Cache Service for Provisioning Resources
When provisioning tenant resources protected by ACS, you will need to use the ACS
management APIs or portal to create the initial admin" account for newly provisioned tenants.
Considerations for Provisioning Storage
One consideration that applies to Windows Azure Tables, blobs, and SQL Azure in a multitenant
solution is geo-distribution. Specifically it is identifying data that must be unique system-wide
(across all data centers used in the solution) and balancing that against maintaining a performant
Note

24
user experience. One option is to build a custom replication strategy to bring such shared data
near to end users (which naturally has to ensure that new data is unique, perhaps by always
inserting to the master data source). The other option is to partition the data, such that the
amount and frequency of access of global data is minimized.
Another consideration for Azure Storage in general is the hard limits imposed by Azure, which
albeit large are not infinite. When planning your multitenant solution, you should take these
scalability limits into account.
Using SQL Azure for Provisioning Resources
In some multitenant scenarios with large amount of data involved, it is best to provision new SQL
Azure databases by copying from an existing SQL Azure reference instance. The enhanced
provisioning speed provided by this must be weighed against the cost of maintaining an extra
SQL Azure database on top of those required by tenants and the system itself.
Provisioning SQL Azure Resources
The options for provisioning new SQL Azure resources for a tenant include:
Use DDL in scripts or embedded as resources within assemblies
Create SQL Server 2008 R2 DAC Packages and deploy them using the API's. You can also
deploy from a DAC package in Windows Azure blob storage to SQL Azure, as shown in this
example.
Copy from a master reference database
Use database Import and Export to provision new databases from a file.

Provisioning Windows Azure BLOB Storage
The approach for provisioning BLOB storage is to first create the container(s), then to each apply
the policy and create and apply Shared Access Keys to protected containers and blobs.
Using Windows Azure Blobs for Provisioning Resources
When it comes to provisioning compute or pre-initialized storage resources for new tenants,
Azure blob storage should be secured using the container level access policy (as described
above) to protect the CS Packages, VHD images and other resources.
Management Resources
Finally, the design of a multitenant application must tackle the extremely important task of
managing the application, tenants and their services, all the data resources, and any connectivity
or security issues they entail. This section discusses common uses of compute, data, and other
services to support multitenant applications while running Windows Azure.

25
Using Windows Azure Roles for Management Resources
The service provider will need a way to monitor and manage the system resources. A web role is
typically dedicated to provide the service provider with tools to manage resources, view logs,
performance counters, and provision manually, etc.
Using ACS for Management Resources
Most multitenant systems will require a namespace created within ACS that is used to secure
system resources, as well as the creation and management of individual namespaces per tenant
(such as for using the AF Cache). This is also accomplished using the ACS management
namespace.
Using Cache Service for Management Resources
If the service provider exposes certain KPIs or computed statistics to all tenants, it may decide to
cache these often requested values. The tenants themselves do not get direct access to the
cache, and must go through some intermediary (such as a WCF service) that retains the actual
authorization key and URL for access to the Cache.
Using SQL Azure for Management Resources
Examples of these are single, system wide and datacenter agnostic membership/roles database
for non-federating tenants or those relying on a custom IP STS configured for use with ACS. For
multitenant systems concerned with multiple geographic distributions, the challenge of a
centralized system for management data surfaces. To solve these problems you can take the
approach previously described for application resources, either by defining your geographical
regions as shards in Hybrid Linear/Expanded Shard architecture or a more simple Linear Shard
architecture. In both cases, leveraging middle-tier logic to fan out and aggregate results for your
monitoring and management queries.
Using Windows Azure Tables for Management Resources
The Windows Azure Diagnostics (WAD) infrastructure by default logs to Windows Azure Tables.
When relying on these WAD tables (Trace, Event Logs and Performance Counters) you need to
consider just how sensitive the logged data may, who has access to them and ultimately choose
if they are isolated (aka provisioned) between customers or shared system-wide. For example,
its unlikely that you would provide all tenants direct access to the diagnostic log tables which
aggregates traces from all instances across the solution and might expose one tenants data to
another tenant.
Using Windows Azure Blobs for Management Resources
The canonical management resources stored in blob storage are IIS Logs and crash dumps. IIS
Logs are transferred from role instances to blob storage. If your system monitoring relies on the
IIS Logs, you will want to ensure that only the system has access to these logs via a container
level access policy. If your tenants require some of this data, you will want to perform some post
processing on the data (perhaps to ensure that only that tenants data is included) and then push

26
the results out to tenants via a tenant specific container. Crash dumps, on the other hand are
something only the service provider system should have access to as they assist in
troubleshooting the infrastructure and will very likely contain data that spans tenants.
Metering
Within the context of multitenant applications, metering is motivated by a desire to bill charges to
tenants that are influenced by usage as well as to collect system-wide KPI's to inform capacity
planning and on-going architecture decisions. What is metered? Typically it boils down to these:
Raw resource consumption: Azure and AppFabric resource use (e.g., compute hours, data
storage size, number of messages)
Specific use of applications features (for example, premium operations charged per use)
Use by tenants own users
The latter two tend to be driven by application specific requirements, so we will focus on the raw
resource consumption that is common to all Azure based service providers. From a raw resource
perspective, for most SaaS applications, compute hours is by far the most significant, followed by
storage size and to a lesser degree, data transfer out from the data centers (egress) -- particularly
as data transfers into Azure datacenters is now free.
So how can you get this data for metering on raw compute or storage? Lets explore some
examples.
Compute Hours
Unfortunately there is no API at present to query for this by the service provider's subscription.
The best approach is to approximate usage in line with how Windows Azure compute time is
billed. For example, for each instance allocated, compute the hours of instance uptime, rounding
up to the nearest hour multiplied by the hourly cost for the instance size.
Data Storage
Again, there is no public billing or management API for Windows Azure Storage (Blobs, Tables,
and to a lesser extent Queues) or Cache and Service Bus Queues) that provides exact usage,
but one can extrapolate by knowing the size of the entities being stored and tracking the number
of entities stored by the tenant on average by the tenant over the billing period. SQL Azure
database size is the exception; You can determine the database size that for which you will be
billed that day by querying sys.database_usage within the master database and aggregating the
result over the month to get at the actual cost.
Scaling Compute for Multitenant Solutions
While specific requirements vary, as a general theme when "auto-scaling" is considered the
approach amounts to increasing or decreasing the instance count according to some heuristic.
This heuristic may depend on some key performance indicator (KPI) derived from sources such
as performance logs, IIS logs or even queue depth. Alternately, it may simply be implemented in
response to a schedule, such as incrementing for a month end burst typical in accounting
applications, and decrementing the instance count at other times.

27
The key factor to consider here is that scaling the instance count up or down is not instantaneous.
Your algorithms should incorporate this in two ways. When scaling up, its best if you can
anticipate the demand-- it doesn't have to be a long range forecast (but on the order of 20
minutes) to allow for your instances to become available. When scaling down, recall that you pay
for partial hour used as a complete hour, so it makes economic sense to keep those unneeded
instances around for that full hour.
Conclusion and Resources

Frameworks and Samples
Having read through the above, you probably agree that building a multitenant solution is a big
investment. From this perspective, starting from a sample or framework is a great idea. Here are
some great starting points.
Microsoft Enterprise Library 5.0 Integration Pack for Windows Azure
Microsoft Enterprise Library is a collection of reusable application blocks that address common
cross-cutting concerns in enterprise software development. The Microsoft Enterprise Library
Integration Pack for Windows Azure is an extension to Microsoft Enterprise Library 5.0 that can
be used with the Windows Azure technology platform. It includes the Autoscaling Application
Block, Transient Fault Handling Application Block, blob configuration source, protected
configuration provider, and learning materials. This application block is a great place to start.
CloudNinja
The CloudNinja Sample available on CodePlex demonstrates the following aspects of multi-
tenancy as discussed in this article.
Multitenant Web Roles
Dedicated Management/Provisioning Worker Roles
Linear Sharding in SQL Azure
Dedicated Windows Azure Tables for Management
Dedicated Azure Queues for Provisioning
Dedicated Public/Private Blob Storage
Multitenant AppFabric Cache
Time & KPI Based Auto-Scaling
Metering
Federated Security with a Custom STS and ACS

Fabrikam Shipping
The Fabrikam Shipping Sample sample available on Codeplex demonstrates many of the aspects
of a multitenant SaaS offering.
Dedicated & Multitenant Web Roles

28
Dedicated Management/Provisioning Worker Roles
Linear Sharding in SQL Azure
Dedicated Public/Private Blob Storage
Federated Security with a Custom STS, ADFS, Facebook, Google, Live ID and ACS
PayPal integration

Enzo SQL Shard Library
The Enzo SQL Shard Library available on CodePlex demonstrates and assists with the numerous
SQL Azure sharding techniques discussed in this article.
Lokad Cloud
Lokad Cloud is a feature rich framework that provides lots of the functionality as described in this
article and a lot more.
Auto-scaling
Strongly typed Azure Blob, Table and Queue storage (they call this an Object to Cloud
mapper).
Task scheduler
Logging
Azure Auto Scaling
If youre looking for a simple, straightforward example from which to build your own auto-scaling,
the Azure Auto Scaling sample available on CodePlex is worth a look.Of course, there are also
some commercial offerings available to help with some of the more difficult aspects. In particular,
Paraleap AzureWatch can help you auto-scale your application.
Links and References
Multitenant Architecture
Developing Applications for the Cloud on the Microsoft Windows Azure Platform
Hosting a Multi-Tenant Application on Windows Azure
Building a Scalable, Multi-Tenant Application for Windows Azure
Architecture Strategies for Catching the Long Tail
Multi-tenancy in the cloud: Why it matters
The Force.com Multitenant Architecture: Understanding the Design of Salesforce.coms
Internet Application Development Platform
Patterns For High Availability, Scalability, And Computing Power With Windows Azure
Related Topics
GeekSpeak: Autoscaling Azure
Performance-Based Scaling in Windows Azure
Accounts and Billing in SQL Azure
Security Resources for Windows Azure
How to Configure a Web Role for Multiple Web Sites

29
How to Configure a Web Role for Multiple Web Sites
Federations: Building Scalable, Elastic, and Multi-tenant Database Solutions with SQL Azure
Managing Databases and Logins in SQL Azure
Inside SQL Azure


30
Packaging and Deploying an Application to
Windows Azure
Authors: Larry Franks, Rama Ramani
This document provides guidance on deploying an application to a Windows Azure hosted
service. It also provides guidance on working with other Windows Azure services that your
application may use, such as SQL Azure and Windows Azure Storage.
Before deploying an application to Windows Azure, you should have an understanding of the
following:
Differences between the Windows Azure Emulator and Windows Azure, SQL Azure, and
Windows Azure Storage
How to configure:
Connection strings for Windows Azure Storage Services and SQL Azure
Endpoints
Role size
Number of instances
How to create an affinity group
Microsofts SLA requirements for hosted services
Development and Production environments for hosted services
How to deploy an application using the Windows Azure Management Portal
Differences between Windows Azure and the Windows Azure
Emulator
The Windows Azure SDK installs the Windows Azure Emulator, which emulates the Windows
Azure hosting and Storage services. Before deploying an application to Windows Azure, you
should first perform testing in the Windows Azure Emulator. While the emulator provides an easy
way to test a hosted application during development, it cannot fully emulate all aspects of the
Windows Azure platform. For example, the connection string used to connect to Windows Azure
Storage differs between the Windows Azure Emulator and Windows Azure. Before deploying an
application to Windows Azure, you should understand the differences between the emulator and
Windows Azure and ensure that your application does not rely on a behavior of the emulator that
is not present in the Windows Azure environment.
For more information on the differences between the emulator and the Windows Azure platform,
see Overview of the Windows Azure SDK Tools.
SQL Azure and Other Windows Azure Services
While the Windows Azure Emulator provides a local testing solution for hosted services and
storage, it does not provide any development equivalent for SQL Azure, the Caching Service,
Service Bus, and other services provided by the Windows Azure Platform.

31
For database design and testing, you can use SQL Server; however there are differences
between SQL Server and SQL Azure that you must be aware of. For a comparison, see Compare
SQL Server with SQL Azure.
If your solution is developed against SQL Server, you should consider whether you will recreate
your databases and associated artifacts in SQL Azure or migrate your SQL Server development
environment to SQL Azure. For information on migration options, see Migrating Databases to
SQL Azure.
For other services such as Caching or Service Bus, you must develop against the live Windows
Azure service.
Pre-Deployment Checklist
The following items should be verified before deploying an application to Windows Azure:

Item to check Description
Number of instances At least two instances must be created to meet
the Windows Azure Compute Service Level
Agreement (SLA) requirements. For more
information on Windows Azure SLAs, see
Service Level Agreements.
Connection strings All connection strings should be checked to
ensure they do not reference development
storage
Virtual machine size The virtual machine size governs available
memory, local storage, processor cores, and
bandwidth for your application. For more
information, see How to Configure Virtual
Machine Sizes.
Endpoints Endpoints determine the ports used for
communications with your hosted services, and
whether the port is public or for internal use
only. For more information, see How to Enable
Role Communication.
Affinity group To ensure that you are deploying to the correct
datacenter, you should consider creating an
affinity group for your project and use this when
provisioning services or deploying to the
Windows Azure platform. If you do not use an
affinity group, you may accidentally deploy
services to different datacenters, which can
impact performance and increase costs.
For more information, see How to Create an

32
Item to check Description
Affinity Group.
Certificates If you wish to enable SSL communications, or
remote desktop functionality for your hosted
service, you must obtain and deploy a
certificate to Windows Azure. For more
information, see How to Add a Certificate to the
Certificate Store and Using Remote Desktop
with Windows Azure.
Co-Administrators Ensure that the co-administrators for your
Windows Azure subscription contain the
appropriate people. For more information, see
How to Add and Remove Co-Administrators for
your Windows Azure Subscription.
Upgrade planning You should familiarize yourself with the
information in the Post-Deployment section of
this article before deployment, as part of
designing your Windows Azure based solution
is creating an upgrade plan.

Deployment
There are three primary methods of deploying an application to Windows Azure. The deployment
methods, and the tools used to perform each type of deployment are described in the following
table:

Deployment Method Tool Requirements
Web based Windows Azure Management
Portal
Browser support for Silverlight
Integrated Development
Environment (IDE)
Visual Studio 2010 and the
Windows Azure SDK
Visual Studio 2010
Command Line The Windows Azure SDK. Command line tools for
deployment are provided as
part of the Windows Azure
SDK

For more information on packaging and deploying an application to Windows Azure, see the
following links:
.NET Languages:Deploying Applications in Windows Azure
Node.js:Windows Azure PowerShell for Node.js Cmdlet Reference

33
PHP:Packaging and Deploying PHP Applications for Windows Azure
Java:Creating a Hello World Application Using the Windows Azure Plugin for Eclipse with
Java
Post-Deployment
If you perform a change to an existing deployment, such as updating a service configuration
setting, upgrading the application, or updating a certificate, this will cause the application
instances to restart. Also, while most changes to a deployment can be made as in place updates
to the existing service, some changes may require you to delete and then redeploy the hosted
service.
For more information on updating an existing deployment, see Overview of Updating a Windows
Azure Service.
For more information on the actions that will cause a restart of the hosted service and how to
minimize the impact of these actions, see Improving Application Availability in Windows Azure.
You are charged for deployments even if they are not running. To ensure that you are not
charged for resources you are not actively using, ensure that you delete any inactive
deployments.
If you are performing tests that involve creating additional instances of your application,
verify that the number of instances is reduced to the normal number after tests have
completed.
If you have configured your deployment to allow remote desktop connections, ensure that
you enable this functionality in the Windows Azure Management Portal only when
needed. Also, If you save the RDP file used for the connection to your local system, you
may be unable to use it to connect to Windows Azure after updating the deployment. If
this is the case, download a new RDP file from the Windows Azure Management Portal.
See Also
Developing Windows Azure Applications
Planning and Designing Windows Azure Applications
Testing, Managing, Monitoring and Optimizing Windows Azure Applications

Note
Note
Note
Bes
Win
Writin
for W
one p
unma
Autho
Bes
App
The fo
when
Verif
Appl
To co
Prope
from t



R
th
m
w
(h
N
st Prac
ndows
ng .NET code
indows applic
platform versu
anaged/native
ors Christian
st Practice
plications
ollowing secti
developing W
fy that the
lication is
onfigure native
erties to disp
the Build tab
Runtime errors
he native code
msvcp100d.dll
which C++ DL
http://go.micro
Note
ctices fo
Azure A
for Windows
cations. There
us the other. T
e code in Wind
Martinez, Tra
es for Ru
s
ions provide r
Windows Azu
Native Cod
Compiled
e code to com
lay the Prope
of the Prope
s stating that
e you deploye
are not redis
L files to redis
osoft.com/fwli
or Runn
Applica
Azure applic
e are subtle d
This documen
dows Azure A
ace Young, a
unning Na
recommendat
re application
de you dep
in Release
mpile in Relea
erties page, a
rties page:
you are miss
ed was comp
stributable DL
stribute see D
ink/p/?LinkId=
ning Un
ations
cations is for t
differences to
nt provides re
Applications.
nd Mark Simm
ative Cod
tions for ensu
ns that call na
ploy with y
e Mode
ase mode righ
and select the
ing msvcr100
iled in Debug
LLs. For more
Determining W
=236016) in th
nmanag
the most part
be aware of w
commendatio
ms
de from W
uring that nati
ative code.
your Windo
ht-click on the
e Release co
0d.dll or msvc
g mode. The f
e information a
Which DLLs to
he C++ 2010
ged Co
just like writin
when writing
ons for runnin
Windows
ve code runs
ows Azure
e native code
nfiguration op
cp100d.dll ind
files msvcr100
about determ
o Redistribute
documentati

3
ode in
ng .NET code
.NET code fo
ng
Azure
correctly

project, selec
ption available

icate that
0d.dll and
ining
e
on.
34
e
or
ct
e

35
Ensure that Native Code can be Located by your Windows Azure
Application when Running in Windows Azure Compute Emulator
Follow these steps to ensure that your Windows Azure Application can locate any referenced
native code when running in Windows Azure Compute Emulator:
1. Set Appropriate Properties for Compiled Native Code Files in Visual Studio
Include the compiled native code file as a project item and on the Properties dialog for the
file set the Copy to Output Directory option to Copy always and the Build Action option to
None.
This will copy the compiled native code file into the \bin directory and ensure that your
Windows Azure Application can locate the compiled native code file when running in
Windows Azure Compute Emulator.



2. When troubleshooting your RoleEntry implementation in the Windows Azure Compute
Emulator it can make it easier to debug issues if you copy the Compiled Native Code
File to the devfabric runtime directory
For example, when using Windows Azure SDK version 1.5 or earlier copy the compiled
native code file to the following directory:

C:\Program Files\Windows Azure SDK\[SDK Version]\bin\devfabric\x64\

Or, when using Windows Azure SDK version 1.6 or later copy the compiled native code
file to this directory:

C:\Program Files\Windows Azure SDK\[SDK Version]\bin\runtimes\base\x64\
3. Ensure that Windows Azure Applications Running in a Web Role Instance can Locate
any Referenced Native Code
If a Windows Azure Application references native code that is wrapped using C++/CLI and
the Windows Azure Application is running in a Web Role instance, use one of the following

36
methods to ensure that the Windows Azure Application can locate the referenced native
code:

The steps below reference the CppCliWebRole project provided with the sample
code in PinvokeCppCliInAzure.zip, available for download at
http://azureunmanagedcode.codeplex.com/. For more information about the sample
code see Sample Code: Running Native Code from Windows Azure Applications.
Method 1: Edit the PATH Environment Variable, then Restart Windows Azure Compute
Emulator and IIS:
a. Stop and exit Windows Azure Compute Emulator.
b. Edit your PATH environment variable to point to a directory containing the native
compiled code.
c. Type iisreset from an elevated command prompt.
d. Press F5 to run the sample code.
Method 2: Use a startup command
In the steps below, the native code is in the file ExampleNativeCode.dll.
a. Stop and exit Windows Azure Compute Emulator.
b. Open the indist.cmd file included with the CppCliWebRole project.
c. Change the following line:
REM copy "%~dps0ExampleNativeCode.dll"
"%windir%\system32\inetsrv"
To:

copy "%~dps0ExampleNativeCode.dll" "%windir%\system32\inetsrv"
d. Save the project.
e. Press F5 to run the sample code.
The use of a startup command works for actual deployments as well. If you prefer
avoiding references to the IIS system directory then you might alternatively create
and run a script to:
a. Change the PATH environment variable to point to the directory containing the native
compiled code.
b. Restart IIS and any processes that depend on it.
Ensure that the Compiled Native Code is 64-bit
Windows Azure is a 64-bit platform, as are the Windows Azure application (worker and web role)
hosts. If you are not using a pure 64-bit development environment and your application
references native code, your application will throw errors. One simple test to verify that all of the
Note
Note

37
native code you are referencing is 64-bit is by testing the native code using console test
applications that are hard coded to run as 64 bit applications.
Use the Native Multi-targeting Feature of Visual Studio 2010 to Build Native
Libraries with the Visual Studio 2008 Platform Toolset
By default, only the Visual C++ runtime libraries for Visual C++ 2008 are installed on Windows
Azure worker and web roles. Therefore, native code compiled against the Visual C++ runtime
library for Visual C++ 2010 or Visual C++ 2005 will fail to load in worker and web role instances. If
you have both Visual Studio 2008 and Visual Studio 2010 installed on the same computer you
can use the native multi-targeting feature of Visual Studio 2010 to build native libraries for your
application with the Visual Studio 2008 platform toolset (compiler, linker, headers, and libraries).
For more information about using Visual Studio 2010 to build an application with the Visual Studio
2008 platform toolset see C++ Native Multi-Targeting
(http://go.microsoft.com/fwlink/p/?LinkId=231170).
Add an Elevated Startup Task to your Worker or Web Role Project to Install
the Native Visual C++ Libraries for Visual C++ 2010 on Worker/Web Role
Instances
If building native libraries with the Visual Studio 2008 platform toolset is not a viable option,
consider adding an elevated startup task to your worker or web role project to copy and install the
required version of the runtime libraries. The following steps describe how to create an elevated
startup task to copy the 64-bit version of the Microsoft Visual C++ 2010 Redistributable Package
to a worker / web role and run the redistributable package to install the Visual C++ libraries for
Visual C++ 2010 onto the worker / web role instances:
1. Create a Startup folder for your worker or web role project.
2. Copy vcredist_x64.exe (http://go.microsoft.com/fwlink/p/?LinkId=225987) to the Startup
folder.
3. Create a startup.cmd file in the Startup folder.
4. Edit startup.cmd and add the following line:

"%~dps0vcredist_x64.exe" /q /norestart

5. Modify the properties of vcredit_x64.exe and startup.cmd from Visual Studio Solution
Explorer. Set the Build Action option to Content and the Copy to Output Directory option
to Copy if Newer.
6. Edit the ServiceDefinition.csdef file for the role by adding the following elevated startup task
to execute startup.cmd:

< Task commandLine ="Startup\Startup.cmd" executionContext ="elevated" taskType
="simple" />

38
Use Process Monitor to Troubleshoot Problems Loading Files in Windows
Azure Compute Emulator
The error message generated when an assembly cannot be loaded by Windows Azure Computer
Emulator may be non-intuitive. To troubleshoot problems loading files in a Worker or Web Role
host instance use Process Monitor as follows:
1. Download Process Monitor from Process Monitor v2.96
(http://go.microsoft.com/fwlink/p/?LinkID=137175).
2. Launch Process Monitor to troubleshoot problems with Windows Azure Applications loading
files when running in Windows Azure Compute Emulator.
3. Set filter parameters as depicted below for the Windows Azure Compute Emulator host. If
troubleshooting problems with Windows Azure Applications running in a worker role project
change the filter to display entries with the process name WaWorkerHost.exe rather than
WaWebHost.exe.





4. Look for any NAME_NOT_FOUND messages. This will help to isolate the missing library file.
Once you have determined which file is missing you can narrow the scope of the search by
applying a filter to isolate only messages related to that file.
Configure Worker/Web Roles with the .NET Full Trust Level
To enable Windows Azure applications to execute 64-bit native code using P/Invoke, first
configure the associated worker or web roles with the .NET Full Trust level. For more information
about calling native code from applications running in Windows Azure worker or web roles see
the following resources:

39
Windows Azure Role Architecture (http://go.microsoft.com/fwlink/p/?LinkId=236012)
Windows Azure and Native Code, Hands-On Lab
(http://go.microsoft.com/fwlink/p/?LinkID=231343)
Windows Azure and Native Code Hands-On Lab
(http://go.microsoft.com/fwlink/p/?LinkId=236013).Provides the contents of the Windows
Azure and Native Code, Hands-On Lab in a single Word document.
Sample Code: Running Native Code from Windows Azure
Applications
This section describes the sample code in PinvokeCppCliInAzure.zip
(http://go.microsoft.com/fwlink/p/?LinkId=236170), available for download at
http://azureunmanagedcode.codeplex.com/
When using Windows Azure Compute Emulator from Visual Studio to work through the
sample code, it is not necessary to run the portion of the startup command that installs
the Visual C++ 2010 runtime libraries, therefore you should comment out the following
line in the startup.cmd file:
REM "%~dps0vcredist_x64.exe" /q /norestart
The sample code contains a project which creates a native DLL called ExampleNativeCode.dll. It
has been compiled as a 64 bit library in Release mode as recommended.
Exit and restart Windows Azure Compute Emulator each time you run any of the sample
code if any changes are made to the code or to Windows environment variables. This will
ensure that any referenced native code is released from memory so that it can be
recompiled and that any environment variable changes are picked up by Windows Azure
Compute Emulator the next time it runs.
The sample code ExampleNativeCode.dll is wrapped using P/Invoke in a project called
ManagedUsingPinvoke and wrapped using C++/CLI in a project called ManagedUsingCppCLI.
The code in the native DLL is a modified version of the default Win32 project template with
exports and contains a single function that answers all known questions in the universe, as long
as the answer for said questions is the number 42:
EXAMPLENATIVECODE_API int fnExampleNativeCode(void)
{
return 42;
}
The P/Invoke code and the C++/CLI code that utilize this function are very similar:
P/Invoke
public class Class1
{
[DllImport("ExampleNativeCode.dll")]
Note
Note

40
static extern int fnExampleNativeCode();

public int GetTheAnswerToEverything()
{
return fnExampleNativeCode();
}
}
C++/CLI
public ref class Class1
{
public:

int GetTheAnswerToEverything()
{
return fnExampleNativeCode();
}
};
Both Worker Role samples (PInvokeWorkerRole and CppCliWorkerRole) in the solution invoke
the code in the same manner:
try
{
Trace.WriteLine(new Class1().GetTheAnswerToEverything());
}
catch (Exception ex)
{
Trace.WriteLine(ex);
}
Both Worker Role samples include the native compiled code and are configured to always copy
the native compiled code to the output directory.
When you run press F5 to execute the code in Windows Azure Compute Emulator you should
see the following output:





Certa
in a W
PInvo
versio
P/Invo
<p>
Th
<%=Env
<%=new
</p>
C++/C
<p>
Th
<%=Env
<%=new
</p>
Runn



Howe


T
co
N
in considerat
Web Role inst
okeWebRole
on of the defa
oke
his is PInvo
vironment.Ge
w Class1().G
CLI
his is C++/C
vironment.Ge
w Class1().G
ing the PInvo
ever, running t
his error occu
ompiled code
Note
ions apply wh
ance as comp
and CppCliW
ault ASP.NET
ke <br />
tEnvironment
etTheAnswerT
LI <br />
tEnvironment
etTheAnswerT
okeWebRole p
the CppCliWe
urs even with
e to the output
hen calling na
pared to runn
WebRole proj
project temp
Variable("PA
oEverything(
Variable("PA
oEverything(
project will res
ebRole projec
the proper pr
t directory.
ative code from
ning in a Work
jects contain
late:
ATH") %><br /
) %>
ATH") %><br /
) %>
sult in the exp
ct without mo
roject settings
m Windows A
ker Role insta
the following
/>
/>
pected output
dification will
s applied to c
Azure Applica
ance. The
code in a slig
t similar to the
result in the f
copy the nativ

4
ations running
ghtly modified
e following:

following erro

e
41
g
d
or:
To ad
Applic
using
simila



ddress this err
cations Runni
one of these
ar to the follow
ror, use one o
ing in a Web
methods you
wing:
of the method
Role Instance
u should see

ds described i
e can Locate
the expected
n Ensure tha
any Referenc
output for the
t Windows Az
ced Native Co
e CppCliWeb

4
zure
ode. After
bRole page,

42

43
Other Third Party Software on Windows Azure
This section contains topics that address or describe how to use various types of third-party
software on Windows Azure. In many cases, the topics describe how to use a library or software
stack; in others, how best to do so.
In This Section

Leveraging Node.js libuv in Windows Azure
This topic describes one way to use the libuv library, which is a common cross-
platform networking library most often used in node.js applications.

Related Sections
Planning and Designing Windows Azure Applications
Testing, Managing, Monitoring and Optimizing Windows Azure Applications
Node.js Developer Guidance (Windows Azure)
PHP Developer Guidance (Windows Azure)
Developing Windows Azure Applications
Leveraging Node.js libuv in Windows Azure
Author: Christian Martinez
Node.js is definitely one of the hottest things going right now and you can find blogs about
running it in Windows Azure such as this one by Nathan Totten. While that is certainly interesting
and cool, the subject of this post was inspired by this post on the Node.js blog by Nodes creator
Ryan Dahl. More specifically, it was this quote that got me interested:
Since Node is totally non-blocking, libuv turns out to be a rather useful library itself: a BSD-
licensed, minimal, high-performance, cross-platform networking library.
I am always interested in cool networking libraries with my current favorite being 0MQ (
http://www.zeromq.org/ ) which is a full featured, multiple MEP supporting, high performance,
general purpose messaging library. Libuv is a very different beast from 0MQ and has different
goals but both projects share the common bond of being open source and of making performance
a first class citizen.
In order to get started on my adventure I downloaded the 5.7 version of the libuv source and
attempted to run the vcbuild.bat file. After that tanked, I downloaded python, then subversion and
checked out the gyp script and.. anyways these shenanigans could probably be a blog in and of
themselves but for now just know that what ended up being generated was a VS 2010 solution
that I was then able to tweak to generate a 64 bit static library. Since libuv and Node are highly
active projects its likely that things have changed since I went through my exercise so if you take
this on yourself your experience may or may not be different.

44
Once I had the library, I had to decide how I wanted to deploy things to Azure. Libuv is native C
code and comes with a test suite that I could have theoretically deployed using the SDK
1.5ProgramEntryPoint element. But, that would have been boring and probably not seen as too
useful by developers who wanted to use the library with their own managed code so what I did
instead was develop a prototype in C++/CLI that allowed me to create a server that runs in a
WorkerRole and looks like this:
public override void Run()
{
long cnt =0;
long byteCnt = 0;
using (var server = new TcpServer())
{
var ep = RoleEnvironment.CurrentRoleInstance.InstanceEndpoints["listen"];
//must assign callback before calling listen
server.OnBytesReceived += (arg) =>
{
Interlocked.Increment(ref cnt);
Interlocked.Add(ref byteCnt, arg.Length);
//Trace.WriteLine(cnt);
};
server.Listen(ep.IPEndpoint.Address.ToString(),
ep.IPEndpoint.Port);
//if this is successful libuv uv_run blocks exit
}
}
That should look very normal to any managed developer and while all I do in the prototype is
increment bytes, and number of calls received youre free to do any interesting thing you choose
in the exposed delegate.
So, how did I get from straight C code to a nice managed experience? Well, the first thing I had to
do was to decide if I wanted to use P/Invoke or C++/CLI. After perusing the libuv source and
realizing the whole thing was callback driven, I decided to go with C++/CLI. While pointer to
member function syntax is always a joy to deal with, I still prefer dealing with that to managing the
DLLImport and marshalling and mapping one does using P/Invoke. This resource was a great
help in figuring out the callback syntax. Essentially what you end up needing is a Delegate
declaration, a Delegate instance variable and a GCHandle. My header file ended up looking like
this:
// LibuvWrapper.h
#pragma once
#include "include\uv.h"
using namespace System;
using namespace System::Collections::Concurrent;
using namespace System::Runtime::InteropServices;

45

namespace LibuvWrapper {

public delegate void BytesReceived(array<Byte>^ stuff);

public ref class TcpServer sealed
{
public:

TcpServer();
~TcpServer();
!TcpServer();
void Listen(String^ address, int port);
//public delegate for passing bits sent
//from client onto implementation code
BytesReceived^ OnBytesReceived;
private:
//delegate declarations. to convert to native callbacks you
//need the delegate declaration, delegate instance,
//gchandle and preferably a typedef
delegate void AfterRead(uv_stream_t* handle, ssize_t nread,
uv_buf_t buf);
delegate void OnConnection(uv_stream_t* server,
int status);
delegate void AfterWrite(uv_write_t* req, int status);
delegate void OnClose(uv_handle_t* peer);
delegate void OnServerClose(uv_handle_t* handle);

//callback methods
void AfterReadImpl(uv_stream_t* handle, ssize_t nread,
uv_buf_t buf);
void OnConnectionImpl(uv_stream_t* server, int status);
void AfterWriteImpl(uv_write_t* req, int status);
void OnCloseImpl(uv_handle_t* peer);
void OnServerCloseImpl(uv_handle_t* handle);


//handles and delegate members that will be
//converted to native callbacks
AfterRead^ afterReadFunc;
GCHandle afterReadHandle;
OnConnection^ onConnectionFunc;

46
GCHandle onConnectionHandle;
AfterWrite^ afterWriteFunc;
GCHandle afterWriteHandle;
OnClose^ onCloseFunc;
GCHandle onCloseHandle;
OnServerClose^ onServerCloseFunc;
GCHandle onServerCloseHandle;


//libuv members
uv_tcp_t * server;
uv_loop_t * loop;
uv_buf_t * ack;
//this is the same response sent to all
//clients as ack. created once and never
//deleted until object destruction
};
}
Wiring up the delegates looks like this:
afterReadFunc = gcnew AfterRead(this, &LibuvWrapper::TcpServer::AfterReadImpl);
afterReadHandle = GCHandle::Alloc(afterReadFunc);

The typedef and passing the callback look like this:
typedef void (__stdcall * AFTERREAD_CB)(uv_stream_t* handle,
ssize_t nread, uv_buf_t buf);

...

if(uv_read_start(stream, echo_alloc,static_cast<AFTERREAD_CB>
(Marshal::GetFunctionPointerForDelegate(afterReadFunc).ToPointer())))
{
throw gcnew System::Exception(
"Error on setting up after read\n");}
It is not super pretty but it got the job done! The rest of the internals are available with the posted
source.
Once I had a server, I built a simple test client using plain old C# , TcpClient and ran that from
another WorkerRole:
var ep = RoleEnvironment.Roles["LibuvServer"].Instances.Select(
i => i.InstanceEndpoints ["listen"]).ToArray();
var cl = new TcpClient();
byte[] msg = Encoding.ASCII.GetBytes(new string('x',msgSize));

47
byte[] gbye = Encoding.ASCII.GetBytes("|^|");
byte[] ackBuf = new byte[1];

//it's easy for the client to start before server so wait
//forever until we connect.
//for something more sophisticated AKA "the greatest retry library
//in the history of man" see
//http://code.msdn.microsoft.com/Transient-Fault-Handling-b209151fwhile (!cl.Connected)
{
try
{
cl.Connect(ep[0].IPEndpoint);
}
catch (Exception)
{
Thread.Sleep(100);
}

}
using(var stream = cl.GetStream())
{
var watch = new Stopwatch();
watch.Start();
for (int i = 0; i < numTimes ; i++)
{

stream.Write(msg, 0, msg.Length);
stream.Read(ackBuf, 0, ackBuf.Length);
}
watch.Stop();
log = string.Format("Sent {0} messages of size {1} and received {0} acks in {2}
milliseconds",
numTimes, msgSize,
watch.ElapsedMilliseconds);
Trace.WriteLine(log, "Information");
stream.Write(gbye,0,gbye.Length);
}
Next, I added a simple script and Startup Task to my server project to ensure the 2010 C
runtime was available:
"%~dps0vcredist_x64.exe" /q /norestart
And finally, I deployed the project and ran it in Windows Azure.

48
The full source for the project can be found here.


49
Business Continuity for Windows Azure
Published: March 2012
Authors: Patrick Wickline, Adam Skewgar, Walter Myers III
Reviewers: Michael Hyatt, Brad Calder, Tina Stewart, Steve Young, Shont Miller, Vlad
Romanenko, Sasha Nosov, Monilee Atkinson, Yousef Khalidi, Brian Goldfarb
Introduction
IT managers spend considerable time and effort ensuring their data and applications are
available when needed. The general approach to business continuity in the cloud is no different
than any other environment. The same principles and goals apply, but the implementation and
process will be different.
Availability problems can be classified into three broad categories:
Failure of individual servers, devices, or network connectivity
Corruption, unwanted modification, or deletion of data
Widespread loss of facilities
This document explains how you think about and plan for availability in all three of these
categories when using Windows Azure.
We start by describing some of the high availability services that Windows Azure provides and
then describe some of the common backup and recovery mechanisms you may choose to
employ. We then move on to describe the services that Windows Azure exposes to help you build
highly-available applications and, finally, discuss strategies for designing applications that can
withstand full site outages.
The fundamental concepts of preparedness and recovery are consistent with what you are
probably already familiar with. However, Windows Azure mitigates some potential failures for you
and provides the infrastructure and capabilities for you to design your own availability and
recovery plans at a much lower cost than traditional solutions.
Security and availability are closely related topics. While we touch on security of the platform
itself, this document is not intended to provide full coverage of that topic and we will not cover
designing secure applications. For more information on these topics, please see the following:
Windows Azure Security Overview
Security Best Practices for Developing Windows Azure Applications
A Cloud Platform Designed for High Availability
By deploying your application on Windows Azure you are taking advantage of many fault
tolerance and secure infrastructure capabilities that you would otherwise have to design, acquire,
implement, and manage. This section covers the things Windows Azure does for you without any
additional expense or complexity to ensure that you are building on a platform you can trust.

50
World Class Data Centers in Multiple Geographies
A basic requirement for business continuity is the availability of well-managed datacenter
infrastructure in diverse geographic locations. If individual data centers are not properly managed,
the most robust application designs may be undermined by gross failures at the infrastructure
level. With Windows Azure, you can take advantage of the extensive experience Microsoft has in
designing, managing, and securing world-class data centers in diverse locations around the
world.
These are the same data centers that run many of the worlds largest online services. These
datacenters are designed and constructed with stringent levels of physical security and access
control, power redundancy and efficiency, environmental control, and recoverability capabilities.
The physical facilities have achieved broad industry compliance, including ISO 27001 and SOC /
SSAE 16 / SAS 70 Type II and within the United States, FISMA certification.
To ensure recoverability of the Windows Azure platform core software components, Microsoft has
established an Enterprise Business Continuity Program based on Disaster Recovery Institute
International (DRII) Professional Practice Statements and Business Continuity Institute (BCI)
Good Practice Guidelines. This program also aligns to FISMA and ISO27001 Continuity Control
requirements. As part of this methodology, recovery exercises are performed on a regular basis
simulating disaster recovery scenarios. In the rare event that a system failure does occur,
Microsoft uses an aggressive, root cause analysis process to deeply understand the cause.
Implementation of improvements learned from outages is a top priority for the engineering
organization. In addition, Microsoft provides post-mortems for every customer impacting incident
upon request.
Infrastructure Redundancy and Data Durability
The Windows Azure platform mitigates outages due to failures of individual devices, such as hard
drives, network interface adapters, or even entire servers. Data durability for SQL Azure and
Windows Azure Storage (blobs, tables, and queues) is facilitated by maintaining multiple copies
of all data on different drives located across fully independent physical storage sub-systems.
Copies of data are continually scanned to detect and repair bit rot, an often overlooked threat to
the integrity of stored data.
Compute availability is maintained by deploying roles on isolated groupings of hardware and
network devices known as fault domains. The health of each compute instance is continually
monitored and roles are automatically relocated to new fault domains in the event of a failure. By
taking advantage of the Windows Azure service model and deploying at least two instances of
each role in your application, your application can remain available as individual instances are
relocated to new fault domains in the event of a failure within a fault domain. This all happens
transparently and without downtime with no need for you to deal with managing the underlying
infrastructure.
Furthermore, Windows Azure provides built in network load balancing, automatic OS and service
patching, and a deployment model that allows you to upgrade your application without downtime
by using upgrade domains, a concept similar to fault domains, which ensures that only a portion
of your service is updated at any time.

51
Data Backup and Recovery
The ability to restore application data in the event of corruption or unwanted modification or
deletion is a fundamental requirement for many applications. The following sections discuss
recommended approaches for implementing point-in-time backups for Windows Azure Blob and
Table storage, and SQL Azure databases.
Blob and Table Storage Backup
While blobs and tables are highly durable, they always represent the current state of the data.
Recovery from unwanted modification or deletion of data may require restoring data to a previous
state. This can be achieved by taking advantage of the capabilities provided by Windows Azure to
store and retain point-in-time copies.
For Windows Azure Blobs, you can perform point-in-time backups using the blob snapshot
feature. For each snapshot, you are only charged for the storage required to store the differences
within the blob since the last snapshot state. The snapshots are dependent on the existence of
the original blob they are based on, so a copy operation to another blob or even another storage
account is advisable to ensure that backup data is properly protected against accidental deletion.
For Windows Azure Tables, you can make point-in-time copies to a different table or to Windows
Azure Blobs. More detailed guidance and examples of performing application-level backups of
tables and blobs can be found here:
Protecting Your Tables Against Application Errors
Protecting Your Blobs Against Application Errors
SQL Azure Backup
Point-in-time backups for SQL Azure databases are achieved with the SQL Azure Database Copy
command. You can use this command to create a transactionally-consistent copy of a database
on the same logical database server or to a different server. In either case, the database copy is
fully functional and completely independent of the source database. Each copy you create
represents a point-in-time recovery option. You can recover the database state completely by
renaming the new database with the source database name. Alternatively, you can recover a
specific subset of data from the new database by using Transact-SQL queries. For additional
details about SQL Azure backup, see Business Continuity in SQL Azure.
Platform Services for Building Highly Available Solutions
On top of a secure and highly redundant infrastructure, there are Windows Azure services that
act as building blocks for designing highly available deployments that span multiple data centers.
These services can each be used on their own or in combination with each other, third-party
services and application-specific logic to achieve the desired balance of availability and recovery
goals.
Geo-Replication of Blob and Table Storage
All Windows Azure blob and table storage is replicated between paired data centers hundreds of
miles apart within a specific geographic region (e.g., between North Central and South Central in

52
the United States or between North and West in Europe). Geo-replication is designed to provide
additional durability in case there is a major data center disaster. With this first version of geo-
replication, Microsoft controls when failover occurs and failover will be limited to major disasters
in which the original primary location is deemed unrecoverable in a reasonable amount of time. In
the future, Microsoft plans to provide the ability for customers to control failover of their storage
accounts on an individual basis. Data is typically replicated within a few minutes, although
synchronization interval is not yet covered by an SLA. More information can be found the blog
post Introducing Geo-replication for Windows Azure Storage.
SQL Azure Export to Blob Using the SQL Azure Import/Export Service
Geo-replication of SQL Azure data can be achieved by exporting the database to an Azure
Storage blob using the SQL Azure Import/Export service. This can be implemented in three
ways:
Export to a blob using storage account in a different data center.
Export to a blob using storage account in the same data center (and rely on Windows Azure
Storage geo-replication to the separate data center).
Import to your on-premises SQL Server.
For implementation details see the MSDN article Business Continuity in SQL Azure.
Traffic Manager
Windows Azure Traffic Manager (currently a Community Technology Preview) allows you to load
balance incoming traffic across multiple hosted Windows Azure services whether they are
running in the same datacenter or across different datacenters around the world. By effectively
managing traffic, you can ensure high performance, availability, and resiliency of your
applications. Traffic routing occurs as a result of policies that you define and that are based on
one of the following three criteria:
Performance traffic is forwarded to the closest hosted service in terms of network latency
Round robin traffic is distributed equally across all hosted services
Failover traffic is sent to a primary service and, if this service is not online, to the next
available service in a list
Traffic Manager constantly monitors hosted services to ensure they are online and will not route
traffic to a service that is unavailable. Using Traffic Manager you can create applications that are
designed to respond to requests in multiple geographic locations and that can therefore survive
entire site outages. Prior to the general availability of a production version of Traffic Manager, you
can implement third-party DNS routing solutions such as those provided by Akamai or Level 3
Communications to build applications that are designed for high-availability. For more information,
see Windows Azure Traffic Manager.
Planning for Site Disasters
Windows Azure greatly reduces what you need to worry about to ensure high availability within a
single data center. However, even the most well designed data center could be rendered
inaccessible in the case of a true disaster. To plan for such disasters, you must think through both

53
the technical and procedural steps required to provide the level of cross-site availability you
require. Much of this is application specific and as the application owner you must make tradeoff
decisions between availability and complexity or cost.
Types of Cross-Site Availability
There are some basic steps that all but the most trivial of applications should take to know that
they can be deployed in a different data center in the event of a site disaster. For many
applications, redeployment from scratch is an acceptable solution. For those that need a quicker
and more predictable recovery a second deployment must be ready and waiting in a second data
center. For an even smaller set of applications, true multi-site high-availability is required. We will
look at each of these classes of applications in order from the least complex and costly to the
most complex and costly. Cross-site availability is a lot like insurance you pay for protection that
you hope you will never have to use. No two applications have precisely the same business
requirements or technical design points, so these classes of applications are meant as general
guidance that should be adapted to your specific needs.
For an application to become available in a secondary data center, three requirements must be
met:
1. The customers application and dependent services must be deployed.
2. The necessary data must be available to the application (typically in the same data center).
3. Any external traffic must be routed to the application.
Each of these three requirements can be accomplished at the time of a disaster or ahead of time,
and each can be accomplished manually or automatically. Every application is different, so its
possible there are other application-specific requirements, such as availability of a dependent
service. In the rest of this section, well describe several common patterns for recovering
availability in the case of a disaster.
Redeploy on Disaster
The simplest form of disaster recovery is to redeploy your application and dependent services,
such as Windows Azure Cache and Windows Azure Service Bus, when a disaster occurs.
Redeployment of applications is accomplished using the same method used when the application
was originally created. This can be done manually via the Windows Azure Portal or can be
automated using the Windows Azure Service Management interface.
To mitigate data loss, the redeployment strategy should be coupled with data backup or
synchronization to make sure the data exists and is usable from a backup storage account or
database. Because the data for the new deployment is in a different account, the application must
be designed with configurable connection strings. This way, on redeploy, only the applications
configuration must be changed.
To meet the third requirement (traffic routing to the new deployment), a custom domain name
must be used (not <name>.cloudapp.net). In the case of a disaster, the custom domain name can
be configured to route to the new application name after the new deployment is completed. A
CNAME record is sufficient to accomplish this, but you should be sure you know the process for
making the necessary changes with your DNS provider.

54
Because no compute resources are reserved for recovery this is the least expensive solution.
However, low cost comes at the expense of increased risk and increased recovery time. In the
event of a large-scale disaster, resources in other data centers are allocated based on real-time
availability.
Active/Passive Deployment
The redeployment strategy described previously takes time and has risks. Some customers need
faster and more predictable recovery and may want to reserve standby resources in an alternate
data center. The active/passive pattern means keeping an always-ready (potentially smaller)
deployment of the application in a secondary data center. At the time of a disaster, the secondary
deployment can be activated and scaled.
Here, the first requirement (deployment of the app) is taken care of ahead of time. The second
requirement (data availability) is typically handled one of the data replication methods discussed
above. The third requirement (routing traffic) is handled by the same way as with the redeploy
pattern (DNS change at the time of disaster). In this pattern, however, the process can be fully
automated using Windows Azure Traffic manager (CTP) or other similar DNS management
solutions, which can be configured to reroute traffic to a secondary deployment only when the first
deployment stops responding.
When using a warm standby, it is important to carefully think through data consistency
implications at the time of the failover. You need to plan for any steps that must be taken before
your application is brought online in an alternate location as well as how and when to return to the
original deployment when it becomes available again. Because the application is not designed to
function simultaneously in multiple data centers, returning service to the primary location should
follow a similar procedure to the one you follow in the case of a failover.
Active/Active Deployment
The active/active pattern is very similar to the active/passive deployment pattern, but the third
requirement (traffic routed to the deployment) is always satisfied because both deployments now
handle incoming traffic at all times . As in the active/passive configuration, this can be
accomplished with Windows Azure Traffic Manager (CTP) or similar solutions, but now using a
performance policy, which routes traffic from a user to whichever data center provides them the
lowest latency or a load-balancing policy, which distributes load evenly to each site. This solution
is the most difficult to implement, as it requires the application to be designed to handle
simultaneous requests across multiple instances of the application that reside in distinct data
centers. However, it is more efficient than the active/passive solution in that all compute
resources are utilized all the time.
To achieve full, multi-site high availability without downtime, another common pattern is the
active/active deployment, so named because two deployments in two data centers are both
active. This meets the first requirement (that the app is deployed). In this model, a full copy of the
data is in both locations, and they are continually synchronized. This meets the second
requirement (that the data is available).

55
Access Control Service (ACS)
Windows Azure Access Control Service (ACS) is a cloud-based service that provides an easy
way of authenticating and authorizing users to gain access to your web applications and services
while allowing the features of authentication and authorization to be factored out of your code.
Instead of implementing an authentication system with user accounts that are specific to your
application, you can let ACS orchestrate the authentication and much of the authorization of your
users. ACS integrates with standards-based identity providers, including enterprise directories
such as Active Directory, and web identities such as Windows Live ID, Google, Yahoo!, and
Facebook.
Restoration of ACS Namespaces in the Event of a Disaster
Being built on the Azure platform, ACS is a highly available system and resilient to failures within
a single data center. However in the event of a disaster rendering a full data center inoperable,
there is the potential for data loss. Should such a disaster occur, all Windows Azure services will
make best efforts to restore you to a pre-disaster state as quickly as possible.
Data Recovery
The ability to restore an ACS subscription is dependent on the version of ACS that you subscribe
to. ACS currently consists of two versions (version 1.0 and version 2.0) with different disaster
recovery and data restoration capabilities. There are two ways to determine the version of an
ACS namespace:
1. See the ACS Version column in the Service Bus, Access Control, & Caching section of the
Windows Azure portal. This column is present for all Access Control and Service Bus
namespaces, the latter of which includes a built-in ACS namespace.
2. You can also determine the ACS version by trying to connect to the ACS 2.0 management
portal in a web browser. The URL to the ACS 2.0 management portal is
https://<tenant>.accesscontol.windows.net/ where <tenant> should be replaced by your
actual ACS namespace name. If you can successfully access the portal or are prompted to
sign in, then the ACS version is 2.0. If you get an HTTP status code 404 or error message
Page cannot be found, then the ACS version is 1.0.
ACS 1.0 Namespaces
While benefiting from the high availability built into Windows Azure, ACS 1.0 namespaces are not
recoverable in the event of a disaster. Should a disaster occur Microsoft will work with ACS 1.0
customers to mitigate the loss of data, however no guarantees are made about the ability to
recover from a disaster.
In December 2011, Microsoft officially deprecated ACS version 1.0. All customers with ACS 1.0
namespaces are encouraged to migrate to ACS 2.0 in advance of December 20, 2012 to avoid
potential service interruptions. For more information about migrating from ACS 1.0 to ACS 2.0,
please see the Guidelines for Migrating an ACS 1.0 Namespace to an ACS 2.0 Namespace.
ACS 2.0 Namespaces
ACS 2.0 takes backups of all namespaces once per day and stores them in a secure offsite
location. When ACS operation staff determines there has been an unrecoverable data loss at one
of ACSs regional data centers, ACS may attempt to recover customers subscriptions by

56
restoring the most recent backup. Due to the frequency of backups data loss up to 24 hours may
occur.
ACS 2.0 customers concerned about potential for data loss are encouraged to review a set of
Windows Azure PowerShell Cmdlets available through the Microsoft hosted Codeplex Open
Source repository. These scripts allow administrators to manage their namespaces and to import
and extract all relevant data. Through use of these scripts, ACS customers have the ability
develop custom backup and restore solutions for a higher level of data consistency than is
currently offered by ACS.
Notification
In the event of a disaster, information will be posted at the Windows Azure Service Dashboard
describing the current status of all Windows Azure services globally. The dashboard will be
updated regularly with information about the disaster. If you want to receive notifications for
interruptions to any the services, you can subscribe to the services RSS feed on the Service
Dashboard. In addition, you can contact customer support by visiting the Support options for
Windows Azure web page and follow the instructions to get technical support for your service(s).
Summary
Meeting your availability and recovery requirements using Windows Azure is a partnership
between you and Microsoft. Windows Azure greatly reduces the things you need to deal with by
providing application resiliency and data durability to survive individual server, device and
network connection failures. Windows Azure also provides many services for implementing
backup and recovery strategies and for designing highly-available applications using world-class
datacenters in diverse geographies. Ultimately, only you know your application requirements and
architecture well enough to ensure that you have an availability and recovery plan that meets
your requirements.
We hope this documents helps you think through the issues and your options. We welcome your
feedback on this document and your ideas for improving the business continuity capabilities we
provide. Please send your comments to [email protected].


57
Messaging


58
Capacity Planning for Service Bus Queues and
Topics
Author: Valery Mizonov and Ralph Squillace
This topic describes:
The often-important difference in queue size limits between Windows Azure queues and
Windows Azure Service Bus queues and topics
How to approximate the correct Service Bus queue capacity when you want to take
advantage of Service Bus queue and topic features
A test run with messages of different sizes to give some idea of the number of messages of
that size that can be placed in a Service Bus queue or topic
Queue and Topic Size Limitations in Windows Azure Service Bus
Both Windows Azure queues and Service Bus queues are implementations of queued access
storage, but each has a slightly different feature set, which means you may choose one or the
other depending on the needs of your particular application. For example, Service Bus has topics
and subscriptions, a pub-sub implementation that is often important in business notification
scenarios; whereas Windows Azure queues can store as much as 100 terabytes of data when
Service Bus queues are currently limited to 5 gigabytes.
The two major areas of difference maximum message size and maximum queue size -- are
shown below:

Comparison Windows Azure Queues Service Bus Queues
Maximum Message Size 64 KB
Note: This includes the
roughly 25% overhead of
base64 encoding.
256 KB
Note: This includes both the
headers and the body, where
the maximum header size is
64 KB.
Maximum Queue Size 100 TB
Note: Maximum size is limited
at the storage account level.
1, 2, 3, 4, or 5 GB
The size is defined upon
creation of a queue or topic.

This last difference the 5 GB queue and topic size limit for Service Bus, compared to 100 TB
limit for Windows Azure queues can be fairly important, because it means that if you want to
use Service Bus features like deadlettering, topics and subscriptions, rules, or actions, you must
assess how much Service Bus queue or topic capacity you will need for your application in
advance to build the solid application you want to build. It can be easy for a very active
application to create a queue or topic with 5 GB of data depending upon the message size, the
size of the message headers, and the size of any custom message properties and how fast
those messages are enqueued or sent to a topic.

59
Finally, with Service Bus queues, you establish the queue size when you create the queue. You
cannot resize the queue after its been created; whereas Windows Azure queues have access to
so much storage that thinking about increasing the size of the Windows Azure queue is almost
not required.
When a queue or topic exceeds its configured capacity, subsequent requests to enqueue a
message result in a Microsoft.ServiceBus.Messaging.QuotaExceededException exception. Your
application should provide the adequate handling for this particular exception type. For example,
one approach would be to temporarily suspend message delivery to a queue or topic and give the
consumers enough time to process the backlog before message publication can be resumed.
It is also important to remember that messages in Service Bus queues are comprised of two
parts: a header and a body. The total size of the entire message (header + body) cannot exceed
256 KB. The Service Bus brokered messaging API uses binary XML serialization (not text XML),
which reduces the output size of the serialized payload, and while this in turn enables storing
messages slightly larger than 256 KB, you must test the reduction in size that you see for your
own application.
Calculating the Capacity of Service Bus Queues and Topics
Lets look at the basic algorithm to calculate the size of messages, and therefore what kind of
Service Bus queues or topics you might need for your application.
An empty message has a body size of 1024 bytes and default header size of 156 bytes; including
other elements, this rounds out to a total of 1635 bytes in a message with no custom headers or
properties and an empty message body.
The following formula can be used to estimate the size requirements that the specified number of
messages can introduce:

Number Of Messages * (Message Body Size + Message Header Size)
In order to determine the body size, use the BrokeredMessage.Size property. Header size can
be trickier, so it depends on how accurate you need to be. The most accurate method is to send
the message (or a test message if you need to know the info before sending multiple messages),
and then query the queue metadata using the NamespaceManager.GetQueue (or
NamespaceManager.GetTopic) method and use the SizeInBytes property (of either the
QueueDescription or TopicDescription objects) to determine how much header was added.
Topic size requires a slightly different algorithm. The formula to determine how much space a
given number of messages consumes in the topic is:
Number Of Messages * (Message Body Size + (Message Header Size * Number Of
Subscriptions))
Note that the size of the header is multiplied by the number of subscriptions to the topic, which
means that if any custom headers are added to your topic messages, the topic size will increase
in line with the number of subscriptions.
For example, if you had a default message with 200 subscriptions, it results in a topic size of 32
KB. However, if you increase the header size to 600 bytes, the topic size will now be 120 KB.
Finally, if you add the ACK messages that flow from each subscription receiver, youre adding in

60
quite a bit. A single message with a 600-byte header matching 200 subscriptions, add in 200 ACK
messages (one for each subscription) and youre looking at 568,635 bytes. Its important to have
thought about this ahead of time.
Validating Capacity Approximations
You can take the following data into consideration when approximating how many messages you
can hold in a single Service Bus queue. This data was captured by a custom utility in order to
determine the queue capacity with different message size options.

Sample
Message Size
1 GB Queue 2 GB Queue 3 GB Queue 4 GB Queue 5 GB Queue
1 KB 1,041,790 2,059,920 3,128,550 4,186,400 5,238,750
10 KB 102,996 208,358 312,537 416,716 520,895
50 KB 20,857 41,792 62,507 83,343 104,179
100 KB 10,466 20,836 31,254 41,672 52,090
250 KB 4,191 8,334 12,501 16,669 20,836

To rule out the potential impact of binary XML serialization on the test results, all sample
messages were initialized with a byte array of the respective size filled in with random values.
See Also
Service Bus Queues
The Developer's Guide to Service Bus
An Introduction to Service Bus Queues
Service Bus Topics and Queues
Queues, Topics, and Subscriptions


61
Best Practices for Handling Large Messages with
Windows Azure Queues
Author: Valery Mizonov
Reviewers: Brad Calder, Larry Franks, Jai Haridas, Sidney Higa, Christian Martinez, Curt
Peterson, and Mark Simms
This article is intended to offer a developer-oriented guidance on the implementation of a
generics-aware storage abstraction layer for the Windows Azure Queue Service. The problem
space addressed by the article is centered on supporting very large messages in Windows Azure
queues and overcoming the message size limitation that exists today. Simply put, this blog and
the associated code will enable you to utilize Windows Azure Queues without having to engage in
the message size bookkeeping imposed by the queues 64KB limit.
Why Large Messages?
There were wonderful times when "640K ought to be enough for anybody." A few kilobytes could
buy a luxurious storage space where a disciplined developer from the past was able to happily
put all of her applications data. Today, the amount of data that modern applications need to be
able to exchange can vary substantially. Whether its a tiny HL7 message or multi-megabyte EDI
document, modern applications have to deal with all sorts of volumetric characteristics evolving
with unpredictable velocity. A business object that was expressed in a multi-byte structure in the
last century may easily present itself today as a storage-hungry artifact several times larger than
its predecessor thanks to modern serialization and representation techniques and formats.
Handling messages in a given solution architecture without imposing technical limitations on
message size is the key to supporting ever-evolving data volumes. Large messages cannot
always be avoided. For instance, if a B2B solution is designed to handle EDI traffic, the solution
needs to be prepared to receive the EDI documents up to several megabytes. Every tier, service
and component in the end-to-end flow needs to accommodate the document size that is being
processed. Successfully accepting a 20MB EDI 846 Inventory Advice document via a Web
Service but failing to store it in a queue for processing due to the queues message size
constraints would be considered as unpleasant discovery during testing.
Why would someone choose to use a queue for large messages on the Windows Azure platform?
Whats wrong with other alternatives such as blobs, tables, cloud drives or SQL Azure databases
to say the least? Mind you, the queues allow implementing certain types of messaging scenarios
that are characterized by asynchronous, loosely-coupled communications between producers
and consumers performed in a scalable and reliable fashion. The use of Windows Azure queues
decouples different parts of a given solution and offers unique semantics such as FIFO (First In,
First Out) and At-Least-Once delivery. Such semantics can be somewhat difficult to implement
using the other alternative data exchange mechanisms. Furthermore, the queues are best suited
as a volatile store for exchanging data between services, tiers and applications, not as persistent
data storage. The respective data exchange requirement can manifest itself in many different
forms such as passing messages between components in asynchronous manner, load leveling,
or scaling out complex compute workloads. Many of these data exchange patterns are not
something that can be straightforward to implement without queues. In summary, the queues are

62
a crucial capability. Not having to worry about what can and cannot go into a queue is a strong
argument for building unified queue-based messaging solutions that can handle any data of any
size.
In this article, Im going to implement a solution that will enable to use a Windows Azure queue
for exchanging large messages. I also intend to simplify the way my solution interacts with
Windows Azure queues by providing an abstraction layer built on top of APIs found in the
StorageClient namespace. This abstraction layer will make it easier to publish and consume
instances of the application-specific entities as opposed to having to deal with byte arrays or
strings which are the only types supported by the Queue Service API today. I am going to make
extensive use of .NET generics, will take advantage of some value-add capabilities such as
transparent stream compression and decompression as well as apply some known best practices
such as handling intermittent faults in order to improve fault-tolerance of storage operations.
Design Considerations
As things stand today, a message that is larger than 64KB (after its serialized and encoded)
cannot be stored in a Windows Azure queue. The client-side API will return an exception if you
attempt to place a message larger than 64KB in a queue. The maximum allowed message size
can be determined by inspecting the MaxMessageSize property from the CloudQueueMessage
class. As of the writing of this post, the message size limit returned by this property is 65536.
The maximum message size defined in CloudQueueMessage.MaxMessageSize property
is not reflective of the maximum allowed payload size. Messages are subject to Base64
encoding when they are transmitted to a queue. The encoded payloads are always larger
than their raw data. The Base64 encoding adds 25% overhead on average. As a result,
the 64KB size limit effectively prohibits from storing any messages with payload larger
than 48KB (75% of 64KB).
Even though its the limit for a single queue item, it can be deemed prohibitive for certain types of
messages, especially those that cannot be broken down into smaller chunks. From a developer
perspective, worrying about whether a given message can be accommodated on a queue doesnt
help my productivity. At the end of the day, the goal is to get my application data to flow between
producers and consumers in the most efficient way, regardless of the data size. While one side
calls Put (or Enqueue) and the other side invokes Get (or Dequeue) against a queue, the rest
should theoretically occur automagically.
Overcoming the message size limitation in Windows Azure queues by employing a smart way of
dealing with large messages is the key premises for the technical challenge elaborated in this
article. This will come at the cost of some additional craftsmanship. In the modern world of
commercial software development, any extra development efforts need to be wisely justified. I am
going to justify the additional investments with the following design goals:
Support for very large messages through eliminating any restrictions imposed by the
Queue Service API as it pertains to the message size.
Support for user-defined generic objects when publishing and consuming messages from
a Windows Azure queue.
Important

63
Transparent overflow into a configurable message store either blob container, distributed
cache or other type of repository capable of storing large messages.
Transparent compression that is intended to increase cost-efficiency by minimizing the
amount of storage space consumed by large messages.
Increased reliability in the form of extensive use of the transient condition handling best
practices when performing queue operations.
The foundation for supporting large messages in size-constrained queues will be the following
pattern. First, I check if a given message can be accommodated on a Windows Azure queue
without performing any extra work. The way to determine whether a message can be safely
stored on a queue without violating size constraints will be through a formula which I wrap into a
helper function as follows:
/// <summary>
/// Provides helper methods to enable cloud application code to invoke common, globally
accessible functions.
/// </summary>
public static class CloudUtility
{
/// <summary>
/// Verifies whether or not the specified message size can be accommodated in a
Windows Azure queue.
/// </summary>
/// <param name="size">The message size value to be inspected.</param>
/// <returns>True if the specified size can be accommodated in a Windows Azure queue,
otherwise false.</returns>
public static bool IsAllowedQueueMessageSize(long size)
{
return size >= 0 && size <= (CloudQueueMessage.MaxMessageSize - 1) / 4 * 3;
}
}
If message size is under the enforced limit, I should simply invoke the Queue Service API to
enqueue the message "as is." If the message size is in excess of the limitation in question, the
data flow becomes quite interesting. The following flowchart visualizes the subsequent steps:

64


In summary, if a message cannot be accommodated on a queue due to its size, it overflows into a
message store capable of storing large messages. A tiny metadata message is then created
consisting of a reference to the item in the overflow store. Finally, the metadata message is put
on a queue. I always choose to compress a message before asserting its suitability for
persistence in a queue. This effectively expands the population of messages that can be queued
without incurring the need to go into the overflow store. A good example is an XML document
slightly larger than 64KB which, after serialization and compression is performed, becomes a

65
perfect candidate to be simply put on a queue. You can modify this behavior in case the default
compression is not desirable. It can be achieved by providing a custom serializer component
elaborated in the next section.
There are several considerations that apply here, mainly from a cost perspective. As it can be
noted in the above flowchart, I attempt to determine whether a large message can first overflow
into Windows Azure Caching Service (referred herein as Caching Service for the sake of brevity).
Since the usage of distributed cloud-based caching service is subject to a charge, the cache
overflow path should be made optional. This is reflected on the flowchart.
In addition, there may be situations when the message is quite large and therefore is not suitable
for being stored in a size-constrained distributed cache. As of the writing of this article, the
maximum cache size is 4GB. Therefore, we must take this into consideration and provide a
failover path should we exceed cache capacityor quotas. The quotas come with eviction behavior
that also needs to be accounted for.
The use of the Windows Azure Caching ServHFice as an overflow store helps reduce
latency and eliminate excessive storage transactions when exchanging a large number of
messages. It offers a highly available, distributed caching infrastructure capable of
replicating and maintaining cached data in memory across multiple cache servers for
durability. These benefits can be outweighed by the cache size limitation and costs
associated with the service usage. It is therefore important to perform a cost-benefit
analysis to assess the pros and cons of introducing the Caching Service as an overflow
store in certain scenarios.
Given that the distributed cache storage is limited, it is essential to set out some further rules that
will enable the efficient use of the cache. In connection to this, one important recommendation
needs to be explicitly called out:
Due to specifics of its eviction behavior, Caching Service does not offer a complete and
ultimate guaranteed durability when compared to the Windows Azure Blob Service. When
used as an overflow store, Caching Service is best suited when individual messages are
volatile in nature and are under 8MB in size. The term volatile means that messages
are published into and subsequently consumed as quickly as possible. The 8MB
recommendation is due to the optimal cache item size that is configured in the Caching
Service by default.
Im going to reflect the above recommendation in the code by providing a helper function that will
determine whether or not the specified item size value can be considered as optimal when storing
an item of the given size in the cache.
public static class CloudUtility
{
private static readonly long optimalCacheItemSize = 8 * 1024 * 1024;

/// <summary>
Important
Important

66
/// Determines whether the specified value can be considered as optimal when storing
an item of a given size in the cache.
/// </summary>
/// <param name="size">The item size value to be inspected.</param>
/// <returns>True if the specified size can be considered as optimal, otherwise
false.</returns>
public static bool IsOptimalCacheItemSize(long size)
{
return size >= 0 && size <= optimalCacheItemSize;
}
}
Now that some initial pre-requisites are considered, its time to switch over to the consumer side
and take a look at the implementation pattern for retrieving large messages from a queue. First,
lets visualize the process flow for the purposes of facilitating overall understanding:

67


To summarize the above flow, a message of an unknown type is fetched from a queue and
compared against a metadata message type. If it is not a metadata message, the flow continues
with decompression logic, so that the original message can be correctly reconstructed before
being presented to the consumer. By contrast, if it was in fact a metadata message, it is
inspected to determine the type of overflow store that was used for storing the actual message. If
it is identified as a message stored in the cache, the respective Caching Service API is invoked
and the real message will be fetched before being decompressed and returned to the consumer.
In case the real message was put into a blob container, the Blob Service API will be targeted to
retrieve the real message from the blob entity, decompressed and handed back to the caller.
In addition to handling Enqueue and Dequeue operations for large messages, there is a need to
make sure that all overflowed payloads are removed from their respective overflow message

68
stores upon the consumers request. To accomplish this, one of the potential implementation
patterns is to couple the removal process with the Delete operation when its being invoked for a
given message. The visual representation of this operation can be depicted as follows:


Before we start implementing the patterns mentioned above, one last consideration worth making
is the definition of a message. What would be considered a message, and what forms will it
manifest itself in? Would it be a byte array, a stream of data, a simple type like a string, or a
complex application-specific object which the developer implements as part of the solution object
model? I truly believe that this is the area where we should not constrain ourselves. Lets just
assume that a message is of generic type <T> meaning its anything the developer wishes to use.
You will see that the end implementation will naturally unfold itself around this idea.
Putting all together, the following diagram summarizes all the three possible travel paths which
are accounted for in the above design:

At this
point
discus
Tech
To fol
shipp
discus
Azure
folder
As no
with W
main
throug
infras
queue
/// <s
/// De
/// </
public
{
//
//
//
//
messag
s point, there
onwards, I wi
ssed above.
hnical Im
llow along, do
ed as part of
ssed in this a
e.Services.Fr
r. This locatio
oted at the be
Window Azure
operations su
gh which the
tructure-level
es since these
summary>
efines a gen
/summary>
c interface
// <summary>
// Puts a si
// </summary
// <typepara
ge.</typepar
seem to be e
ill switch the f
mplementa
ownload the fu
a larger end-
rticle. Once d
ramework pr
n contains all
ginning, the o
e queues. I ap
upported by m
contract surfa
l functions fro
e do not add
erics-aware
ICloudQueueS

ngle message
>
m name="T">T
am>
enough input
focus to the s
ation
ull sample co
to-end refere
downloaded a
roject under C
l the main cod
original idea w
pproach this r
my custom sto
aces to consu
om the code s
significant va
abstraction
torage : IDi
on a queue.
he type of t
to start bringi
source code re
de from the M
nce impleme
and unzipped,
Contoso.Clou
de artifacts di
was to abstrac
requirement b
orage abstrac
umers is show
snippet below
alue at this tim
layer for Wi
isposable

the payload a
ing the techni
equired to im
MSDN Code G
ntation which
, navigate to t
ud.Integratio
scussed belo
ct the way a c
by providing a
ction layer. Th
wn below. I in
such as crea
me.
indows Azure
associated wi
ical design to
plement the p
Gallery. The s
h is powered b
the
on and expan
ow.
cloud applicat
a contract tha
he programmi
tentionally om
ation and dele
Queue storag
ith the

6

life. From thi
patterns
sample is
by the pattern
d the Storage
tion interacts
at governs the
ing interface
mitted a few
etion of
ge.
69
s
ns
e
e

70
/// <param name="queueName">The target queue name on which message will be
placed.</param>
/// <param name="message">The payload to be put into a queue.</param>
void Put<T>(string queueName, T message);

/// <summary>
/// Retrieves a single message from the specified queue and applies the default
visibility timeout.
/// </summary>
/// <typeparam name="T">The type of the payload associated with the
message.</typeparam>
/// <param name="queueName">The name of the source queue.</param>
/// <returns>An instance of the object that has been fetched from the
queue.</returns>
T Get<T>(string queueName);

/// <summary>
/// Gets a collection of messages from the specified queue and applies the specified
visibility timeout.
/// </summary>
/// <typeparam name="T">The type of the payload associated with the
message.</typeparam>
/// <param name="queueName">The name of the source queue.</param>
/// <param name="count">The number of messages to retrieve.</param>
/// <param name="visibilityTimeout">The timeout during which retrieved messages will
remain invisible on the queue.</param>
/// <returns>The list of messages retrieved from the specified queue.</returns>
IEnumerable<T> Get<T>(string queueName, int count, TimeSpan visibilityTimeout);

/// <summary>
/// Gets a collection of messages from the specified queue and applies the default
visibility timeout.
/// </summary>
/// <typeparam name="T">The type of the payload associated with the
message.</typeparam>
/// <param name="queueName">The name of the source queue.</param>
/// <param name="count">The number of messages to retrieve.</param>
/// <returns>The list of messages retrieved from the specified queue.</returns>
IEnumerable<T> Get<T>(string queueName, int count);

/// <summary>
/// Deletes a single message from a queue.

71
/// </summary>
/// <typeparam name="T">The type of the payload associated with the
message.</typeparam>
/// <param name="message">The message to be deleted from a queue.</param>
/// <returns>A flag indicating whether or not the specified message has actually been
deleted.</returns>
bool Delete<T>(T message);
}
There is also a need for one extra contract (interface) which will abstract access to the large-
message overflow store. Two components implement the contract, one for each overflow store
(blob storage and distributed cache). The contract is comprised of the following operations:
/// <summary>
/// Defines a generics-aware abstraction layer for Windows Azure Blob storage.
/// </summary>
public interface ICloudBlobStorage : IDisposable
{
/// <summary>
/// Puts a blob into the underlying storage, overwrites if the blob with the same
name already exists.
/// </summary>
/// <typeparam name="T">The type of the payload associated with the blob.</typeparam>
/// <param name="containerName">The target blob container name into which a blob will
be stored.</param>
/// <param name="blobName">The custom name associated with the blob.</param>
/// <param name="blob">The blob's payload.</param>
/// <returns>True if the blob was successfully put into the specified container,
otherwise false.</returns>
bool Put<T>(string containerName, string blobName, T blob);

/// <summary>
/// Puts a blob into the underlying storage. If the blob with the same name already
exists, overwrite behavior can be applied.
/// </summary>
/// <typeparam name="T">The type of the payload associated with the blob.</typeparam>
/// <param name="containerName">The target blob container name into which a blob will
be stored.</param>
/// <param name="blobName">The custom name associated with the blob.</param>
/// <param name="blob">The blob's payload.</param>
/// <param name="overwrite">The flag indicating whether or not overwriting the
existing blob is permitted.</param>
/// <returns>True if the blob was successfully put into the specified container,
otherwise false.</returns>

72
bool Put<T>(string containerName, string blobName, T blob, bool overwrite);

/// <summary>
/// Retrieves a blob by its name from the underlying storage.
/// </summary>
/// <typeparam name="T">The type of the payload associated with the blob.</typeparam>
/// <param name="containerName">The target blob container name from which the blob
will be retrieved.</param>
/// <param name="blobName">The custom name associated with the blob.</param>
/// <returns>An instance of <typeparamref name="T"/> or default(T) if the specified
blob was not found.</returns>
T Get<T>(string containerName, string blobName);

/// <summary>
/// Deletes the specified blob.
/// </summary>
/// <param name="containerName">The target blob container name from which the blob
will be deleted.</param>
/// <param name="blobName">The custom name associated with the blob.</param>
/// <returns>True if the blob was deleted, otherwise false.</returns>
bool Delete(string containerName, string blobName);
}
Both contracts heavily rely on generic type <T>. It enables you to tailor the message type to any
.NET type of your choice. I will however have to handle some extreme use cases, namely types
that require special treatment, such as streams. I expand on this later on.
Regardless of the message type chosen, one important requirement will apply; the object type
that represents a message on a queue must be serializable. All objects passing through the
storage abstraction layer are subject to serialization before they land on a queue or overflow
store. In my implementation, serialization and deserialization are also coupled with compression
and decompression, respectively. This approach increases efficiency from a cost and bandwidth
perspective. The cost-related benefit comes from the fact that compressed large messages
inherently consume less storage, resulting in a decrease in storage costs. The bandwidth
efficiency arises from savings on payload size thanks to compression, which in turn makes
payloads smaller on the wire as they flow to and from the Windows Azure storage.
The requirement for serialization and deserialization is declared in a specialized interface. Any
component that implements this interface must provide the specific compression, serialization,
deserialization, and decompression functionality. An example of this interface is shown:
/// <summary>
/// Defines a contract that must be supported by a component which performs serialization
and
/// deserialization of storage objects such as Azure queue items, blobs and table
entries.

73
/// </summary>
public interface ICloudStorageEntitySerializer
{
/// <summary>
/// Serializes the object to the specified stream.
/// </summary>
/// <param name="instance">The object instance to be serialized.</param>
/// <param name="target">The destination stream into which the serialized object will
be written.</param>
void Serialize(object instance, Stream target);

/// <summary>
/// Deserializes the object from specified data stream.
/// </summary>
/// <param name="source">The source stream from which serialized object will be
consumed.</param>
/// <param name="type">The type of the object that will be deserialized.</param>
/// <returns>The deserialized object instance.</returns>
object Deserialize(Stream source, Type type);
}
For compression and decompression, I use the DeflateStream component in the .NET
Framework. This class represents the Deflate algorithm, an industry standard RFC-compliant
algorithm for lossless file compression and decompression. In comparison to the GZipStream
class, the former produces more optimal compressed images and generally delivers better
performance. By contrast, the GZipStream class uses the GZIP data format, which includes a
cyclic redundancy check (CRC) value for detecting data corruption. Behind the scenes, the GZIP
data format uses the same compression algorithm as the DeflateStream class. In summary,
GZipStream = DeflateStream + the cost of calculating and storing CRC checksums.
My implementation of the contract is included below. Note that compression algorithms can be
easily toggled by replacing DeflateStream class with GZipStream and vice versa.
/// <summary>
/// Provides a default implementation of ICloudStorageEntitySerializer which performs
serialization and
/// deserialization of storage objects such as Azure queue items, blobs and table
entries.
/// </summary>
internal sealed class CloudStorageEntitySerializer : ICloudStorageEntitySerializer
{
/// <summary>
/// Serializes the object to the specified stream.
/// </summary>
/// <param name="instance">The object instance to be serialized.</param>

74
/// <param name="target">The destination stream into which the serialized object will
be written.</param>
public void Serialize(object instance, Stream target)
{
Guard.ArgumentNotNull(instance, "instance");
Guard.ArgumentNotNull(target, "target");

XDocument xmlDocument = null;
XElement xmlElement = null;
XmlDocument domDocument = null;
XmlElement domElement = null;

if ((xmlElement = (instance as XElement)) != null)
{
// Handle XML element serialization using separate technique.
SerializeXml<XElement>(xmlElement, target, (xml, writer) => {
xml.Save(writer); });
}
else if ((xmlDocument = (instance as XDocument)) != null)
{
// Handle XML document serialization using separate technique.
SerializeXml<XDocument>(xmlDocument, target, (xml, writer) => {
xml.Save(writer); });
}
else if ((domDocument = (instance as XmlDocument)) != null)
{
// Handle XML DOM document serialization using separate technique.
SerializeXml<XmlDocument>(domDocument, target, (xml, writer) => {
xml.Save(writer); });
}
else if ((domElement = (instance as XmlElement)) != null)
{
// Handle XML DOM element serialization using separate technique.
SerializeXml<XmlElement>(domElement, target, (xml, writer) => {
xml.WriteTo(writer); });
}
else
{
var serializer = GetXmlSerializer(instance.GetType());

using (var compressedStream = new DeflateStream(target,
CompressionMode.Compress, true))

75
using (var xmlWriter =
XmlDictionaryWriter.CreateBinaryWriter(compressedStream, null, null, false))
{
serializer.WriteObject(xmlWriter, instance);
}
}
}

/// <summary>
/// Deserializes the object from specified data stream.
/// </summary>
/// <param name="source">The source stream from which serialized object will be
consumed.</param>
/// <param name="type">The type of the object that will be deserialized.</param>
/// <returns>The deserialized object instance.</returns>
public object Deserialize(Stream source, Type type)
{
Guard.ArgumentNotNull(source, "source");
Guard.ArgumentNotNull(type, "type");

if (type == typeof(XElement))
{
// Handle XML element deserialization using separate technique.
return DeserializeXml<XElement>(source, (reader) => { return
XElement.Load(reader); });
}
else if (type == typeof(XDocument))
{
// Handle XML document deserialization using separate technique.
return DeserializeXml<XDocument>(source, (reader) => { return
XDocument.Load(reader); });
}
else if (type == typeof(XmlDocument))
{
// Handle XML DOM document deserialization using separate technique.
return DeserializeXml<XmlDocument>(source, (reader) => { var xml = new
XmlDocument(); xml.Load(reader); return xml; });
}
else if (type == typeof(XmlElement))
{
// Handle XML DOM element deserialization using separate technique.

76
return DeserializeXml<XmlElement>(source, (reader) => { var xml = new
XmlDocument(); xml.Load(reader); return xml.DocumentElement; });
}
else
{
var serializer = GetXmlSerializer(type);

using (var compressedStream = new DeflateStream(source,
CompressionMode.Decompress, true))
using (var xmlReader =
XmlDictionaryReader.CreateBinaryReader(compressedStream, XmlDictionaryReaderQuotas.Max))
{
return serializer.ReadObject(xmlReader);
}
}
}

private XmlObjectSerializer GetXmlSerializer(Type type)
{
if (FrameworkUtility.GetDeclarativeAttribute<DataContractAttribute>(type) !=
null)
{
return new DataContractSerializer(type);
}
else
{
return new NetDataContractSerializer();
}
}

private void SerializeXml<T>(T instance, Stream target, Action<T, XmlWriter>
serializeAction)
{
using (var compressedStream = new DeflateStream(target, CompressionMode.Compress,
true))
using (var xmlWriter = XmlDictionaryWriter.CreateBinaryWriter(compressedStream,
null, null, false))
{
serializeAction(instance, xmlWriter);

xmlWriter.Flush();
compressedStream.Flush();

77
}
}

private T DeserializeXml<T>(Stream source, Func<XmlReader, T> deserializeAction)
{
using (var compressedStream = new DeflateStream(source,
CompressionMode.Decompress, true))
using (var xmlReader = XmlDictionaryReader.CreateBinaryReader(compressedStream,
XmlDictionaryReaderQuotas.Max))
{
return deserializeAction(xmlReader);
}
}
}
One of the powerful capabilities in the CloudStorageEntitySerializer implementation is the
ability to apply special treatment when handling XML documents of both flavors: XmlDocument
and XDocument. The other area worth highlighting is the optimal serialization and deserialization
of the XML data. Here I decided to take advantage of the XmlDictionaryReader and
XmlDictionaryWriter classes which are known to .NET developers as a superb choice when it
comes to performing efficient serialization and deserialization of XML payloads by using the .NET
Binary XML format.
The decision as to the type of overflow message store is the responsibility of the consumer which
calls into the custom, storage abstraction layer. Along these lines, Im going to provide an option
to select the desired message store type by adding the following constructors in the type that
implements the ICloudQueueStorage interface:
/// <summary>
/// Provides reliable generics-aware access to the Windows Azure Queue storage.
/// </summary>
public sealed class ReliableCloudQueueStorage : ICloudQueueStorage
{
private readonly RetryPolicy retryPolicy;
private readonly CloudQueueClient queueStorage;
private readonly ICloudStorageEntitySerializer dataSerializer;
private readonly ICloudBlobStorage overflowStorage;
private readonly ConcurrentDictionary<object, InflightMessageInfo> inflightMessages;

/// <summary>
/// Initializes a new instance of the <see cref="ReliableCloudQueueStorage"/> class
using the specified storage account information,
/// custom retry policy, custom implementation of <see
cref="ICloudStorageEntitySerializer"/> interface and custom implementation of
/// the large message overflow store.

78
/// </summary>
/// <param name="storageAccountInfo">The storage account that is projected through
this component.</param>
/// <param name="retryPolicy">The specific retry policy that will ensure reliable
access to the underlying storage.</param>
/// <param name="dataSerializer">The component which performs serialization and
deserialization of storage objects.</param>
/// <param name="overflowStorage">The component implementing overflow store that will
be used for persisting large messages that
/// cannot be accommodated in a queue due to message size constraints.</param>
public ReliableCloudQueueStorage(StorageAccountInfo storageAccountInfo, RetryPolicy
retryPolicy, ICloudStorageEntitySerializer dataSerializer, ICloudBlobStorage
overflowStorage)
{
Guard.ArgumentNotNull(storageAccountInfo, "storageAccountInfo");
Guard.ArgumentNotNull(retryPolicy, "retryPolicy");
Guard.ArgumentNotNull(dataSerializer, "dataSerializer");
Guard.ArgumentNotNull(overflowStorage, "overflowStorage");

this.retryPolicy = retryPolicy;
this.dataSerializer = dataSerializer;
this.overflowStorage = overflowStorage;

CloudStorageAccount storageAccount = new CloudStorageAccount(new
StorageCredentialsAccountAndKey(storageAccountInfo.AccountName,
storageAccountInfo.AccountKey), true);
this.queueStorage = storageAccount.CreateCloudQueueClient();

// Configure the Queue storage not to enforce any retry policies since this is
something that we will be dealing ourselves.
this.queueStorage.RetryPolicy = RetryPolicies.NoRetry();

this.inflightMessages = new ConcurrentDictionary<object,
InflightMessageInfo>(Environment.ProcessorCount * 4,
InflightMessageQueueInitialCapacity);
}
}
The above constructors are not performing any complex work; they simply initialize the internal
members and configure the client component that will be accessing a Windows Azure queue. Its
worth noting however that I explicitly tell the queue client not to enforce any retry policy. In order
to provide a robust and reliable storage abstraction layer, I need more granular control over
transient issues when performing operations against Windows Azure queues. Therefore, there

79
will be a separate component that recognizes and is able to handle a much larger variety of
intermittent faults.
Lets now take a look at the internals of the ReliableCloudQueueStorage class which I drafted
above. Specifically, lets review its implementation of the Put operation since this is the location of
the transparent overflow into a large message store.
/// <summary>
/// Puts a single message on a queue.
/// </summary>
/// <typeparam name="T">The type of the payload associated with the message.</typeparam>
/// <param name="queueName">The target queue name on which message will be
placed.</param>
/// <param name="message">The payload to be put into a queue.</param>
public void Put<T>(string queueName, T message)
{
Guard.ArgumentNotNullOrEmptyString(queueName, "queueName");
Guard.ArgumentNotNull(message, "message");

// Obtain a reference to the queue by its name. The name will be validated against
compliance with storage resource names.
var queue =
this.queueStorage.GetQueueReference(CloudUtility.GetSafeContainerName(queueName));

CloudQueueMessage queueMessage = null;

// Allocate a memory buffer into which messages will be serialized prior to being put
on a queue.
using (MemoryStream dataStream = new
MemoryStream(Convert.ToInt32(CloudQueueMessage.MaxMessageSize)))
{
// Perform serialization of the message data into the target memory buffer.
this.dataSerializer.Serialize(message, dataStream);

// Reset the position in the buffer as we will be reading its content from the
beginning.
dataStream.Seek(0, SeekOrigin.Begin);

// First, determine whether the specified message can be accommodated on a queue.
if (CloudUtility.IsAllowedQueueMessageSize(dataStream.Length))
{
queueMessage = new CloudQueueMessage(dataStream.ToArray());
}
else

80
{
// Create an instance of a large queue item metadata message.
LargeQueueMessageInfo queueMsgInfo = LargeQueueMessageInfo.Create(queueName);

// Persist the stream of data that represents a large message into the
overflow message store.
this.overflowStorage.Put<Stream>(queueMsgInfo.ContainerName,
queueMsgInfo.BlobReference, dataStream);

// Invoke the Put operation recursively to enqueue the metadata message.
Put<LargeQueueMessageInfo>(queueName, queueMsgInfo);
}
}
// Check if a message is available to be put on a queue.
if (queueMessage != null)
{
Put(queue, queueMessage);
}
}
A new code artifact that has just manifested itself in the snippet above is the
LargeQueueMessageInfo class. This custom type is ultimately our metadata message that
describes the location of a large message. This class is marked as internal as its not intended to
be visible to anyone outside the storage abstraction layer implementation. The class is defined as
follows:
/// <summary>
/// Implements an object holding metadata related to a large message which is stored in
/// the overflow message store such as Windows Azure blob container.
/// </summary>
[DataContract(Namespace = WellKnownNamespace.DataContracts.General)]
internal sealed class LargeQueueMessageInfo
{
private const string ContainerNameFormat = "LargeMsgCache-{0}";

/// <summary>
/// Returns the name of the blob container holding the large message payload.
/// </summary>
[DataMember]
public string ContainerName { get; private set; }

/// <summary>
/// Returns the unique reference to a blob holding the large message payload.
/// </summary>

81
[DataMember]
public string BlobReference { get; private set; }

/// <summary>
/// The default constructor is inaccessible, the object needs to be instantiated
using its Create method.
/// </summary>
private LargeQueueMessageInfo() { }

/// <summary>
/// Creates a new instance of the large message metadata object and allocates a
globally unique blob reference.
/// </summary>
/// <param name="queueName">The name of the Windows Azure queue on which a reference
to the large message will be stored.</param>
/// <returns>The instance of the large message metadata object.</returns>
public static LargeQueueMessageInfo Create(string queueName)
{
Guard.ArgumentNotNullOrEmptyString(queueName, "queueName");

return new LargeQueueMessageInfo() { ContainerName =
String.Format(ContainerNameFormat, queueName), BlobReference =
Guid.NewGuid().ToString("N") };
}
}
Moving forward, I need to implement a large message overflow store that will leverage the
Windows Azure Blob Storage service. As I pointed out earlier, this component must support the
ICloudBlobStorage interface that will be consumed by the ReliableCloudQueueStorage
component to relay messages into the ICloudBlobStorage implementation whenever these
cannot be accommodated on a queue due to message size limitation. To set the stage for the
next steps, I will include the constructors implementation only:
/// <summary>
/// Implements reliable generics-aware layer for Windows Azure Blob storage.
/// </summary>
public class ReliableCloudBlobStorage : ICloudBlobStorage
{
private readonly RetryPolicy retryPolicy;
private readonly CloudBlobClient blobStorage;
private readonly ICloudStorageEntitySerializer dataSerializer;

/// <summary>

82
/// Initializes a new instance of the ReliableCloudBlobStorage class using the
specified storage account info, custom retry
/// policy and custom implementation of ICloudStorageEntitySerializer interface.
/// </summary>
/// <param name="storageAccountInfo">The access credentials for Windows Azure storage
account.</param>
/// <param name="retryPolicy">The custom retry policy that will ensure reliable
access to the underlying storage.</param>
/// <param name="dataSerializer">The component which performs
serialization/deserialization of storage objects.</param>
public ReliableCloudBlobStorage(StorageAccountInfo storageAccountInfo, RetryPolicy
retryPolicy, ICloudStorageEntitySerializer dataSerializer)
{
Guard.ArgumentNotNull(storageAccountInfo, "storageAccountInfo");
Guard.ArgumentNotNull(retryPolicy, "retryPolicy");
Guard.ArgumentNotNull(dataSerializer, "dataSerializer");

this.retryPolicy = retryPolicy;
this.dataSerializer = dataSerializer;

CloudStorageAccount storageAccount = new CloudStorageAccount(new
StorageCredentialsAccountAndKey(storageAccountInfo.AccountName,
storageAccountInfo.AccountKey), true);
this.blobStorage = storageAccount.CreateCloudBlobClient();

// Configure the Blob storage not to enforce any retry policies since this is
something that we will be dealing ourselves.
this.blobStorage.RetryPolicy = RetryPolicies.NoRetry();

// Disable parallelism in blob upload operations to reduce the impact of multiple
concurrent threads on parallel upload feature.
this.blobStorage.ParallelOperationThreadCount = 1;
}
}
Earlier in this article, I have shown the implementation of the Put operation which ensures that
small messages will always be placed on a queue whereas large messages will be transparently
routed into the overflow store. For the sake of continuity, lets now review the mechanics behind
the counterpart Put operation implemented by the overflow store.
/// <summary>
/// Puts a blob into the underlying storage, overwrites the existing blob if the blob
with the same name already exists.
/// </summary>

83
private bool Put<T>(string containerName, string blobName, T blob, bool overwrite, string
expectedEtag, out string actualEtag)
{
Guard.ArgumentNotNullOrEmptyString(containerName, "containerName");
Guard.ArgumentNotNullOrEmptyString(blobName, "blobName");
Guard.ArgumentNotNull(blob, "blob");

var callToken = TraceManager.CloudStorageComponent.TraceIn(containerName, blobName,
overwrite, expectedEtag);

// Verify whether or not the specified blob is already of type Stream.
Stream blobStream = IsStreamType(blob.GetType()) ? blob as Stream : null;
Stream blobData = null;
actualEtag = null;

try
{
// Are we dealing with a stream already? If yes, just use it as is.
if (blobStream != null)
{
blobData = blobStream;
}
else
{
// The specified blob is something else rather than a Stream, we perform
serialization of T into a new stream instance.
blobData = new MemoryStream();
this.dataSerializer.Serialize(blob, blobData);
}

var container =
this.blobStorage.GetContainerReference(CloudUtility.GetSafeContainerName(containerName));
StorageErrorCode lastErrorCode = StorageErrorCode.None;

Func<string> uploadAction = () =>
{
var cloudBlob = container.GetBlobReference(blobName);
return UploadBlob(cloudBlob, blobData, overwrite, expectedEtag);
};

try
{

84
// First attempt - perform upload and let the UploadBlob method handle any
retry conditions.
string eTag = uploadAction();

if (!String.IsNullOrEmpty(eTag))
{
actualEtag = eTag;
return true;
}
}
catch (StorageClientException ex)
{
lastErrorCode = ex.ErrorCode;

if (!(lastErrorCode == StorageErrorCode.ContainerNotFound || lastErrorCode ==
StorageErrorCode.ResourceNotFound || lastErrorCode ==
StorageErrorCode.BlobAlreadyExists))
{
// Anything other than "not found" or "already exists" conditions will be
considered as a runtime error.
throw;
}
}

if (lastErrorCode == StorageErrorCode.ContainerNotFound)
{
// Failover action #1: create the target container and try again. This time,
use a retry policy to wrap calls to the
// UploadBlob method.
string eTag = this.retryPolicy.ExecuteAction<string>(() =>
{
CreateContainer(containerName);
return uploadAction();
});

return !String.IsNullOrEmpty(actualEtag = eTag);
}

if (lastErrorCode == StorageErrorCode.BlobAlreadyExists && overwrite)
{
// Failover action #2: Overwrite was requested but BlobAlreadyExists has
still been returned.

85
// Delete the original blob and try to upload again.
string eTag = this.retryPolicy.ExecuteAction<string>(() =>
{
var cloudBlob = container.GetBlobReference(blobName);
cloudBlob.DeleteIfExists();

return uploadAction();
});

return !String.IsNullOrEmpty(actualEtag = eTag);
}
}
finally
{
// Only dispose the blob data stream if it was newly created.
if (blobData != null && null == blobStream)
{
blobData.Dispose();
}

TraceManager.CloudStorageComponent.TraceOut(callToken, actualEtag);
}

return false;
}
In summary, the above code takes a blob of type <T> and first checks if this is already a
serialized image of a message in the form of a Stream object. All large messages that are
relayed to the overflow storage by the ReliableCloudQueueStorage component will arrive as
streams, ready for persistence. Next, the UploadBlob action is invoked, which in turn calls into
Blob Service Client API, specifically its UploadFromStream operation. If a large message blob
fails to upload successfully, the code inspects the error returned by the Blob Service and provides
a failover path for 2 conditions: ContainerNotFound and BlobAlreadyExists. In the event that the
target blob container is not found, the code will attempt to create the missing container. It
performs this action within a retry-aware scope to improve reliability and increase resilience to
transient failures. The second failover path is intended to handle a situation where a blob with the
same name already exists. The code will remove the existing blob, provided the overwrite
behavior is enabled. After removal, the upload of the new blob will be retried. Again, this
operation is performed inside a retry-aware scope for increased reliability.
Now that I can store large messages in a blob container, its time to design another
implementation of the ICloudBlobStorage interface that will leverage the Windows Azure
Caching Service. For consistency, lets start off with its constructors:
/// <summary>

86
/// Implements reliable generics-aware layer for Windows Azure Caching Service.
/// </summary>
public class ReliableCloudCacheStorage : ICloudBlobStorage
{
private readonly RetryPolicy retryPolicy;
private readonly ICloudStorageEntitySerializer dataSerializer;
private readonly DataCacheFactory cacheFactory;
private readonly DataCache cache;

/// <summary>
/// Initializes a new instance of the ReliableCloudCacheStorage class using the
specified storage account information
/// custom retry policy and custom implementation of ICloudStorageEntitySerializer
interface.
/// </summary>
/// <param name="endpointInfo">The endpoint details for Windows Azure Caching
Service.</param>
/// <param name="retryPolicy">The custom retry policy that will ensure reliable
access to the Caching Service.</param>
/// <param name="dataSerializer">The component which performs custom serialization
and deserialization of cache items.</param>
public ReliableCloudCacheStorage(CachingServiceEndpointInfo endpointInfo, RetryPolicy
retryPolicy, ICloudStorageEntitySerializer dataSerializer)
{
Guard.ArgumentNotNull(endpointInfo, "endpointInfo");
Guard.ArgumentNotNull(retryPolicy, "retryPolicy");
Guard.ArgumentNotNull(dataSerializer, "dataSerializer");

this.retryPolicy = retryPolicy;
this.dataSerializer = dataSerializer;

var cacheServers = new List<DataCacheServerEndpoint>(1);
cacheServers.Add(new DataCacheServerEndpoint(endpointInfo.ServiceHostName,
endpointInfo.CachePort));

var cacheConfig = new DataCacheFactoryConfiguration()

87
{
Servers = cacheServers,
MaxConnectionsToServer = 1,
IsCompressionEnabled = false,
SecurityProperties = new
DataCacheSecurity(endpointInfo.SecureAuthenticationToken, endpointInfo.SslEnabled),
// The ReceiveTimeout value has been modified as per recommendations provided
in
// http://blogs.msdn.com/b/akshar/archive/2011/05/01/azure-appfabric-caching-
errorcode-lt-errca0017-gt-substatus-lt-es0006-gt-what-to-do.aspx
TransportProperties = new DataCacheTransportProperties() { ReceiveTimeout =
TimeSpan.FromSeconds(45) }
};

this.cacheFactory = new DataCacheFactory(cacheConfig);
this.cache = this.retryPolicy.ExecuteAction<DataCache>(() =>
{
return this.cacheFactory.GetDefaultCache();
});
}
}
If you recall from earlier considerations, one of the key technical design decisions was to take
advantage of both the Blob Service and Caching Service for storing large messages. The cache
option is mostly suited for transient objects not exceeding the recommended payload size of
8MB. The blob option is essentially for everything else. Overall, this decision introduces the need
for a hybrid overflow store. The foundation for building a hybrid store is already in the
codebase. Its just the matter of marrying the existing artifacts together as follows:
/// <summary>
/// Implements reliable generics-aware storage layer combining Windows Azure Blob storage
and
/// Windows Azure Caching Service in a hybrid mode.
/// </summary>
public class ReliableHybridBlobStorage : ICloudBlobStorage
{
private readonly ICloudBlobStorage blobStorage;
private readonly ICloudBlobStorage cacheStorage;
private readonly ICloudStorageEntitySerializer dataSerializer;
private readonly IList<ICloudBlobStorage> storageList;


88
/// <summary>
/// Initializes a new instance of the ReliableHybridBlobStorage class using the
specified storage account information, caching
/// service endpoint, custom retry policies and a custom implementation of
ICloudStorageEntitySerializer interface.
/// </summary>
/// <param name="storageAccountInfo">The access credentials for Windows Azure storage
account.</param>
/// <param name="storageRetryPolicy">The custom retry policy that will ensure
reliable access to the underlying blob storage.</param>
/// <param name="cacheEndpointInfo">The endpoint details for Windows Azure Caching
Service.</param>
/// <param name="cacheRetryPolicy">The custom retry policy that will ensure reliable
access to the Caching Service.</param>
/// <param name="dataSerializer">The component which performs serialization and
deserialization of storage objects.</param>
public ReliableHybridBlobStorage(StorageAccountInfo storageAccountInfo, RetryPolicy
storageRetryPolicy, CachingServiceEndpointInfo cacheEndpointInfo, RetryPolicy
cacheRetryPolicy, ICloudStorageEntitySerializer dataSerializer)
{
Guard.ArgumentNotNull(storageAccountInfo, "storageAccountInfo");
Guard.ArgumentNotNull(storageRetryPolicy, "storageRetryPolicy");
Guard.ArgumentNotNull(cacheEndpointInfo, "cacheEndpointInfo");
Guard.ArgumentNotNull(cacheRetryPolicy, "cacheRetryPolicy");
Guard.ArgumentNotNull(dataSerializer, "dataSerializer");

this.dataSerializer = dataSerializer;
this.storageList = new List<ICloudBlobStorage>(2);

this.storageList.Add(this.cacheStorage = new
ReliableCloudCacheStorage(cacheEndpointInfo, cacheRetryPolicy, dataSerializer));
this.storageList.Add(this.blobStorage = new
ReliableCloudBlobStorage(storageAccountInfo, storageRetryPolicy, dataSerializer));
}
}
At this point, I conclude the saga by including one more code snippet showing the implementation
of the Put operation in the hybrid overflow store.

/// <summary>
/// Puts a blob into the underlying storage. If the blob with the same name already
exists, overwrite behavior can be customized.

89
/// </summary>
/// <typeparam name="T">The type of the payload associated with the blob.</typeparam>
/// <param name="containerName">The target blob container name into which a blob will be
stored.</param>
/// <param name="blobName">The custom name associated with the blob.</param>
/// <param name="blob">The blob's payload.</param>
/// <param name="overwrite">The flag indicating whether or not overwriting the existing
blob is permitted.</param>
/// <returns>True if the blob was successfully put into the specified container,
otherwise false.</returns>
public bool Put<T>(string containerName, string blobName, T blob, bool overwrite)
{
Guard.ArgumentNotNull(blob, "blob");

bool success = false;
Stream blobData = null;
bool treatBlobAsStream = false;

try
{
// Are we dealing with a stream already? If yes, just use it as is.
if (IsStreamType(blob.GetType()))
{
blobData = blob as Stream;
treatBlobAsStream = true;
}
else
{
// The specified item type is something else rather than a Stream, we perform
serialization of T into a new stream instance.
blobData = new MemoryStream();

this.dataSerializer.Serialize(blob, blobData);
blobData.Seek(0, SeekOrigin.Begin);
}

try
{
// First, make an attempt to store the blob in the distributed cache.
// Only use cache if blob size is optimal for this type of storage.
if (CloudUtility.IsOptimalCacheItemSize(blobData.Length))
{

90
success = this.cacheStorage.Put<Stream>(containerName, blobName,
blobData, overwrite);
}
}
finally
{
if (!success)
{
// The cache option was unsuccessful, fail over to the blob storage as
per design decision.
success = this.blobStorage.Put<Stream>(containerName, blobName, blobData,
overwrite);
}
}
}
finally
{
if (!treatBlobAsStream && blobData != null)
{
// Only dispose the blob data stream if it was newly created.
blobData.Dispose();
}
}

return success;
}
This article would be considered incomplete if I failed to provide some examples of how the
storage abstraction layer discussed above can be consumed from a client application. I will
combine these examples with a test application that will also validate the technical
implementation.
Validation
In order to prove that large messages can successfully pass back and forth through the newly
implemented storage abstraction layer, a very simple console application was put together. In the
first step, it takes a sample XML document of 90MB in size and puts it on a Windows Azure
queue. In the second step, it consumes a message from the queue. The message should indeed
be the original XML document which is written back to the disk under a different name to be able
to compare the file size and its content. In between these steps, the application enters a pause
mode during which you can explore the content of the queue and respective message overflow
store such as cache or blob container. The source code for the test application is provided below.
using System;
using System.IO;

91
using System.Configuration;
using System.Xml.Linq;

using Contoso.Cloud.Integration.Framework;
using Contoso.Cloud.Integration.Framework.Configuration;
using Contoso.Cloud.Integration.Azure.Services.Framework.Storage;

namespace LargeQueueMessageTest
{
class Program
{
static void Main(string[] args)
{
// Check if command line arguments were in fact supplied.
if (null == args || args.Length == 0) return;

// Read storage account and caching configuration sections.
var cacheServiceSettings =
ConfigurationManager.GetSection("CachingServiceConfiguration") as
CachingServiceConfigurationSettings;
var storageAccountSettings =
ConfigurationManager.GetSection("StorageAccountConfiguration") as
StorageAccountConfigurationSettings;

// Retrieve cache endpoint and specific storage account definitions.
var cacheServiceEndpoint =
cacheServiceSettings.Endpoints.Get(cacheServiceSettings.DefaultEndpoint);
var queueStorageAccount =
storageAccountSettings.Accounts.Get(storageAccountSettings.DefaultQueueStorage);
var blobStorageAccount =
storageAccountSettings.Accounts.Get(storageAccountSettings.DefaultBlobStorage);

PrintInfo("Using storage account definition: {0}",
queueStorageAccount.AccountName);
PrintInfo("Using caching service endpoint name: {0}",
cacheServiceEndpoint.Name);

string fileName = args[0], queueName = "LargeMessageQueue";
string newFileName = String.Format("{0}_Copy{1}",
Path.GetFileNameWithoutExtension(fileName), Path.GetExtension(fileName));

long fileSize = -1, newFileSize = -1;

92

try
{
// Load the specified file into XML DOM.
XDocument largeXmlDoc = XDocument.Load(fileName);

// Instantiate the large message overflow store and use it to instantiate
a queue storage abstraction component.
using (var overflowStorage = new
ReliableHybridBlobStorage(blobStorageAccount, cacheServiceEndpoint))
using (var queueStorage = new
ReliableCloudQueueStorage(queueStorageAccount, overflowStorage))
{
PrintInfo("\nAttempting to store a message of {0} bytes in size on a
Windows Azure queue", fileSize = (new FileInfo(fileName)).Length);

// Enqueue the XML document. The document's size doesn't really
matter any more.
queueStorage.Put<XDocument>(queueName, largeXmlDoc);

PrintSuccess("The message has been succcessfully placed into a
queue.");
PrintWaitMsg("\nYou can now inspect the content of the {0} queue and
respective blob container...", queueName);

// Dequeue a message from the queue which is expected to be our
original XML document.
XDocument docFromQueue = queueStorage.Get<XDocument>(queueName);

// Save it under a new name.
docFromQueue.Save(newFileName);

// Delete the message. Should remove the metadata message from the
queue as well as blob holding the message data.
queueStorage.Delete<XDocument>(docFromQueue);

PrintInfo("\nThe message retrieved from the queue is {0} bytes in
size.", newFileSize = (new FileInfo(newFileName)).Length);

// Perform very basic file size-based comparison. In the reality, we
should have checked the document structurally.
if (fileSize > 0 && newFileSize > 0 && fileSize == newFileSize)

93
{
PrintSuccess("Test passed. This is expected behavior in any code
written by CAT.");
}
else
{
PrintError("Test failed. This should have never happened in the
code written by CAT.");
}
}
}
catch (Exception ex)
{
PrintError("ERROR: {0}", ExceptionTextFormatter.Format(ex));
}
finally
{
Console.ReadLine();
}
}

private static void PrintInfo(string format, params object[] parameters)
{
Console.ForegroundColor = ConsoleColor.White;
Console.WriteLine(format, parameters);
Console.ResetColor();
}

private static void PrintSuccess(string format, params object[] parameters)
{
Console.ForegroundColor = ConsoleColor.Green;
Console.WriteLine(format, parameters);
Console.ResetColor();
}

private static void PrintError(string format, params object[] parameters)
{
Console.ForegroundColor = ConsoleColor.Red;
Console.WriteLine(format, parameters);
Console.ResetColor();
}


94
private static void PrintWaitMsg(string format, params object[] parameters)
{
Console.ForegroundColor = ConsoleColor.Gray;
Console.WriteLine(format, parameters);
Console.ResetColor();
Console.ReadLine();
}
}
}
For the sake of completeness, below is the application configuration file that was used during
testing. If you are going to try the test application out, please make sure you modify your copy of
app.config and add the actual storage account credentials and caching service endpoint
information.
<?xml version="1.0"?>
<configuration>
<configSections>
<section name="CachingServiceConfiguration"
type="Contoso.Cloud.Integration.Framework.Configuration.CachingServiceConfigurationSettin
gs, Contoso.Cloud.Integration.Framework, Version=1.0.0.0, Culture=neutral,
PublicKeyToken=23eafc3765008062"/>
<section name="StorageAccountConfiguration"
type="Contoso.Cloud.Integration.Framework.Configuration.StorageAccountConfigurationSettin
gs, Contoso.Cloud.Integration.Framework, Version=1.0.0.0, Culture=neutral,
PublicKeyToken=23eafc3765008062"/>
</configSections>

<CachingServiceConfiguration defaultEndpoint="YOUR-CACHE-NAMESPACE-GOES-HERE">
<add name="YOUR-CACHE-NAMESPACE-GOES-HERE" authToken="YOUR-CACHE-SECURITYTOKEN-GOES-
HERE"/>
</CachingServiceConfiguration>

<StorageAccountConfiguration defaultBlobStorage="My Azure Storage"
defaultQueueStorage="My Azure Storage">
<add name="My Azure Storage" accountName="YOUR-STORAGE-ACCOUNT-NAME-GOES-HERE"
accountKey="YOUR-STORAGE-ACCOUNT-KEY-GOES-HERE"/>
</StorageAccountConfiguration>
</configuration>
Provided the test application has been successfully compiled and executed, the output similar to
the following is expected to appear in the console windows:

If you
appea

Since
scree
applic

Note
the 87
serial
first a
Once
expec
scree
peek into the
ar on the que
e the test mes
nshot depicts
cation is paus
how the origin
7% savings o
ization. Given
nd best choic
the test appl
cted to be rem
nshot below:
e storage acc
ue:
ssage was lar
s the expected
sed:
nal 90MB XM
on storage a
n the target cl
ce.
ication procee
moved along w
count used by
ge enough to
d content insi
ML document
nd bandwidt
lass of scena
eds with dele
with the blob
y the test appl
o overflow dire
ide the respec
used in my te
th which was
rios, XML bin
tion of the qu
holding the m
ication, the fo
ectly into the
ctive blob con
est became a
the result of
nary serializat
ueue message
message data
ollowing mess
blob storage,
ntainer while t
11MB blob. T
applying XML
ion + compre
e, the metada
a as shown on

9

sage will

the following
the test

This reflects
L binary
ession is the
ata message
n the
95
g
is

The e
intend
routin
guess
Con
As we
larger
Servic
with a
quality
T
th
T
G
A
As I m
overfl
carefu
projec
cache
While
Azure
are no
T
co
T
go
A
bl
A
th
st
example show
ded to highlig
ng into blob st
s, its now the
nclusion
e have seen,
r than 64KB b
ce without ad
a little extra w
y of life impro
ransparent m
he datacenter
ransparent,
Generics sup
Automatic ha
mentioned ear
ow storage, t
ully evaluate t
cted number o
e.
e this solution
e queues, the
ot incorporate
he ability to c
onfiguration.
he additional
oals or functio
An item in the
lob storage a
A garbage col
he overflow m
torage abstra
wn above refle
ht the fundam
tore, transpar
e right time to
the use of Wi
by leveraging
ding any add
work you can e
ovements suc
message com
r.
easily custo
port that allow
ndling of tra
rlier, while thi
the use of Win
the storage re
of messages
provides an e
re is always r
ed in this solu
configure the t
custom seria
onal needs (fo
blobs metad
nd quickly fin
llector compo
message store
ction layer im
ects a simplis
mentals of the
ent compress
jump to a con
indows Azure
the Windows
itional technic
enhance the m
ch as:
mpression to
mizable over
ws you to eas
ansient cond
s solution can
ndows Azure
equirements o
and message
easy to use m
room for impr
ution and whic
type of large
alizers in case
or instance, y
ata acting as
d out if you h
onent that wil
e (in case que
mplemented h
tic view on th
e storage abst
sion, automat
nclusion.
e Queues can
s Azure Cach
cal restriction
messaging ex
o reduce stora
rflow of large
sily store any
itions for imp
n use both dis
Caching Ser
of your projec
e size before
means of supp
ovement. Som
ch you may w
message ove
e the default o
you dont need
a breadcrum
ave any orph
ll ensure time
eues are also
ere).
he lifecycle of
traction layer
tic removal of
n be extended
ing Service a
s on the clien
xperience for
age costs and
e messages to
object type.
proved reliabi
stributed cach
vice incurs ad
ct and perform
deciding to e
porting large
me examples
wish to add are
erflow store in
one does not
d the default
mb allowing yo
haned large m
ely removal of
accessed by
a large mess
such as large
f both messag
d to support m
nd Windows A
nt. In fact, I ha
the client by
d save bandw
o Cache or B
lity.
he and blob s
dditional costs
m a cost analy
enable overflo
messages on
s of value-add
e:
n the applicati
meet your pe
compression
ou to scan thr
message blobs
f any orphane
y components

9

sage. It is
e message
ge parts. I
messages
Azure Blob
ave shown tha
providing
width into/out o
lob storage.
store for
s. You should
ysis based on
ow using
n Windows
d features tha
ion
erformance
).
ough your
s (zombies).
ed blobs from
other than
96
at
of
d
n
t

97
The accompanying sample code is available for download from the MSDN Code Gallery. Note
that all source code files are governed by the Microsoft Public License as explained in the
corresponding legal notices.
Additional Resources/References
For more information on the topic discussed in this article, please refer to the following:
Data Storage Offerings on the Windows Azure Platform article in the TechNet Wiki Library.
Windows Azure Storage Architecture Overview post on the Windows Azure Storage Team
blog.
Best Practices for Maximizing Scalability and Cost Effectiveness of Queue-Based Messaging
Solutions on Windows Azure topic in the MSDN Library.


98
Best Practices for Maximizing Scalability and Cost
Effectiveness of Queue-Based Messaging
Solutions on Windows Azure
Authored by: Valery Mizonov
Reviewed by: Brad Calder, Sidney Higa, Christian Martinez, Steve Marx, Curt Peterson, Paolo
Salvatori, and Trace Young
This article offers prescriptive guidance and best practices for building scalable, highly efficient
and cost effective queue-based messaging solutions on the Windows Azure platform. The
intended audience for this article includes solution architects and developers designing and
implementing cloud-based solutions which leverage the Windows Azure platforms queue storage
services.
Abstract
A traditional queue-based messaging solution utilizes the concept of a message storage location
known as a message queue, which is a repository for data that will be sent to or received from
one or more participants, typically via an asynchronous communication mechanism.
The queue-based data exchange represents the foundation of a reliable and highly scalable
messaging architecture capable of supporting a range of powerful scenarios in the distributed
computing environment. Whether its high-volume work dispatch or durable messaging, a
message queuing technology can step in and provide first-class capabilities to address the
different requirements for asynchronous communication at scale.
The purpose of this article is to examine how developers can take advantage of particular design
patterns in conjunction with capabilities provided by the Windows Azure platform to build
optimized and cost-effective queue-based messaging solutions. The article takes a deeper look at
most commonly used approaches to implementing queue-based interactions in Windows Azure
solutions, and provides recommendations for improving performance, increasing scalability and
reducing operating expense.
The underlying discussion is mixed with relevant best practices, hints and recommendations
where appropriate. The scenario described in this article highlights a technical implementation
that is based upon a real-world customer project.
Customer Scenario
For the sake of a concrete example, we will generalize a real-world customer scenario as follows.
A SaaS solution provider launches a new billing system implemented as a Windows Azure
application servicing the business needs for customer transaction processing at scale. The key
premise of the solution is centered upon the ability to offload compute-intensive workload to the
cloud and leverage the elasticity of the Windows Azure infrastructure to perform the
computationally intensive work.
The on-premises element of the end-to-end architecture consolidates and dispatches large
volumes of transactions to a Windows Azure hosted service regularly throughout the day.
Volumes vary from a few thousands to hundreds of thousands transactions per submission,
reach
SLA-d
The s
comp
storag
instan
Windo
Workl
items
instan
optim
The p
the Pr
mining
The s

The d
worklo
typica
each
funda
hing millions o
driven require
solution archit
rised of a mu
ge for work di
nce, decompo
ows Azure qu
load processi
from queues
nces employ m
al performanc
processed wo
rocess Contro
g, reporting a
solution archit
diagram above
oads. The qu
al for many ot
other via que
amental comp
of transactions
ement for a gu
tecture is foun
lti-instance w
ispatch. Trans
osed (de-batc
ueues for the
ing is handled
s and passing
multi-threade
ce.
rk items are r
oller worker ro
and analysis.
tecture can be
e depicts a ty
eue-based m
her Windows
eues. This ena
ponents involv
s per day. Ad
uaranteed ma
nded on the d
worker role-ba
saction batch
ched) into sma
purposes of l
d by multiple i
them through
d queue listen
routed into a d
ole instance,
e depicted as
ypical architec
message exch
Azure applic
ables taking a
ved in a queu
ditionally, ass
aximum proce
distributed ma
ased cloud tie
hes are receiv
aller work item
oad distributi
instances of t
h computation
ners to implem
dedicated que
aggregated a
follows:
cture for scalin
ange pattern
cations and se
a canonical ap
e-based mes
sume that the
essing latency
ap-reduce des
r using the W
ved by Proces
ms and enque
on.
the processin
nal procedure
ment parallel
eue from whic
and persisted
ng out large o
adopted by t
ervices which
pproach to ex
ssage exchan
e solution mus
y.
sign pattern a
Windows Azure
ss Initiator wo
eued into a co
g worker role
es. The proce
data process
ch these are
into a data s
or complex co
his architectu
need to com
xamining spec
ge.

9
st satisfy a
and is
e queue
orker role
ollection of
e fetching wor
essing
sing for
dequeued by
tore for data

ompute
ure is also ver
mmunicate wit
cific
99
rk
y
ry
h

100
Queue-Based Messaging Fundamentals
A typical messaging solution that exchanges data between its distributed components using
message queues includes publishers depositing messages into queues and one or more
subscribers intended to receive these messages. In most cases, the subscribers, sometimes
referred to as queue listeners, are implemented as single- or multi-threaded processes, either
continuously running or initiated on demand as per a scheduling pattern.
At a higher level, there are two primary dispatch mechanisms used to enable a queue listener to
receive messages stored on a queue:
Polling (pull-based model): A listener monitors a queue by checking the queue at regular
intervals for new messages. When the queue is empty, the listener continues polling the
queue, periodically backing off by entering a sleep state.
Triggering (push-based model): A listener subscribes to an event that is triggered (either
by the publisher itself or by a queue service manager) whenever a message arrives on a
queue. The listener in turn can initiate message processing thus not having to poll the queue
in order to determine whether or not any new work is available.
It is also worth mentioning that there are different flavors of both mechanisms. For instance,
polling can be blocking and non-blocking. Blocking keeps a request on hold until a new message
appears on a queue (or timeout is encountered) whereas a non-blocking request completes
immediately if there is nothing on a queue. With a triggering model, a notification can be pushed
to the queue listeners either for every new message, only when the very first message arrives to
an empty queue or when queue depth reaches a certain level.
The dequeue operations supported by Windows Azure Queue Service API are non-
blocking. This means that the API methods such as GetMessage or GetMessages will
return immediately if there is no message found on a queue. By contrast, the Windows
Azure Service Bus queues offer blocking receive operations which block the calling
thread until a message arrives on a queue or a specified timeout period has elapsed.
The most common approach to implementing queue listeners in Windows Azure solutions today
can be summarized as follows:
1. A listener is implemented as an application component that is instantiated and executed as
part of a worker role instance.
2. The lifecycle of the queue listener component would often be bound to the run time of the
hosting role instance.
3. The main processing logic is comprised of a loop in which messages are dequeued and
dispatched for processing.
4. Should no messages be received, the listening thread enters a sleep state the duration of
which is often driven by an application-specific back-off algorithm.
5. The receive loop is being executed and a queue is being polled until the listener is notified to
exit the loop and terminate.
The following flowchart diagram depicts the logic commonly used when implementing a queue
listener with a polling mechanism in Windows Azure applications:
Note

F
th
The u
using
transa
N
or purposes o
he use of a ce
use of a class
Windows Az
actions in term
Note
of this article,
entral queue m
ic queue liste
ure queues b
ms of applicat
more comple
manager (bro
ener with a po
because the W
tion requests
ex design pat
oker) are not u
olling mechan
Windows Azur
performed ag
tterns, for exa
used.
ism may not
re pricing mo
gainst the que
ample those th
be the optima
del measures
eue, regardle

10

hat require
al choice whe
s storage
ess of if the
01
en

102
queue is empty or not. The purpose of the next sections is to discuss some techniques for
maximizing performance and minimizing the cost of queue-based messaging solutions on the
Windows Azure platform.
Best Practices for Performance, Scalability & Cost Optimization
In this section we must examine how to improve the relevant design aspects to achieve higher
performance, better scalability and cost efficiency.
Perhaps, the easiest way of qualifying an implementation pattern as a more efficient solution
would be through the design which meets the following goals:
Reduces operational expenditures by removing a significant portion of storage transactions
that dont derive any usable work.
Eliminates excessive latency imposed by a polling interval when checking a queue for new
messages.
Scales up and down dynamically by adapting processing power to volatile volumes of
work.
The implementation pattern should also meet these goals without introducing a level of
complexity that effectively outweighs the associated benefits.
Best Practices for Optimizing Storage Transaction Costs
When evaluating the total cost of ownership (TCO) and return on investment (ROI) for a solution
deployed on the Windows Azure platform, the volume of storage transactions is one of the main
variables in the TCO equation. Reducing the number of transactions against Windows Azure
queues decreases the operating costs as it relates to running solutions on Windows Azure.
In the context of a queue-based messaging solution, the volume of storage transactions can be
reduced using a combination of the following methods:
1. When putting messages in a queue, group related messages into a single larger batch,
compress and store the compressed image in a blob storage and use the queue to keep a
reference to the blob holding the actual data.
2. When retrieving messages from a queue, batch multiple messages together in a single
storage transaction. The GetMessages method in the Queue Service API enables de-
queuing the specified number of messages in a single transaction (see the note below).
3. When checking the presence of work items on a queue, avoid aggressive polling intervals
and implement a back-off delay that increases the time between polling requests if a queue
remains continuously empty.
4. Reduce the number of queue listeners when using a pull-based model, use only 1 queue
listener per role instance when a queue is empty. To further reduce the number of queue
listeners per role instance to zero, use a notification mechanism to instantiate queue listeners
when the queue receives work items.
5. If queues remain empty for most of the time, automatically scale down the number of role
instances and continue to monitor relevant system metrics to determine if and when the
application should scale up the number of instances to handle increasing workload.

103
Most of the above recommendations can be translated into a fairly generic implementation that
handles message batches and encapsulates many of the underlying queue/blob storage and
thread management operations. Later in this article, we will examine how to do this.
When retrieving messages via the GetMessages method, the maximum batch size
supported by Queue Service API in a single dequeue operation is limited to 32.
Generally speaking, the cost of Windows Azure queue transactions increases linearly as the
number of queue service clients increases, such as when scaling up the number of role instances
or increasing the number of dequeue threads. To illustrate the potential cost impact of a solution
design that does not take advantage of the above recommendations; we will provide an example
backed up by concrete numbers.
The Cost Impact of Inefficient Design
If the solution architect does not implement relevant optimizations, the billing system architecture
described above will likely incur excessive operating expenses once the solution is deployed and
running on the Windows Azure platform. The reasons for the possible excessive expense are
described in this section.
As noted in the scenario definition, the business transaction data arrives at regular intervals.
However, lets assume that the solution is busy processing workload just 25% of the time during a
standard 8-hour business day. That results in 6 hours (8 hours * 75%) of idle time when there
may not be any transactions coming through the system. Furthermore, the solution will not
receive any data at all during the 16 non-business hours every day.
During the idle period totaling 22 hours, the solution is still performing attempts to dequeue work
as it has no explicit knowledge when new data arrives. During this time window, each individual
dequeue thread will perform up to 79,200 transactions (22 hours * 60 min * 60 transactions/min)
against an input queue, assumed a default polling interval of 1 second.
As previously mentioned, the pricing model in the Windows Azure platform is based upon
individual storage transactions. A storage transaction is a request made by a user application to
add, read, update or delete storage data. As of the writing of this whitepaper, storage transactions
are billed at a rate of $0.01 for 10,000 transactions (not taking into account any promotional
offerings or special pricing arrangements).
When calculating the number of queue transactions, keep in mind that putting a single
message on a queue would be counted as 1 transaction, whereas consuming a message
is often a 2-step process involving the retrieval followed by a request to remove the
message from the queue. As a result, a successful dequeue operation will attract 2
storage transactions. Please note that even if a dequeue request results in no data being
retrieved; it still counts as a billable transaction.
The storage transactions generated by a single dequeue thread in the above scenario will add
approximately $2.38 (79,200 / 10,000 * $0.01 * 30 days) to a monthly bill. In comparison, 200
dequeue threads (or, alternatively, 1 dequeue thread in 200 worker role instances) will increase
the cost by $457.20 per month. That is the cost incurred when the solution was not performing
Important
Important

104
any computations at all, just checking on the queues to see if any work items are available. The
above example is abstract as no one would implement their service this way, which is why it is
important to do the optimizations described next.
Best Practices for Eliminating Excessive Latency
To optimize performance of queue-based Windows Azure messaging solutions one approach is
to use the publish/subscribe messaging layer provided with the Windows Azure Service Bus, as
described in this section.
In this approach, developers will need to focus on creating a combination of polling and real-time
push-based notifications, enabling the listeners to subscribe to a notification event (trigger) that is
raised upon certain conditions to indicate that a new workload is put on a queue. This approach
enhances the traditional queue polling loop with a publish/subscribe messaging layer for
dispatching notifications.
In a complex distributed system, this approach would necessitate the use of a message bus or
message-oriented middleware to ensure that notifications can be reliably relayed to one or more
subscribers in a loosely coupled fashion. Windows Azure Service Bus is a natural choice for
addressing messaging requirements between loosely coupled distributed application services
running on Windows Azure and running on-premises. It is also a perfect fit for a message bus
architecture that will enable exchanging notifications between processes involved in queue-based
communication.
The processes engaged in a queue-based message exchange could employ the following
pattern:

Speci
subsc
instan
excha
Comm
T
ac
su
It
in
ev
S
ad
F
se
While
furthe
next s
Im
ifically, and as
cribers, the sa
nces would m
ange. We hav
munication Us
he use of the
ccount the vo
uch as a queu
is therefore i
ntroducing the
valuating whe
ervice Bus w
dditional deve
or more inform
ections in Win
e the impact o
er cost reducti
section.
mportant
s it relates to
ame principles
meet the major
ve already cov
sing Windows
Windows Az
olume of mess
ue or a topic.
mportant to p
e Service Bus
ether or not th
ould, in fact, l
elopment effo
mation on the
ndows Azure
on latency is fa
ion could be r
the interactio
s that apply to
rity of require
vered these fu
s Azure Servi
zure Service B
saging operat
perform a cos
s into a given
he introductio
lead to cost re
orts.
e pricing mod
Platform FAQ
airly easy to a
realized by us
on between qu
o the commun
ments for pus
undamentals
ce Bus.
Bus is subject
tions against
t-benefit anal
architecture.
n of the notifi
eduction that
el for Service
Qs.
address with
sing dynamic
ueue service
nication betw
sh-based noti
in How to Sim
t to a pricing m
a Service Bu
ysis to asses
Along those
cation dispatc
can justify th
e Bus, please
a publish/sub
(elastic) scal
publishers an
ween Windows
ification mess
mplify & Scale
model that ta
s messaging
ss the pros an
lines, it is wor
ch layer base
e investment
refer to the re
bscribe messa
ing, as descr

10

nd
s Azure role
sage
e Inter-Role
kes into
entity
nd cons of
rth
ed on the
ts and
elevant
aging layer, a
ribed in the
05
a

106
Best Practices for Dynamic Scaling
The Windows Azure platform makes it possible for customers to scale up and down faster and
easier than ever before. The ability to adapt to volatile workloads and variable traffic is one of the
primary value propositions of the cloud platform. This means that scalability is no longer an
expensive IT vocabulary term, it is now an out-of-the-box feature that can be programmatically
enabled on demand in a well-architected cloud solution.
Dynamic scaling is the technical capability of a given solution to adapt to fluctuating workloads by
increasing and reducing working capacity and processing power at runtime. The Windows Azure
platform natively supports dynamic scaling through the provisioning of a distributed computing
infrastructure on which compute hours can be purchased as needed.
It is important to differentiate between the following 2 types of dynamic scaling on the Windows
Azure platform:
Role instance scaling refers to adding and removing additional web or worker role instances
to handle the point-in-time workload. This often includes changing the instance count in the
service configuration. Increasing the instance count will cause Windows Azure runtime to
start new instances whereas decreasing the instance count will in turn cause it to shut down
running instances.
Process (thread) scaling refers to maintaining sufficient capacity in terms of processing
threads in a given role instance by tuning the number of threads up and down depending on
the current workload.
Dynamic scaling in a queue-based messaging solution would attract a combination of the
following general recommendations:
1. Monitor key performance indicators including CPU utilization, queue depth, response
times and message processing latency.
2. Dynamically increase or decrease the number of role instances to cope with the spikes
in workload, either predictable or unpredictable.
3. Programmatically expand and trim down the number of processing threads to adapt to
variable load conditions handled by a given role instance.
4. Partition and process fine-grained workloads concurrently using the Task Parallel
Library in the .NET Framework 4.
5. Maintain a viable capacity in solutions with highly volatile workload in anticipation of sudden
spikes to be able to handle them without the overhead of setting up additional instances.
The Service Management API makes it possible for a Windows Azure hosted service to modify
the number of its running role instances by changing deployment configuration at runtime.
The maximum number of Windows Azure small compute instances (or the equivalent
number of other sized compute instances in terms of number of cores) in a typical
subscription is limited to 20 by default. Any requests for increasing this quota should be
raised with the Windows Azure Support team. For more information, see the Windows
Azure Platform FAQs
Dynamic scaling of the role instance count may not always be the most appropriate choice for
handling load spikes. For instance, a new role instance can take a few seconds to spin up and
there are currently no SLA metrics provided with respect to spin-up duration. Instead, a solution
Note
may n
increa
and d
proce
A
tr
pe
th
W
ba
re
sc
In mo
instan
archit
scalin
that g
or dec

It is w
on Wi
servic
Im
need to simply
ase. While wo
determine whe
esses.
At present, the
ansactions/se
erforming que
hreads, it may
When this occu
ack-off delay
ecommended
cale across m
ost cases, aut
nce. By contra
tecture that is
ng actions. Th
gathers and an
commission id
worth noting th
indows Azure
ce will be able
mportant
y increase the
orkload is bein
ether it needs
e scalability ta
ec. If an appli
eue operation
y result in HTT
urs, the applic
algorithm. Ho
to use multip
multiple queue
o-scaling the
ast, role insta
responsible f
he diagram be
nalyzes load
dle instances
hat the scaling
e or as an on-
e to access th
e number of w
ng processed
s to dynamica
arget for a sin
cation attemp
ns from multip
TP 503 Serv
cation should
owever, if the
ple queues an
es.
worker proce
nce scaling o
for monitoring
elow depicts a
metrics to de
.
g agent servic
-premises ser
he Windows A
worker thread
d, the solution
ally reduce or
gle Windows
pts to exceed
ple role instan
ver Busy resp
d implement a
e HTTP 503 e
nd implement
esses is the re
often involves
g performanc
a service com
termine whet
ce can be dep
rvice. Irrespec
Azure queues
ds to deal with
n will monitor t
increase the
Azure queue
this target, fo
nce running h
ponse from th
a retry mecha
rrors are occu
t a partitioning
esponsibility o
a central ele
e metrics and
mponent called
ther it needs t
ployed either
ctively of the d
.
h the volatile w
the relevant lo
number of wo
e is constrain
or example, th
undreds of de
he storage ser
nism using ex
urring regular
g-based strate
of an individu
ment of the s
d taking the a
d Dynamic Sc
to provision n
as a worker r
deployment to

10
workload
oad metrics
orker
ned at 500
hrough
equeue
rvice.
xponential
rly, it is
egy to
al role
solution
ppropriate
caling Agent
ew instances

role running
opology, the
07
s

108
To implement a dynamic scaling capability, consider the use of the Microsoft Enterprise Library
Autoscaling Application Block that enables automatic scaling behavior in the solutions running on
Windows Azure. The Autoscaling Application Block provides all of the functionality needed to
define and monitor autoscaling in a Windows Azure application.
Now that we have covered the latency impact, storage transaction costs and dynamic scale
requirements, it is a good time to consolidate our recommendations into a technical
implementation.
Technical Implementation
In the previous sections, we have examined the key characteristics attributed to a well-designed
messaging architecture based on the Windows Azure queue storage queues. We have looked at
three main focus areas that help reduce processing latency, optimize storage transaction costs
and improve responsiveness to fluctuating workloads.
This section is intended to provide a starting point to assist Windows Azure developers with
implementing some of the patterns referenced in this whitepaper from a programming
perspective.
This section will focus on building an auto-scalable queue listener that supports both pull-
based and push-based models. For advanced techniques in dynamic scaling at the role
instance level, please refer to the Enterprise Library Autoscaling Application Block.
In addition, for the sake of brevity, we will only focus on some core functional elements
and avoid undesired complexity by omitting much of the supporting infrastructure code
from the code samples below. For the purposes of clarification, its also worth pointing out
that the technical implementation discussed below is not the only solution to a given
problem space. It is intended to serve the purpose of a starting point upon which
developers may derive their own more elegant solutions.
From this point onwards, this whitepaper will focus on the source code required to implement the
patterns discussed above.
Building Generic Queue Listener
First, we define a contract that will be implemented by a queue listener component that is hosted
by a worker role and listens on a Windows Azure queue.
/// Defines a contract that must be implemented by an extension responsible for listening
on a Windows Azure queue.
public interface ICloudQueueServiceWorkerRoleExtension
{
/// Starts a multi-threaded queue listener that uses the specified number of dequeue
threads.
void StartListener(int threadCount);

Note

109
/// Returns the current state of the queue listener to determine point-in-time load
characteristics.
CloudQueueListenerInfo QueryState();

/// Gets or sets the batch size when performing dequeue operation against a Windows
Azure queue.
int DequeueBatchSize { get; set; }

/// Gets or sets the default interval that defines how long a queue listener will be
idle for between polling a queue.
TimeSpan DequeueInterval { get; set; }

/// Defines a callback delegate which will be invoked whenever the queue is empty.
event WorkCompletedDelegate QueueEmpty;
}
The QueueEmpty event is intended to be used by a host. It provides the mechanism for the host
to control the behavior of the queue listener when the queue is empty. The respective event
delegate is defined as follows:
/// <summary>
/// Defines a callback delegate which will be invoked whenever an unit of work has been
completed and the worker is
/// requesting further instructions as to next steps.
/// </summary>
/// <param name="sender">The source of the event.</param>
/// <param name="idleCount">The value indicating how many times the worker has been
idle.</param>
/// <param name="delay">Time interval during which the worker is instructed to sleep
before performing next unit of work.</param>
/// <returns>A flag indicating that the worker should stop processing any further units
of work and must terminate.</returns>
public delegate bool WorkCompletedDelegate(object sender, int idleCount, out TimeSpan
delay);
Handling queue items is easier if a listener can operate with generics as opposed to using bare
metal SDK classes such as CloudQueueMessage. Therefore, we define a new interface that will
be implemented by a queue listener capable of supporting generics-based access to queues:
/// <summary>
/// Defines a contract that must be supported by an extension that implements a generics-
aware queue listener.
/// </summary>
/// <typeparam name="T">The type of queue item data that will be handled by the queue
listener.</typeparam>

110
public interface ICloudQueueListenerExtension<T> : ICloudQueueServiceWorkerRoleExtension,
IObservable<T>
{
}
Note that we also enabled the generics-aware listener to push queue items to one or more
subscribers through the implementation of the Observer design pattern by leveraging the
IObservable<T> interface available in the .NET Framework 4.
We intend to keep a single instance of a component implementing the
ICloudQueueListenerExtension<T> interface. However, we need to be able to run multiple
dequeue threads (worker processes, or tasks for simplicity). Therefore, we add support for multi-
threaded dequeue logic in the queue listener component. This is where we take advantage of the
Task Parallel Library (TPL). The StartListener method will be responsible for spinning up the
specified number of dequeue threads as follows:

/// <summary>
/// Starts the specified number of dequeue tasks.
/// </summary>
/// <param name="threadCount">The number of dequeue tasks.</param>
public void StartListener(int threadCount)
{
Guard.ArgumentNotZeroOrNegativeValue(threadCount, "threadCount");

// The collection of dequeue tasks needs to be reset on each call to this method.
if (this.dequeueTasks.IsAddingCompleted)
{
this.dequeueTasks = new BlockingCollection<Task>(this.dequeueTaskList);
}

for (int i = 0; i < threadCount; i++)
{
CancellationToken cancellationToken = this.cancellationSignal.Token;
CloudQueueListenerDequeueTaskState<T> workerState = new
CloudQueueListenerDequeueTaskState<T>(Subscriptions, cancellationToken,
this.queueLocation, this.queueStorage);

// Start a new dequeue task and register it in the collection of tasks internally
managed by this component.
this.dequeueTasks.Add(Task.Factory.StartNew(DequeueTaskMain, workerState,
cancellationToken, TaskCreationOptions.LongRunning, TaskScheduler.Default));
}

// Mark this collection as not accepting any more additions.

111
this.dequeueTasks.CompleteAdding();
}
The DequeueTaskMain method implements the functional body of a dequeue thread. Its main
operations are the following:
/// <summary>
/// Implements a task performing dequeue operations against a given Windows Azure queue.
/// </summary>
/// <param name="state">An object containing data to be used by the task.</param>
private void DequeueTaskMain(object state)
{
CloudQueueListenerDequeueTaskState<T> workerState =
(CloudQueueListenerDequeueTaskState<T>)state;

int idleStateCount = 0;
TimeSpan sleepInterval = DequeueInterval;

try
{
// Run a dequeue task until asked to terminate or until a break condition is
encountered.
while (workerState.CanRun)
{
try
{
var queueMessages = from msg in
workerState.QueueStorage.Get<T>(workerState.QueueLocation.QueueName, DequeueBatchSize,
workerState.QueueLocation.VisibilityTimeout).AsParallel() where msg != null select msg;
int messageCount = 0;

// Process the dequeued messages concurrently by taking advantage of the
above PLINQ query.
queueMessages.ForAll((message) =>
{
// Reset the count of idle iterations.
idleStateCount = 0;

// Notify all subscribers that a new message requires processing.
workerState.OnNext(message);

// Once successful, remove the processed message from the queue.
workerState.QueueStorage.Delete<T>(message);


112
// Increment the number of processed messages.
messageCount++;
});

// Check whether or not we have done any work during this iteration.
if (0 == messageCount)
{
// Increment the number of iterations when we were not doing any work
(e.g. no messages were dequeued).
idleStateCount++;

// Call the user-defined delegate informing that no more work is
available.
if (QueueEmpty != null)
{
// Check if the user-defined delegate has requested a halt to any
further work processing.
if (QueueEmpty(this, idleStateCount, out sleepInterval))
{
// Terminate the dequeue loop if user-defined delegate
advised us to do so.
break;
}
}

// Enter the idle state for the defined interval.
Thread.Sleep(sleepInterval);
}
}
catch (Exception ex)
{
if (ex is OperationCanceledException)
{
throw;
}
else
{
// Offload the responsibility for handling or reporting the error to
the external object.
workerState.OnError(ex);

// Sleep for the specified interval to avoid a flood of errors.

113
Thread.Sleep(sleepInterval);
}
}
}
}
finally
{
workerState.OnCompleted();
}
}
A couple of points are worth making with respect to the DequeueTaskMain method
implementation.
First, we are taking advantage of the Parallel LINQ (PLINQ) when dispatching messages for
processing. The main advantage of PLINQ here is to speed up message handling by executing
the query delegate on separate worker threads on multiple processors in parallel whenever
possible.
Since query parallelization is internally managed by PLINQ, there is no guarantee that
PLINQ will utilize more than a single core for work parallelization. PLINQ may run a query
sequentially if it determines that the overhead of parallelization will slow down the query.
In order to benefit from PLINQ, the total work in the query has to be sufficiently large to
benefit from the overhead of scheduling the work on the thread pool.
Second, we are not fetching a single message at a time. Instead, we ask the Queue Service API
to retrieve a specific number of messages from a queue. This is driven by the DequeueBatchSize
parameter that is passed to the Get<T> method. When we enter the storage abstraction layer
implemented as part of the overall solution, this parameter is handed over to the Queue Service
API method. In addition, we run a safety check to ensure that the batch size doesnt exceed the
maximum size supported by the APIs. This is implemented as follows:
/// This class provides reliable generics-aware access to the Windows Azure Queue
storage.
public sealed class ReliableCloudQueueStorage : ICloudQueueStorage
{
/// The maximum batch size supported by Queue Service API in a single Get operation.
private const int MaxDequeueMessageCount = 32;

/// Gets a collection of messages from the specified queue and applies the specified
visibility timeout.
public IEnumerable<T> Get<T>(string queueName, int count, TimeSpan visibilityTimeout)
{
Guard.ArgumentNotNullOrEmptyString(queueName, "queueName");
Guard.ArgumentNotZeroOrNegativeValue(count, "count");

Note

114
try
{
var queue =
this.queueStorage.GetQueueReference(CloudUtility.GetSafeContainerName(queueName));

IEnumerable<CloudQueueMessage> queueMessages =
this.retryPolicy.ExecuteAction<IEnumerable<CloudQueueMessage>>(() =>
{
return queue.GetMessages(Math.Min(count, MaxDequeueMessageCount),
visibilityTimeout);
});

// ... There is more code after this point ...
And finally, we are not going to run the dequeue task indefinitely. We provisioned an explicit
checkpoint implemented as a QueueEmpty event which is raised whenever a queue becomes
empty. At that point, we consult to a QueueEmpty event handler to determine whether or not it
permits us to finish the running dequeue task. A well-designed implementation of the
QueueEmpty event handler allows supporting the auto scale-down capability as explained in the
following section.
Auto Scaling Down Dequeue Tasks
The purpose of QueueEmpty event handler is a two-fold. First, it is responsible for providing
feedback to the source dequeue task instructing it to enter a sleep state for a given time interval
(as defined in the delay output parameter in the event delegate). Secondly, it indicates to the
dequeue task whether or not it must gracefully shut itself down (as prescribed by the Boolean
return parameter).
The following implementation of the QueueEmpty event handler solves the two challenges
highlighted earlier in this whitepaper. It calculates a random exponential back-off interval and tells
the dequeue task to exponentially increase the delay between queue polling requests. Note that
the back-off delay will not exceed 1 second as configured in our solution, since it really isnt
necessary to have a long delay between polling when auto-scaling is implemented well enough.
In addition, it queries the queue listener state to determine the number of active dequeue tasks.
Should this number be in excess of 1, the event handler advise the originating dequeue task to
complete its polling loop provided the back-off interval has also reached its specified maximum.
Otherwise, the dequeue task will not be terminated, leaving exactly 1 polling thread running at a
time per a single instance of the queue listener. This approach helps reduce the number of
storage transactions and therefore decrease the transaction costs as explained earlier.
private bool HandleQueueEmptyEvent(object sender, int idleCount, out TimeSpan delay)
{
// The sender is an instance of the ICloudQueueServiceWorkerRoleExtension, we can
safely perform type casting.

115
ICloudQueueServiceWorkerRoleExtension queueService = sender as
ICloudQueueServiceWorkerRoleExtension;

// Find out which extension is responsible for retrieving the worker role
configuration settings.
IWorkItemProcessorConfigurationExtension config =
Extensions.Find<IWorkItemProcessorConfigurationExtension>();

// Get the current state of the queue listener to determine point-in-time load
characteristics.
CloudQueueListenerInfo queueServiceState = queueService.QueryState();

// Set up the initial parameters, read configuration settings.
int deltaBackoffMs = 100;
int minimumIdleIntervalMs =
Convert.ToInt32(config.Settings.MinimumIdleInterval.TotalMilliseconds);
int maximumIdleIntervalMs =
Convert.ToInt32(config.Settings.MaximumIdleInterval.TotalMilliseconds);

// Calculate a new sleep interval value that will follow a random exponential back-
off curve.
int delta = (int)((Math.Pow(2.0, (double)idleCount) - 1.0) * (new
Random()).Next((int)(deltaBackoffMs * 0.8), (int)(deltaBackoffMs * 1.2)));
int interval = Math.Min(minimumIdleIntervalMs + delta, maximumIdleIntervalMs);

// Pass the calculated interval to the dequeue task to enable it to enter into a
sleep state for the specified duration.
delay = TimeSpan.FromMilliseconds((double)interval);

// As soon as interval reaches its maximum, tell the source dequeue task that it must
gracefully terminate itself
// unless this is a last deqeueue task. If so, we are not going to keep it running
and continue polling the queue.
return delay.TotalMilliseconds >= maximumIdleIntervalMs &&
queueServiceState.ActiveDequeueTasks > 1;
}
At a higher level, the above described "dequeue task scale-down" capability can be explained as
follows:
1. Whenever there is anything in the queue, the dequeue tasks will ensure that the workload will
be processed as soon as possible. There will be no delay between requests to dequeue
messages from a queue.

116
2. As soon as the source queue becomes empty, each dequeue task will raise a QueueEmpty
event.
3. The QueueEmpty event handler will calculate a random exponential back-off delay and
instruct the dequeue task to suspend its activity for a given interval.
4. The dequeue tasks will continue polling the source queue at computed intervals until the idle
duration exceeds its allowed maximum.
5. Upon reaching the maximum idle interval, and provided that the source queue is still empty,
all active dequeue tasks will start gracefully shutting themselves down this will not occur all
at once, since the dequeue tasks are backing off at different points in the back off algorithm.
6. At some point in time, there will be only one active dequeue task waiting for work. As the
result, no idle polling transactions will occur against a queue except only by that single task.
To elaborate on the process of collecting point-in-time load characteristics, it is worth mentioning
the relevant source code artifacts. First, there is a structure holding the relevant metrics that
measure the result of the load that is being applied to the solution. For the purposes of simplicity,
we included a small subset of metrics that will be used further in the sample code.
/// Implements a structure containing point-in-time load characteristics for a given
queue listener.
public struct CloudQueueListenerInfo
{
/// Returns the approximate number of items in the Windows Azure queue.
public int CurrentQueueDepth { get; internal set; }

/// Returns the number of dequeue tasks that are actively performing work or waiting
for work.
public int ActiveDequeueTasks { get; internal set; }

/// Returns the maximum number of dequeue tasks that were active at a time.
public int TotalDequeueTasks { get; internal set; }
}
Secondly, there is a method implemented by a queue listener which returns its load metrics as
depicted in the following example:
/// Returns the current state of the queue listener to determine point-in-time load
characteristics.
public CloudQueueListenerInfo QueryState()
{
return new CloudQueueListenerInfo()
{
CurrentQueueDepth = this.queueStorage.GetCount(this.queueLocation.QueueName),
ActiveDequeueTasks = (from task in this.dequeueTasks where task.Status !=
TaskStatus.Canceled && task.Status != TaskStatus.Faulted && task.Status !=
TaskStatus.RanToCompletion select task).Count(),
TotalDequeueTasks = this.dequeueTasks.Count

117
};
}
Auto Scaling Up Dequeue Tasks
In the previous section, we introduced the ability to reduce the number of active dequeue tasks to
a single instance in order to minimize the impact of idle transactions on the storage operation
costs. In this section, we are going to walk through a contrast example whereby we implement
the auto scale-up capability to bring the processing power back when its needed.
First, we define an event delegate that will help track state transitions from an empty to a non-
empty queue for the purposes of triggering relevant actions:
/// <summary>
/// Defines a callback delegate which will be invoked whenever new work arrived to a
queue while the queue listener was idle.
/// </summary>
/// <param name="sender">The source of the event.</param>
public delegate void WorkDetectedDelegate(object sender);
We then extend the original definition of the ICloudQueueServiceWorkerRoleExtension interface
to include a new event that will be raised every time a queue listener detects new work items,
essentially when queue depth changes from zero to any positive value:
public interface ICloudQueueServiceWorkerRoleExtension
{
// ... The other interface members were omitted for brevity. See the previous code
snippets for reference ...

// Defines a callback delegate to be invoked whenever a new work has arrived to a
queue while the queue listener was idle.
event WorkDetectedDelegate QueueWorkDetected;
}
Also, we determine the right place in the queue listeners code where such an event will be
raised. We are going to fire the QueueWorkDetected event from within the dequeue loop
implemented in the DequeueTaskMain method which needs to be extended as follows:
public class CloudQueueListenerExtension<T> : ICloudQueueListenerExtension<T>
{
// An instance of the delegate to be invoked whenever a new work has arrived to a
queue while the queue listener was idle.
public event WorkDetectedDelegate QueueWorkDetected;

private void DequeueTaskMain(object state)
{
CloudQueueListenerDequeueTaskState<T> workerState =
(CloudQueueListenerDequeueTaskState<T>)state;

118

int idleStateCount = 0;
TimeSpan sleepInterval = DequeueInterval;

try
{
// Run a dequeue task until asked to terminate or until a break condition is
encountered.
while (workerState.CanRun)
{
try
{
var queueMessages = from msg in
workerState.QueueStorage.Get<T>(workerState.QueueLocation.QueueName, DequeueBatchSize,
workerState.QueueLocation.VisibilityTimeout).AsParallel() where msg != null select msg;
int messageCount = 0;

// Check whether or not work items arrived to a queue while the
listener was idle.
if (idleStateCount > 0 && queueMessages.Count() > 0)
{
if (QueueWorkDetected != null)
{
QueueWorkDetected(this);
}
}

// ... The rest of the code was omitted for brevity. See the previous
code snippets for reference ...
In the last step, we provide a handler for the QueueWorkDetected event. The implementation of
this event handler will be supplied by a component which instantiates and hosts the queue
listener. In our case, its a worker role. The code responsible for instantiation and implementation
of event handler is comprised of the following:
public class WorkItemProcessorWorkerRole : RoleEntryPoint
{
// Called by Windows Azure to initialize the role instance.
public override sealed bool OnStart()
{
// ... There is some code before this point ...

// Instantiate a queue listener for the input queue.

119
var inputQueueListener = new
CloudQueueListenerExtension<XDocument>(inputQueueLocation);

// Configure the input queue listener.
inputQueueListener.QueueEmpty += HandleQueueEmptyEvent;
inputQueueListener.QueueWorkDetected += HandleQueueWorkDetectedEvent;
inputQueueListener.DequeueBatchSize =
configSettingsExtension.Settings.DequeueBatchSize;
inputQueueListener.DequeueInterval =
configSettingsExtension.Settings.MinimumIdleInterval;

// ... There is more code after this point ...
}

// Implements a callback delegate to be invoked whenever a new work has arrived to a
queue while the queue listener was idle.
private void HandleQueueWorkDetectedEvent(object sender)
{
// The sender is an instance of the ICloudQueueServiceWorkerRoleExtension, we can
safely perform type casting.
ICloudQueueServiceWorkerRoleExtension queueService = sender as
ICloudQueueServiceWorkerRoleExtension;

// Get the current state of the queue listener to determine point-in-time load
characteristics.
CloudQueueListenerInfo queueServiceState = queueService.QueryState();

// Determine the number of queue tasks that would be required to handle the
workload in a queue given its current depth.
int dequeueTaskCount =
GetOptimalDequeueTaskCount(queueServiceState.CurrentQueueDepth);

// If the dequeue task count is less than computed above, start as many dequeue
tasks as needed.
if (queueServiceState.ActiveDequeueTasks < dequeueTaskCount)
{
// Start the required number of dequeue tasks.
queueService.StartListener(dequeueTaskCount -
queueServiceState.ActiveDequeueTasks);
}
} // ... There is more code after this point ...

120
In light of the above example, the GetOptimalDequeueTaskCount method is worth taking a
deeper look at. This method is responsible for computing the number of dequeue tasks that would
be considered optimal for handling the workload in a queue. When invoked, this method should
determine (through any appropriate decision-making mechanisms) how much "horsepower" the
queue listener needs in order to process the volume of work either awaiting on or expected to
arrive to a given queue.
For instance, the developer could take a simplistic approach and embed a set of static rules
directly into the GetOptimalDequeueTaskCount method. Using the known throughput and
scalability characteristics of the queuing infrastructure, average processing latency, payload size
and other relevant inputs, the rule set could take an optimistic view and decide on an optimal
dequeue task count.
In the example below, an intentionally over-simplified technique is being used for determining the
number of dequeue tasks:
/// <summary>
/// Returns the number of queue tasks that would be required to handle the workload in a
queue given its current depth.
/// </summary>
/// <param name="currentDepth">The approximate number of items in the queue.</param>
/// <returns>The optimal number of dequeue tasks.</returns>
private int GetOptimalDequeueTaskCount(int currentDepth)
{
if (currentDepth < 100) return 10;
if (currentDepth >= 100 && currentDepth < 1000) return 50;
if (currentDepth >= 1000) return 100;

// Return the minimum acceptable count.
return 1;
}
It is worth reiterating that the example code above is not intended to be a one size fits all
approach. A more ideal solution would be to invoke an externally configurable and manageable
rule which performs the necessary computations.
At this point, we have a working prototype of a queue listener capable of automatically scaling
itself up and down as per fluctuating workload. Perhaps, as a final touch, it needs to be enriched
with the ability to adapt itself to variable load while its being processed. This capability can be
added by applying the same pattern as it was being followed when adding support for the
QueueWorkDetected event.
Now, lets switch focus to another important optimization that will help reduce latency in the
queue listeners.
Implementing Publish/Subscribe Layer for Zero-Latency Dequeue
In this section, we are going to enhance the above implementation of a queue listener with a
push-based notification mechanism built on top of the Service Bus one-way multicast capability.

121
The notification mechanism is responsible for triggering an event telling the queue listener to start
performing dequeue work. This approach helps avoid polling the queue to check for new
messages and therefore eliminate the associated latency.
First, we define a trigger event that will be received by our queue listener in case a new workload
is deposited into a queue:
/// Implements a trigger event indicating that a new workload was put in a queue.
[DataContract(Namespace = WellKnownNamespace.DataContracts.Infrastructure)]
public class CloudQueueWorkDetectedTriggerEvent
{
/// Returns the name of the storage account on which the queue is located.
[DataMember]
public string StorageAccount { get; private set; }

/// Returns a name of the queue where the payload was put.
[DataMember]
public string QueueName { get; private set; }

/// Returns a size of the queue's payload (e.g. the size of a message or the number
of messages in a batch).
[DataMember]
public long PayloadSize { get; private set; }

// ... The constructor was omitted for brevity ...
}
Next, we enable the queue listener implementations to act as subscribers to receive a trigger
event. The first step is to define a queue listener as an observer for the
CloudQueueWorkDetectedTriggerEvent event:
/// Defines a contract that must be implemented by an extension responsible for listening
on a Windows Azure queue.
public interface ICloudQueueServiceWorkerRoleExtension :
IObserver<CloudQueueWorkDetectedTriggerEvent>
{
// ... The body is omitted as it was supplied in previous examples ...
}
The second step is to implement the OnNext method defined in the IObserver<T> interface. This
method gets called by the provider to notify the observer about a new event:
public class CloudQueueListenerExtension<T> : ICloudQueueListenerExtension<T>
{
// ... There is some code before this point ...

/// <summary>

122
/// Gets called by the provider to notify this queue listener about a new trigger
event.
/// </summary>
/// <param name="e">The trigger event indicating that a new payload was put in a
queue.</param>
public void OnNext(CloudQueueWorkDetectedTriggerEvent e)
{
Guard.ArgumentNotNull(e, "e");

// Make sure the trigger event is for the queue managed by this listener,
otherwise ignore.
if (this.queueLocation.StorageAccount == e.StorageAccount &&
this.queueLocation.QueueName == e.QueueName)
{
if (QueueWorkDetected != null)
{
QueueWorkDetected(this);
}
}
}

// ... There is more code after this point ...
}
At it can be seen in the above example, we purposefully invoke the same event delegate as it is
used in the previous steps. The QueueWorkDetected event handler already provides the
necessary application logic for instantiating optimal number of dequeue tasks. Therefore, the
same event handler is reused when handling the CloudQueueWorkDetectedTriggerEvent
notification.
As noted in the previous sections, we dont have to maintain a continuously running dequeue task
when a push-based notification is employed. Therefore, we can reduce the number of queue
tasks per a queue listener instance to zero and use a notification mechanism to instantiate
dequeue tasks when the queue receives work items. In order to make sure that we are not
running any idle dequeue tasks, the following straightforward modification in the QueueEmpty
event hander is required:
private bool HandleQueueEmptyEvent(object sender, int idleCount, out TimeSpan delay)
{
// ... There is some code before this point ...

// As soon as interval reaches its maximum, tell the source dequeue task that it must
gracefully terminate itself.
return delay.TotalMilliseconds >= maximumIdleIntervalMs;
}

123
In summary, we are no longer detecting whether or not there is a single active dequeue task
remaining. The result of the revised QueueEmpty event hander only takes into account the fact of
exceeding the maximum idle interval upon which all active dequeue tasks will be shut down.
To receive the CloudQueueWorkDetectedTriggerEvent notifications, we leverage the
publish/subscribe model that is implemented as loosely coupled messaging between Windows
Azure role instances. In essence, we hook on the same inter-role communication layer and
handle the incoming events as follows:
public class InterRoleEventSubscriberExtension : IInterRoleEventSubscriberExtension
{
// ... Some code here was omitted for brevity. See the corresponding guidance on
Windows Azure CAT team blog for reference ...

public void OnNext(InterRoleCommunicationEvent e)
{
if (this.owner != null && e.Payload != null)
{
// ... There is some code before this point ...

if (e.Payload is CloudQueueWorkDetectedTriggerEvent)
{
HandleQueueWorkDetectedTriggerEvent(e.Payload as
CloudQueueWorkDetectedTriggerEvent);
return;
}

// ... There is more code after this point ...
}
}

private void HandleQueueWorkDetectedTriggerEvent(CloudQueueWorkDetectedTriggerEvent
e)
{
Guard.ArgumentNotNull(e, "e");

// Enumerate through registered queue listeners and relay the trigger event to
them.
foreach (var queueService in
this.owner.Extensions.FindAll<ICloudQueueServiceWorkerRoleExtension>())
{
// Pass the trigger event to a given queue listener.
queueService.OnNext(e);
}

124
}
}
Multicasting a trigger event defined in the CloudQueueWorkDetectedTriggerEvent class is the
ultimate responsibility of a publisher, namely, the component depositing work items on a queue.
This event can be triggered either before the very first work item is enqueued or after last item is
put in a queue. In the example below, we publish a trigger event upon completing putting work
items into the input queue:
public class ProcessInitiatorWorkerRole : RoleEntryPoint
{
// The instance of the role extension which provides an interface to the inter-role
communication service.
private volatile IInterRoleCommunicationExtension interRoleCommunicator;

// ... Some code here was omitted for brevity. See the corresponding guidance on
Windows Azure CAT team blog for reference ...

private void HandleWorkload()
{
// Step 1: Receive compute-intensive workload.
// ... (code was omitted for brevity) ...

// Step 2: Enqueue work items into the input queue.
// ... (code was omitted for brevity) ...

// Step 3: Notify the respective queue listeners that they should expect work to
arrive.
// Create a trigger event referencing the queue into which we have just put work
items.
var trigger = new CloudQueueWorkDetectedTriggerEvent("MyStorageAccount",
"InputQueue");

// Package the trigger into an inter-role communication event.
var interRoleEvent = new
InterRoleCommunicationEvent(CloudEnvironment.CurrentRoleInstanceId, trigger);

// Publish inter-role communication event via the Service Bus one-way multicast.
interRoleCommunicator.Publish(interRoleEvent);
}
}
Now that we have built a queue listener that is capable of supporting multi-threading, auto-scaling
and push-based notifications, its time to consolidate all recommendations pertaining to the
design of queue-based messaging solutions on the Windows Azure platform.

125
Conclusion
To maximize the efficiency and cost effectiveness of queue-based messaging solutions running
on the Windows Azure platform, solution architects and developers should consider the following
recommendations.
As a solution architect, you should:
Provision a queue-based messaging architecture that uses the Windows Azure queue
storage service for high-scale asynchronous communication between tiers and services
in cloud-based or hybrid solutions.
Recommend partitioned queuing architecture to scale beyond 500 transactions/sec.
Understand the fundamentals of Windows Azure pricing model and optimize solution to
lower transaction costs through a series of best practices and design patterns.
Consider dynamic scaling requirements by provisioning an architecture that is adaptive to
volatile and fluctuating workloads.
Employ the right auto-scaling techniques and approaches to elastically expand and shrink
compute power to further optimize the operating expense.
Evaluate the cost-benefit ratio of reducing latency through taking dependency on Windows
Azure Service Bus for real-time push-based notification dispatch.
As a developer, you should:
Design a messaging solution that employs batching when storing and retrieving data
from Windows Azure queues.
Implement an efficient queue listener service ensuring that queues will be polled by a
maximum of one dequeue thread when empty.
Dynamically scale down the number of worker role instances when queues remain
empty for a prolonged period of time.
Implement an application-specific random exponential back-off algorithm to reduce the
effect of idle queue polling on storage transaction costs.
Adopt the right techniques that prevent from exceeding the scalability targets for a single
queue when implementing highly multi-threaded multi-instance queue publishers and
consumers.
Employ a robust retry policy capable of handling a variety of transient conditions when
publishing and consuming data from Windows Azure queues.
Use the one-way eventing capability provided by Windows Azure Service Bus to support
push-based notifications in order to reduce latency and improve performance of the queue-
based messaging solution.
Explore the new capabilities of the .NET Framework 4 such as TPL, PLINQ and Observer
pattern to maximize the degree of parallelism, improve concurrency and simplify the
design of multi-threaded services.
The accompanying sample code is available for download from the MSDN Code Gallery. The
sample code also includes all the required infrastructure components such as generics-aware
abstraction layer for the Windows Azure queue service which were not supplied in the above
code snippets. Note that all source code files are governed by the Microsoft Public License as
explained in the corresponding legal notices.

126
Additional Resources/References
For more information on the topic discussed in this whitepaper, please refer to the following:
Understanding Windows Azure Storage Billing Bandwidth, Transactions, and Capacity post
on the Windows Azure Storage team blog.
Service Management API article in the MSDN Library.
About the Service Management API in Windows Azurepost on Neil Mackenzies blog.
Windows Azure Service Management CmdLets project on the MSDN CodePlex.
Windows Azure Storage Abstractions and their Scalability Targets post on the Windows
Azure Storage team blog.
Queue Read/Write Throughput study published by eXtreme Computing Group at Microsoft
Research.
The Transient Fault Handling Framework for Azure Storage, Service Bus & SQL Azure
project on the MSDN Code Gallery.
The Autoscaling Application Block in the MSDN library.
Windows Azure Storage Transaction - Unveiling the Unforeseen Cost and Tips to Cost
Effective Usage post on Wely Laus blog.


127
How to Integrate a BizTalk Server Application with
Service Bus Queues and Topics
Author: Paolo Salvatori
Reviewers: Ralph Squillace, Thiago Almeida
This article shows how to integrate a BizTalk Server 2010 application with Windows Azure
Service Bus queues, topics, and subscriptions to exchange messages with external systems in a
reliable, flexible, and scalable manner. Queues and topics, introduced in the September 2011
Windows Azure AppFabric SDK, are the foundation of a new cloud-based messaging and
integration infrastructure that provides reliable message queuing and durable publish/subscribe
messaging capabilities to both cloud and on-premises applications based on Microsoft and non-
Microsoft technologies. .NET applications can use the new messaging functionality from either a
brand-new managed API (Microsoft.ServiceBus.Messaging) or via WCF thanks to a new binding
(NetMessagingBinding), and any Microsoft or non-Microsoft applications can use a REST style
API to access these features.
Microsoft BizTalk Server enables organizations to connect and extend heterogeneous systems
across the enterprise and with trading partners. The Service Bus is part of Windows Azure
AppFabric and is designed to provide connectivity, queuing, and routing capabilities not only for
cloud applications but also for on-premises applications. Using both together enables a significant
number of scenarios in which you can build secure, reliable and scalable hybrid solutions that
span the cloud and on premises environments:
1. Exchange electronic documents with trading partners.
2. Expose services running on-premises behind firewalls to third parties.
3. Enable communication between spoke branches and a hub back office system.
In this article you will learn how to use WCF in a .NET and BizTalk Server application to execute
the following operations:
1. Send messages to a Service Bus topic.
2. Receive messages from a Service Bus queue.
3. Receive messages from a Service Bus subscription.
In this article you will also learn how to translate the explicit and user-defined properties of a
BrokeredMessage object into the context properties of a BizTalk message and vice versa. Before
describing how to perform these actions, I will start with a brief introduction of the elements that
compose the solution:
In This Section
Service Bus Queues
Service Bus Topics
BrokeredMessage
NetMessagingBinding
BizTalk WCF Adapters
Scenarios
Soluti
Testin
Imple
Concl
See
Now
Queu
AppF
Unde
Blog.
Build
the ch
Serv
Secu
Serv

Queu
applic
reliab

Queu
size o
allows
throug
1. T
as
2. W
3. S
ion
ng the Solutio
menting a Me
lusion
e Also
Available: Th
ues, Topics, a
Fabric Service
erstanding Wi
ding loosely-c
hannel9 site.
ice Bus Topic
uring Service
vice Bus
es provide m
cations runnin
le fashion acr
es are hosted
of a queue is 5
s you to creat
gh the followi
he new mess
ssembly.
Windows Com
ervice Bus R
on
essage Fan-O
he Service Bu
and Subscript
e Bus topic o
indows Azure
oupled apps w
cs And Queue
Bus with ACS
Queues
essaging cap
ng on premise
ross network
d in Windows
5GB. The ma
te unlimited-s
ng APIs:
saging API av
mmunication F
EST API.
Out Scenario
us September
tions article o
on the MSDN
e AppFabric Q
with Windows
es video on t
S video on th
pabilities that
es or in the clo
and trust bou
Azure by a r
aximum mess
size sequence
vailable throug
Foundation (W
r 2011 Releas
on the MSDN
site.
Queues (and T
s Azure Servi
the channel9
he channel9 s
enable a larg
oud to exchan
undaries.
eplicated, dur
age size is in
es of related m
gh the new M
WCF) through
se article on
N site.
Topics) video
ice Bus Topic
site.
site.
ge and hetero
nge message
rable store in
nstead 256KB
messages. Qu
Microsoft.Serv
the new NetM
the Windows
o on the AppF
cs and Queue
geneous clas
es in a flexible
frastructure. T
B, but the use
ueues are ac
iceBus.Messa
MessagingBin

12
s Azure Blog
Fabric Team
es video on
ss of
e, secure and

The maximum
of sessions
cessed
aging
nding class.
28

m

129
Queue entities provide the following capabilities:
1. Session-based correlation, allowing you to build multiplexed request/reply paths easily.
2. Reliable delivery patterns using the PeekLock receive mode.
3. Transactions support to ensure that batches of messaging operations (both send and
receive) are committed atomically.
4. Detection of inbound message duplicates, allowing clients to send the same message
multiple times without adverse consequences.
5. Dead-letter facility for messages that cannot be processed or expire before being received.
6. Deferring messages for later processing. This functionality is particularly handy in the
following cases:
a. When messages are received out of the expected sequence and need to be saved while
a process waits for a particular message to permit further progress.
b. When messages must be processed based on a set of properties that define their
priorities during a traffic peak.
In the .NET API the message entity is modeled by the BrokeredMessage class, which exposes
properties (such as MessageId, SessionID, and CorrelationId) that you can use to exploit
capabilities such as automatic duplicate detection or session-enabled communications. You can
use a QueueDescription object to specify the metadata that models the behavior of the queue
being created:
1. The DefaultMessageTimeToLive property specifies the default message time-to-live value.
2. The DuplicateDetectionHistoryTimeWindow property defines the duration of the duplicate
detection history.
3. The EnableDeadLetteringOnMessageExpiration property allows you to enable or disable the
dead-letter facility on message expiration.
4. The LockDuration property defines the duration of the lock used by a consumer when using
the PeekLock receive mode. In the ReceiveAndDelete receive mode, a message is deleted
from the queue as soon as a consumer reads it. Conversely, in the PeekLock receive mode,
a message is hidden from other receivers until the timeout defined by the LockDuration
property expires. By that time the receiver should have deleted the message invoking the
Complete method on the BrokeredMessage object.
5. The MaxSizeInMegabytes property defines the maximum queue size in megabytes.
6. The RequiresDuplicateDetection property enables\disables duplicate message detection. The
RequiresSession property enables\disables sessions.
7. The MessageCount return the number of messages in the queue. This is property is
extremely useful because, based on its value, an intelligent system can decide to increase or
decrease the number of competing consumers that concurrently receive and processes
messages from the queue.
Since metadata cannot be changed once a messaging entity is created, modifying the
duplicate detection behavior requires deleting and recreating the queue. The same
principle applies to any other metadata.
Using queues permits you to flatten highly-variable traffic into a predictable stream of work and
distribute the load across a set of worker processes size of which can vary dynamically to
Warning
accom
publis
the m
In ser
intera
conte
increa
loose



See A
Unde
Blog.
An In
Wind
Wind
Serv
Topic
capab

A topi
numb
mess
subsc
mmodate the
sher writes a m
message, but o
rvice-oriented
actions betwee
xt, the use of
asing the agili
coupling of in
Also
erstanding Wi
ntroduction to
dows Azure A
dows Azure A
vice Bus
cs extend the
bilities.
ic entity cons
ber is subject t
age copies to
cription can de
incoming me
message to a
only one of th
d or service bu
en autonomo
f messaging e
ity, scalability
ndividual syst
indows Azure
Service Bus
AppFabric Ser
AppFabric Lea
Topics
messaging fe
ists of a sequ
to vary in the
o a poll of wor
efine one or m
ssage volume
a queue, mult
hem will receiv
us architectur
us systems a
entities like qu
y and flexibility
tems.
e AppFabric Q
Queues arti
rvice Bus Que
arning Series
eatures provid
uential messa
future) concu
rker processe
multiple rule e
e. In a compe
iple consume
ve and proces
re composed
are asynchron
ueues and top
y of the overa
Queues (and T
cle on the Ap
eues and Top
available on
ded by queue
ge store like
urrent and du
es. As depicte
entities.
eting consume
ers compete w
ss the messa
of multiple, h
nous and loos
pics (see the
all architecture
Topics) video
ppFabric Team
pics on Neil M
CodePlex.
es with the ad
a queue, but
rable subscri
ed in the follow
ers scenario,
with each othe
age in questio
eterogeneous
sely-coupled.
next section)
e and helps d
o on the AppF
m Blog.
Mackenzie's B
dition of publ
it supports up
ptions, which
wing figure, e

13
when a
er to receive
on.
s systems,
In this
allows
decreasing the
Fabric Team
Blog.
ish-subscribe

p to 2000 (this
relay
each
30
e
e
s

131


Each rule specifies a filter expression that is used to filter messages that pass through the
subscription and a filter action that can modify message properties. In particular, the SqlFilter
class allows a SQL92-like condition to be defined on message properties, as in the following
examples:
1. OrderTotal > 5000 OR ClientPriority > 2
2. ShipDestinationCountry = USA AND ShipDestinationState = WA
Conversely, the SqlRuleAction class can be used to modify, add or remove properties to a
BrokeredMessage object using syntax similar to that used by the SET clause of an UPDATE
SQL-command, as in the following examples.
1. SET AuditRequired = 1
2. SET Priority = 'High', Severity = 1
Each matching rule generates a separate copy of the published message, so any
subscription can potentially generate more copies of the same message, one for each
matching rule.
Topics, like queues, support a competing consumers scenario: in this context, a subscription can
have a single consumer that receives all messages or a set of competing consumers that fetch
messages on a first-come-first-served basis. Topics are a useful messaging solution for
broadcasting messages to many consumers.
Warning

132




See Also

Understanding Windows Azure AppFabric Queues (and Topics) video on the AppFabric Team
Blog.
An Introduction to Service Bus Topics article on the AppFabric Team Blog.
Windows Azure AppFabric Service Bus Queues and Topics on Neil Mackenzie's Blog.
Windows Azure AppFabric Learning Series available on CodePlex
BrokeredMessage
The BrokeredMessage class models the messages exchanged by applications that communicate
through queues and topics. The class provides four different public constructors:
1. The BrokeredMessage() constructor initializes a new instance of the BrokeredMessage class
with an empty payload. A BrokeredMessage object exposes a collection of key-value pairs in
the Properties, which allows you to define a set of custom properties. It is common for a
message to have an empty body because user-defined properties can carry the message
payload.
2. The BrokeredMessage(Object) constructor initializes a new instance of the BrokeredMessage
class from a given object by using a DataContractSerializer object with a binary
XmlDictionaryWriter object.
3. BrokeredMessage(Stream, Boolean) constructor initializes a new instance of the
BrokeredMessage class using the supplied stream as its body.
4. Finally, the BrokeredMessage(Object, XmlObjectSerializer) constructor initializes an instance
of the BrokeredMessage class from a given object using the provided XmlObjectSerializer
object.
The class exposes an interesting set of methods that allow execution of a wide range of actions
at the message level:
1. In the PeekLock receive mode, the Abandon method can release the lock on a peek-locked
message. The Complete method commits the receive operation of a message and indicates
that the message should be marked as processed and then either deleted or archived. The
Defer method indicates that the receiver wants to defer processing for a message. As
mentioned before, deferring messages is a convenient way to handle situations where
messages are received out of the expected sequence and need to be safely parked while the
applications waits for a particular message before proceeding with message processing.
2. The DeadLetter and DeadLetter(String, String) methods allows an application to explicitly
move a message to the dead-letter queue of a queue or a subscription. Take into account
that when you create a queue entity using the managing API or the Windows Azure
Management Portal, you can configure it to automatically move expired messages to the

133
dead-letter queue. In the same way, you can configure a subscription to move expired
messages and messages to its dead-letter queue if they fail filter evaluation.
The BrokeredMessage class exposes a wide range of properties:

Property Description
ContentType Specifies the type of the content.
MessageId Indicates the identifier of the message.
CorrelationId Implement a request-reply message exchange
pattern in which the client application uses the
MessageId property of an outgoing request
message and the CorrelationId property of an
incoming response message to correlate the
two messages. We will see an implementation
of this technique later in this article.
SessionId Sets or gets the session identifier for a
message. A competing consumer scenario
where multiple worker processes receive
messages from the same session-enabled
queue or subscription guarantees that
messages sharing the same SessionId are
received by the same consumer. In this
context, when a client application sends a flow
of request messages to a server application
using a session-enabled queue or topic and
waits for the correlated response messages on
a separate session-enabled queue or a
subscription, the client application can assign
the receive sessions ID to the
ReplyToSessionId property of outgoing
messages. This property value gives the server
application the value to assign to the SessionId
property of response messages.
ReplyTo Gets or sets the address of the queue to reply
to. In an asynchronous request-reply scenario a
client application can send a request message
to a server application through a Service Bus
queue or topic and wait for a response
message. In this scenario, by convention the
client application can use the ReplyTo property
of the request message to indicate to the server
application the address of the queue or topic to
send the response to. We will see an

134
application of this technique when I introduce
the second test case where an orchestration
reads the ReplyTo address from a context
property of the request message and assign its
value to the Address property of a dynamic
send port to return the response message to
the expected queue or topic.
Label Gets or sets an application specific label.
SequenceNumber Returns the unique number that the Service
Bus assigns to a message. Use this property to
retrieve a deferred message from a queue or a
subscription.
TimeToLive Allows you to set or get the messages time-to-
live value. The Service Bus does not enforce a
maximum lifetime for messages waiting to be
processed in a queue or a subscription.
Nevertheless, you can define a default time to
live for messages when you create a queue,
topic, or subscription, or you can explicitly
define an expiration timeout at the message
level using the TimeToLive property.
DeliveryCount Returns the number of message deliveries.
Properties A collection that allows you to define
application specific message properties. This
collection is probably the most important
feature of a BrokeredMessage entity because
user-defined properties can be used for the
following:
1. Carry the payload of a message. In this
context, the body of the message could be
empty.
2. Define application specific properties that
can be used by a worker process to decide
how to process the current message.
3. Specify filter and action expressions that
can be used to define routing and data
enrichment rules at a subscription level

If you are familiar with context properties for a BizTalk message, it is helpful to think of the user-
defined properties contained in the BrokeredMessageProperties collection as the context
properties of a BizTalk message. In fact, these properties can carry a piece of information or even
the entire message payload. If you are using topics and subscriptions, these properties can route

135
the message to the proper destination. Hence, in a scenario where a third party system
exchanges messages with a BizTalk application through the Service Bus, its very important to
translate the application specific properties carried by a BrokeredMessage object to and from the
context properties of a BizTalk message. In the article and in the companion code Ill show you
how to achieve this result.
As indicated in Windows Azure AppFabric Service Bus Quotas, the maximum size for
each property is 32K. Cumulative size of all properties cannot exceed 64K. This applies
to the entire header of the BrokeredMessage, which has both user properties as well as
system properties (such as SequenceNumber, Label, MessageId, and so on). The space
occupied by properties counts towards the overall size of the message and its maximum
size of 256K. If an application exceeds any of the limits mentioned above, a
SerializationException exception is generated, so you should expect to handle this error
condition.




NetMessagingBinding
Service Bus Brokered Messaging supports the WCF programming model and provides a new
binding called NetMessagingBinding that can be used by WCF-enabled applications to send and
receive messages through queues, topics and subscriptions. NetMessagingBinding is the new
name for the binding for queues and topics that provides full integration with WCF. From a
functional perspective, NetMessagingBinding is similar to NetMsmqBinding, which supports
queuing by using Message Queuing (MSMQ) as a transport and enables support for loosely-
coupled applications. On the service-side, NetMessagingBinding provides an automatic
message-pump that pulls messages off a queue or subscription and is integrated with WCFs
ReceiveContext mechanism.
The new binding supports the standard interfaces IInputChannel, IOutputChannel, and
IInputSessionChannel. When an application uses WCF and NetMessagingBinding to send a
message to a queue or a topic, the message is wrapped in a SOAP envelope and encoded. To
set BrokeredMessage-specific properties, you need to create a BrokeredMessageProperty object,
set the properties on it, and add it to the Properties collection of the WCF message, as shown in
the following example. When using the NetMessagingBinding to write a message to queue or a
topic, the ServiceBusOutputChannel internal class looks for the BrokeredMessageProperty
property in the Properties collection of the WCF message and copies all its properties to the
BrokeredMessage object it creates. Then it copies the payload from the WCF message to the
BrokeredMessage object and finally, publishes the resulting message to the target queue or topic.
static void Main(string[] args)
{
Warning

136
try
{
// Create the
var channelFactory = new ChannelFactory<IOrderService>("orderEndpoint");
var clientChannel = channelFactory.CreateChannel();

// Create a order object
var order = new Order()
{
ItemId = "001",
Quantity = 10
};

// Use the OperationContextScope to create a block within which to access the
current OperationScope
using (var scope = new OperationContextScope((IContextChannel)clientChannel))
{
// Create a new BrokeredMessageProperty object
var property = new BrokeredMessageProperty();

// Use the BrokeredMessageProperty object to set the BrokeredMessage
properties
property.Label = "OrderItem";
property.MessageId = Guid.NewGuid().ToString();
property.ReplyTo = "sb://acme.servicebus.windows.net/invoicequeue";

// Use the BrokeredMessageProperty object to define application-specific
properties
property.Properties.Add("ShipCountry", "Italy");
property.Properties.Add("ShipCity", "Milan");

// Add BrokeredMessageProperty to the OutgoingMessageProperties bag provided
// by the current Operation Context

OperationContext.Current.OutgoingMessageProperties.Add(BrokeredMessageProperty.Name,
property);
clientChannel.SendOrder(order);
}
}
catch (Exception ex)
{
Console.WriteLine(ex.Message);

137
}
}
Likewise, when you use a NetMessagingBinding-based service endpoint to receive messages
from a queue or a topic, an application can retrieve the BrokeredMessageProperty object from
the Properties collection of the incoming WCF Message, as shown in the following example. In
particular, when the service endpoint receives messages, the ServiceBusInputChannel and
ServiceBusInputSessionChannel internal classes create a new WCF Message and copy the
payload from the body on the inbound to the body of the newly created WCF message.
(ServiceBusInputSessionChannel is used to receive messages from sessionful queues and
subscriptions). Then they copy the properties from the inbound BrokeredMessage to a new
instance of the BrokeredMessageProperty class and, finally, add the new
BrokeredMessageProperty object to the Properties collection of the incoming WCF Message.
[ServiceBehavior]
public class OrderService : IOrderService
{
[OperationBehavior]
public void ReceiveOrder(Order order)
{
// Get the BrokeredMessageProperty from the current OperationContext
var incomingProperties = OperationContext.Current.IncomingMessageProperties;
var property = incomingProperties[BrokeredMessageProperty.Name] as
BrokeredMessageProperty;

...
}
}
Because the Service Bus does not support IInputSessionChannel, all applications sending
messages to session-enabled queues must use a service contract where the value of the
SessionMode property is different from that of the SessionMode.Required. However, because the
Service Bus WCF runtime supports IInputSessionChannel and other interfaces order to receive
messages from a sessionful queue or subscription that uses WCF and NetMessagingBinding,
applications must implement a session-aware service contract. The following code snippet
provides an example of a WCF service that receives messages from a sessionful
queue/subscription.
// ServiceBus does not support IOutputSessionChannel.
// All senders sending messages to sessionful queue must use a contract which does not
enforce SessionMode.Required.
// Sessionful messages are sent by setting the SessionId property of the
BrokeredMessageProperty object.
[ServiceContract]
public interface IOrderService
{

138
[OperationContract(IsOneWay = true)]
[ReceiveContextEnabled(ManualControl = true)]
void ReceiveOrder(Order order);
}

// ServiceBus supports both IInputChannel and IInputSessionChannel.
// A sessionful service listening to a sessionful queue must have SessionMode.Required in
its contract.
[ServiceContract(SessionMode = SessionMode.Required)]
public interface IOrderServiceSessionful : IOrderService
{
}

[ServiceBehavior(InstanceContextMode = InstanceContextMode.PerSession, ConcurrencyMode =
ConcurrencyMode.Single)]
public class OrderService : IOrderServiceSessionful
{
[OperationBehavior]
public void ReceiveOrder(Order order)
{
// Get the BrokeredMessageProperty from the current OperationContext
var incomingProperties = OperationContext.Current.IncomingMessageProperties;
var property = incomingProperties[BrokeredMessageProperty.Name] as
BrokeredMessageProperty;

...

//Complete the Message
ReceiveContext receiveContext;
if (ReceiveContext.TryGet(incomingProperties, out receiveContext))
{
receiveContext.Complete(TimeSpan.FromSeconds(10.0d));
...
}
else
{
throw new InvalidOperationException("...");
}
}
}
Note that the ManualControl property of the ReceiveContextEnabled operation attribute is set to
true. This setting makes the service explicitly invoke the ReceiveContext.Complete method to

139
commit the receive operation. In fact, when the ManualControl property is set to true, the
message received from the channel is delivered to the service operation with a lock for the
message. The service implementation must either call Complete(TimeSpan) or
Abandon(TimeSpan) to signal that the message has been received completely. Failure to call one
of these methods results in the lock being held on the message until the lock timeout interval
elapses. Once the lock is released (either by a call to Abandon(TimeSpan) or by a lock timeout)
the message is dispatched again from the channel to the service. Calling Complete(TimeSpan)
marks the message as successfully received.
Note also that the OrderService class sets the ServiceBehavior.InstanceContextMode property to
InstanceContextMode.PerSession and the ConcurrencyMode property to
ConcurrencyMode.Single. This way the ServiceHost creates a new service instance every time a
new session is available in the specified queue or subscription, and the ServiceHost uses a single
thread to receive messages sequentially from the queue or subscription. The SessionIdleTimeout
property of the NetMessagingBinding controls the lifetime of the service. For more information on
how to use WCF exchange messages through a sessionful queue, I strongly suggest you to look
at the WcfServiceSessionSample sample included in the Windows Azure AppFabric SDK V1.5.





BizTalk WCF Adapters
Each WCF receive location in BizTalk Server is an instance of a WCF service class called
BizTalkServiceInstance, which is hosted by a separate instance of a ServiceHost-derived class:
1. The BtsServiceHost for receive locations (RLs) running in an in-process host.
2. The WebServiceHost for RLs running in an isolated host.
When you enable a WCF receive location, the adapter initializes and opens the ServiceHost,
which dynamically builds the WCF runtime components within the BizTalk service process
(BtsNtSvc.exe for in in-process host, w3wp.exe for an isolated host). These components include
the channel stack, the dispatcher, and the singleton service instance. WCF receive locations and
send ports are message-type agnostic or, if you prefer, untyped. This design comes in handy
when you need to configure a single receive port to accept numerous message types or versions
that you can normalize (using BizTalk maps or a custom pipeline component) into a common
message type before the messages are posted to the BizTalkMsgBoxDb. However, this design
also implies that the WCF adapters must build on generic service contracts in order to remain
message-type agnostic within the WCF implementation.
WCF receive locations must receive the incoming message bytes, perform any necessary SOAP
and WS-* processing, and publish the message (or some part of it) to the message box. The
WCF Adapters create a separate ServiceHost and singleton service object of this class for each
receive location that handles client requests for the lifetime of the BizTalk Host instance that runs

140
WCF receive locations. The service object uses multiple threads to process messages
concurrently unless the WCF-NetMsmq receive locations are used with the Ordered processing
property enabled.. As shown in the picture below, the class is decorated with the ServiceBehavior
attribute:
//BizTalkServiceInstance Class

[ServiceBehavior(InstanceContextMode = InstanceContextMode.Single, ConcurrencyMode =
ConcurrencyMode.Multiple)]
internal sealed class BizTalkServiceInstance : ITwoWayAsync,
ITwoWayAsyncVoid,
IOneWayAsync,
IOneWayAsyncTxn,
ITwoWayAsyncVoidTxn
{
...
}
In particular, you can note that:
1. InstanceContextMode = InstanceContextMode.Single
2. ConcurrencyMode = ConcurrencyMode.Multiple
Hence, all incoming messages to a WCF receive location are received and processed by a single
well-known instance of the BizTalkServiceInstance class. This use of a single service instance
avoids service activation and deactivation costs and improves performance and scalability.
The BizTalkServiceInstance class implements multiple untyped, generic service contracts.
IOneWayAsync
IOneWayAsyncTxn
ITwoWayAsync
ITwoWayAsyncVoid
ITwoWayAsyncVoidTxn
Each contract was designed for a different scope (as suggested by their names):
One-way vs. two-way Message Exchange Pattern.
Transactional vs. nontransactional communication.
[ServiceContract(Namespace = "http://www.microsoft.com/biztalk/2006/r2/wcf-adapter")]
public interface IOneWayAsync
{
// Methods
[OperationContract(AsyncPattern = true, IsOneWay = true, Action = "*")]
IAsyncResult BeginOneWayMethod(Message message, AsyncCallback callback, object
state);
[OperationContract(IsOneWay = true, Action = "BizTalkSubmit")]
void BizTalkSubmit(Message message);
void EndOneWayMethod(IAsyncResult result);

141
}

[ServiceContract(Namespace = "http://www.microsoft.com/biztalk/2006/r2/wcf-adapter")]
public interface IOneWayAsyncTxn
{
// Methods
[OperationContract(AsyncPattern = true, IsOneWay = true, Action = "*")]
IAsyncResult BeginOneWayMethod(Message message, AsyncCallback callback, object
state);
[OperationContract(IsOneWay = true, Action = "BizTalkSubmit")]
void BizTalkSubmit(Message message);
void EndOneWayMethod(IAsyncResult result);
}

[ServiceContract(Namespace = "http://www.microsoft.com/biztalk/2006/r2/wcf-adapter")]
public interface ITwoWayAsync
{
// Methods
[OperationContract(AsyncPattern = true, IsOneWay = false, Action = "*", ReplyAction =
"*")]
IAsyncResult BeginTwoWayMethod(Message message, AsyncCallback callback, object
state);
[OperationContract(IsOneWay = false, Action = "BizTalkSubmit")]
Message BizTalkSubmit(Message message);
Message EndTwoWayMethod(IAsyncResult result);
}

[ServiceContract(Namespace = "http://www.microsoft.com/biztalk/2006/r2/wcf-adapter")]
public interface ITwoWayAsyncVoid
{
// Methods
[OperationContract(AsyncPattern = true, IsOneWay = false, Action = "*", ReplyAction =
"*")]
IAsyncResult BeginTwoWayMethod(Message message, AsyncCallback callback, object
state);
[OperationContract(IsOneWay = false, Action = "BizTalkSubmit")]
void BizTalkSubmit(Message message);
void EndTwoWayMethod(IAsyncResult result);
}

[ServiceContract(Namespace = "http://www.microsoft.com/biztalk/2006/r2/wcf-adapter")]
public interface ITwoWayAsyncVoidTxn

142
{
// Methods
[TransactionFlow(TransactionFlowOption.Mandatory)]
[OperationContract(AsyncPattern = true, IsOneWay = false, Action = "*", ReplyAction =
"*")]
IAsyncResult BeginTwoWayMethod(Message message, AsyncCallback callback, object
state);
[TransactionFlow(TransactionFlowOption.Mandatory), OperationContract(IsOneWay =
false, Action = "BizTalkSubmit")]
void BizTalkSubmit(Message message);
void EndTwoWayMethod(IAsyncResult result);
}
All the methods exposed by these service contracts are generic, asynchronous, and untyped. In
fact, as shown in the previous example, each method is decorated by the OperationContract
attribute and, in particular:
AsyncPattern = True indicates that an operation is implemented asynchronously using a
BeginmethodName and EndmethodName method pair. In a service contract Action = *
means that the method accepts a message with any action.
ReplyAction = * means that the method can return a message with any action.
Every method accepts as a parameter or returns a generic WCF message
As a consequence, each WCF receive location can accept multiple message types, and versions
can be normalized into a canonical format using a different map before being published to the
MessageBox. Also WCF send ports are message-type agnostic.
Scenarios
As I mentioned in the introduction, the main objective of this article is to demonstrate how to
integrate a BizTalk Server application with the new messaging entities supplied by the Service
Bus. To this purpose I designed and implemented a solution made up of several projects in which
a Windows Forms client application sends a request message to a BizTalk application and waits
for a response. The solution implements the following communication patterns:
The client application sends a request message directly to BizTalk Server using the
BasicHttpBinding and a synchronous request-response message exchange pattern. In this
context, the BizTalk application uses a WCF-BasicHttp request-response receive location to
receive requests and send responses. I will not cover this pattern in the article as its out of
scope and is mainly used to verify that the application works as expected
The client application sends a request message to BizTalk Server through the Service Bus
using the NetTcpRelayBinding and a synchronous request-response message exchange
pattern. In this case, to receive requests and send responses, the BizTalk application uses a
WCF-Custom request-response receive location that is configured to use the
NetTcpRelayBinding and the TransportClientEndpointBehavior.
TransportClientEndpointBehavior specifies the Service Bus credentials that the receive
location uses to authenticate with the Access Control Service. Even if this pattern uses the
Service Bus, I will not cover it in the article as it does use Relay Messaging and not the

143
Brokered Messaging features of the Service Bus. You can consider it as a sort of bonus
track!
The client application exchanges request and response messages with the BizTalk
application using Service Bus queues, topics, and subscriptions. In the remainder of the
article well analyze this case.
The Windows Azure AppFabric SDK V1.5 requires the .NET Framework version 4. As a
consequence, in order to exploit the new messaging features provided by the Service
Bus in a BizTalk application, you need to use BizTalk Server 2010, which is the only
version to support the .NET Framework version 4.




StaticSendPortOrchestration
In this example, the StaticSendPortOrchestration can receive request messages from a queue
A or a subscription X and return the corresponding response message to a queue B or topic Y.
The latter are defined in two separate WCF-Custom static one-way send ports. Besides, the
StaticSendPortOrchestration assigns the MessageId of the request message to the
CorrelationId of the response to allow the client application to correlate the response to the initial
request.
1. The client application sends the request message to a queue A and waits for the response
message on a separate queue B. In a competing consumer scenario, the queue B is shared
by multiple client applications to receive response messages. As a consequence, the
response message could be read by a client other than that which sent the initial request.
2. The client application sends the request message to a queue A and waits for the response
message on a subscription Z. In this case, the StaticSendPortOrchestration writes the
response message to statically defined topic Y. In a competing scenario, different instances
of the client application should use separate subscriptions defined by complementary rules to
make sure that the response message is received by the same instance that initiated the
request. Conversely, to implement a decoupled message fan-out scenario where the BizTalk
application sends a notification message to multiple subscribers, you can create a separate
subscription for each of the consumers and properly define rules to receive only those
messages that satisfy a certain condition.
3. The client application sends the request message to a topic A and waits for the response
message on a queue B. The same considerations made at point 1 apply to this case.
4. The client application sends the request message to a topic X and waits for the response
message on a subscription Z. Even in this case, the StaticSendPortOrchestration writes
the response message to statically defined topic Y. The same considerations made at point 2
apply to this case.
Warning
The fo

1. T
re
2. T
S
C
m
ollowing figur
he Windows
equestqueue
he BizTalk ap
ervice Bus. In
Custom.NetM
messages from
e shows the a
Forms applic
or to the requ
pplication use
n particular, a
MessagingBin
m the reques
architecture o
ation uses a W
uesttopic.
es a one-way
a WCF-Custom
nding.Queue
tqueue, whe
of the StaticS
WCF proxy to
receive port t
m receive loc
e.ReceiveLoc
reas another
SendPortOrch
o send a requ
to receive req
cation called S
cation is used
WCF-Custom
hestration ex
uest message
quest messag
ServiceBusS
d to read requ
m receive loca

14
xample.

e to the
ges from the
Sample.WCF-
uest
ation called
44
-

145
ServiceBusSample.WCF-Custom.NetMessagingBinding.Subscription.ReceiveLocation
is used to receive request messages from the italyMilan subscription. Both receive locations
are configured to use the following components:
The NetMessagingBinding is used to receive messages from a queue or a subscription.
A custom WCF Message Inspector called ServiceBusMessageInspector. At runtime,
this component reads the BrokeredMessageProperty from the Properties collection of the
inbound WCF Message and transforms its public properties (such as MessageId and
ReplyTo) and application specific properties (the key/value pairs contained in the
Properties collection of the BrokeredMessageProperty) into context properties of the
BizTalk message.
The TransportClientEndpointBehavior specifies the Service Bus credentials (in this
sample I use SharedSecret credentials) used to authenticate with the Access Control
Service.
The ListenUriEndpointBehavior is used by the WCF-Custom receive location that
retrieves messages from the ItalyMilan subscription defined on the requesttopic. This
custom component is used to set the value of the ListenUri property of the service
endpoint. See later in this article for more information about this component.
If an application retrieves messages from a sessionful queue or subscription, the
SessionChannelEndpointBehavior must be added to the configuration of the WCF-
Custom receive location. This custom component is used to make sure that at runtime
the WCF-Custom adapter creates an IInputSessionChannel in place of an IInputChannel.
This is a mandatory requirement to receive messages from a sessionful messaging
entity.
3. The Message Agent submits the request message to the MessageBox (BizTalkMsgBoxDb).
4. The inbound request starts a new instance of the StaticSendPortOrchestration. The latter
uses a Direct Port to receive request messages that satisfy the following filter expression:
Microsoft.WindowsAzure.CAT.Schemas.Method == StaticSendPort. The orchestration invokes a
method exposed by the RequestManager helper component to calculate the response
message. Then it copies the context properties from the request message to the response
message and then assigns the value of the MessageId context property of the request
message to the CorrelationId property of the response message. Finally, it checks the string
value contained in the ReplyTo context property: if it contains the word "queue," the
orchestration returns the response message through a logical port bound to a WCF-Custom
send port configured to send messages to the responsequeue, otherwise it publishes the
response to the MessageBox via a logical send port bound to a WCF-Custom send port
configured to send messages to the responsetopic.
5. The Message Agent submits the request message to the MessageBox (BizTalkMsgBoxDb).
6. The response message is consumed by one of the following WCF-Custom send ports:
ServiceBusSample.WCF-Custom.NetMessagingBinding.Queue.SendPort: this send
port writes message to the responsequeue.
ServiceBusSample.WCF-Custom.NetMessagingBinding.Topic.SendPort: this send
port writes message to the responsetopic
7. The selected send port writes the response message to the responsequeue or
responsetopic. Both send ports are configured to use the following components:
The NetMessagingBinding is used to send messages to the Service Bus.

146
The ServiceBusMessageInspector transforms the message context properties into
BrokeredMessage properties
The TransportClientEndpointBehavior specifies the Service Bus credentials (in this
sample I use SharedSecret credentials) used by to authenticate with the Access Control
Service.
8. The client application uses a WCF service with 2 endpoints to retrieve the reply message
from the responsequeue or from the responsetopic. In an environment with multiple client
applications, each of them should use a separate queue or subscription to receive response
messages from BizTalk.




DynamicSendPortOrchestration
In this example, the DynamicSendPortOrchestration can receive request messages from a
queue A or a subscription X and return the corresponding response message to a queue or a
topic whose address has been provided by the client in the ReplyTo property of the
BrokeredMessage. Different clients can use distinct queues or subscriptions defined on different
topics to receive response messages. In my demo, the only requirement is that they all belong to
the same Service Bus namespace. By the way, this constraint can be easily removed by using a
repository (such as SQL Server) to store credentials for different namespaces. This test case
demonstrates how to implement a dynamic request-reply message exchange pattern in which the
sender communicates to the receiver the address of the messaging entity (queue or topic) where
to send the response and the receiver (the orchestration) assigns the MessageId of the request
message to the CorrelationId of the response to allow the client application to correlate the
response to the initial request.
1. The client application specifies the address of the response queue B in the ReplyTo property
of the request message sends the latter to a queue A and waits for the response message on
a queue B. The DynamicSendPortOrchestration retrieves the URL of the response queue
from the ReplyTo context property of the request message and assign it to the address of the
dynamic send port used to send the response. In a competing consumer scenario, different
clients can use distinct queues to receive response messages. This way, the response
message will always be received by the client that sent the initial request.
2. The client application specifies the address of the response topic Y in the ReplyTo property
of the request message sends the latter to a queue A and waits for the response message on
a subscription Z. The DynamicSendPortOrchestration retrieves the URL of the response
topic from the ReplyTo context property of the request message and assign it to the address
of the dynamic send port used to send the response. In a competing scenario, different
instances of the client application should use separate subscriptions defined by
complementary rules to make sure that the response message is received by the same
instance that initiated the request. However, the subscription they use to receive responses
from the BizTalk application can be associated to different topics. Conversely, to implement a

147
decoupled message fan-out scenario where the BizTalk application sends a notification
message to multiple subscribers, you can create a separate subscription for each of the
consumers and properly define rules to receive only those messages that satisfy a certain
condition.
3. The client application specifies the address of the response queue B in the ReplyTo property
of the request message sends the latter to a topic X and waits for the response message on
a queue B. The same considerations made at point 1 apply to this case.
4. The client application specifies the address of the response topic Y in the ReplyTo property
of the request message sends the latter to a topic X and waits for the response message on
a subscription Z. The same considerations made at point 2 apply to this case.
The following figure shows the architecture of the DynamicSendPortOrchestration example.

1. T
re
2. T
S
C
m
S
he Windows
equestqueue
he BizTalk ap
ervice Bus. In
Custom.NetM
messages from
erviceBusSa
Forms applic
e or to the req
pplication use
n particular, a
MessagingBin
m the reques
ample.WCF-
ation uses a W
questtopic.
es a one-way
a WCF-Custo
nding.Queue
tqueue, whe
WCF proxy to
receive port t
om receive loc
e.ReceiveLoc
reas another
o send a requ
to receive req
cation called
cation is used
WCF-Custom
uest message
quest messag
ServiceBusS
d to read requ
m receive loc

14
e to the
ges from the
Sample.WCF
uest
cation called
48
F-


149
Custom.NetMessagingBinding.Subscription.ReceiveLocationis used to receive request
messages from the italyMilan subscription. Both receive locations are configured to use the
following components:
The NetMessagingBinding is used to receive messages from a queue or a subscription
A custom WCF message inspector called ServiceBusMessageInspector. At runtime,
this component reads the BrokeredMessageProperty from the Properties collection of the
inbound WCF Message and transforms its public properties (such as MessageId and
ReplyTo) and application specific properties (the key/value pairs contained in the
Properties collection of the BrokeredMessageProperty) into context properties of the
BizTalk message.
The TransportClientEndpointBehavior specifies the Service Bus credentials (in this
sample I use SharedSecret credentials) used to authenticate with the Access Control
Service.
The ListenUriEndpointBehavior is used by the WCF-Custom receive location that
retrieves messages from the italyMilan subscription defined on the requesttopic. This
custom component is used to set the value of the ListenUri property of the service
endpoint. See later in this article for more information about this component.
When an application retrieves messages from a sessionful queue or subscription, the
SessionChannelEndpointBehavior must be added to the configuration of the WCF-
Custom receive location. This custom component is used to make sure that at runtime
the WCF-Custom adapter creates an IInputSessionChannel in place of an
IInputChannel. This is a mandatory requirement to receive messages from a sessionful
messaging entity.
3. The Message Agent submits the request message to the MessageBox
(BizTalkMsgBoxDb).
4. The inbound request starts a new instance of the DynamicSendPortOrchestration. The
latter uses a Direct Port to receive request messages that satisfy the following filter
expression: Microsoft.WindowsAzure.CAT.Schemas.Method == DynamicSendPort. The
orchestration invokes a method exposed by the RequestManager helper component to
calculate the response message. Then it copies the context properties from the request
message to the response message and then assign the value of the MessageId context
property of the request message to the CorrelationId property of the response message.
Then it reads from the ReplyTo context property the address of the messaging entity (queue
or topic) to which it should send the response. Finally, the orchestration configures the
context properties of the response message and the properties of the dynamic send port to
use the following components:
The NetMessagingBinding is used to send messages to the Service Bus.
The ServiceBusMessageInspector transforms the message context properties into
BrokeredMessage properties
The TransportClientEndpointBehavior specifies the Service Bus credentials (in this
sample I use SharedSecret credentials) used to authenticate with the Access Control
Service.
5. The Message Agent submits the response message to the MessageBox
(BizTalkMsgBoxDb).
6. The response message is consumed by the dynamic one-way send port named
ServiceBusSample.Dynamic.SendPort.

150
7. The send port writes the response message to the messaging entity (in my sample
responsequeue or responsetopic) whose address has been indicated by the client
application in the ReplyTo property of the BrokeredMessageProperty.
8. The client application uses a WCF service with 2 endpoints to retrieve the reply message
from the responsequeue or from the responsetopic. In an environment with multiple client
applications, each of them should use a separate queue or subscription to receive response
messages from BizTalk.




Solution
Now that we described the scenarios we can analyze in detail the main components of the
solution.
In This Section
Queues, Topics and Subscriptions
Data and Message Contracts
Service Contracts
Client Application
Response Handler Service
ListenUri Endpoint Behavior
SessionChannel Endpoint Behavior
Property Schemas
Schemas
Service Bus Message Inspector
Installing Components in GAC
Registering Components in the machine.config file
Receive Locations
Send Ports
Orchestrations
Queues, Topics and Subscriptions
The first operation to perform to properly configure the environment is creating the messaging
entities used by the demo. The first operation to execute is to provision a new Service Bus
namespace or modify an existing namespace to include the Service Bus. You can accomplish
this ta
button

After y
the Se
the de
easies
Mana

ask from the W
n, respectively
you provision
ervice Bus, th
emo. As I me
st way is by u
age Entities c
Windows Azu
y, highlighted
n a new Servic
he next step is
ntioned in the
using the Win
command bar
ure Managem
d in the figure
ce Bus name
s to create th
e introduction
ndows Azure
r highlighted i
ment Portal b
below.
space or mod
e queue, topi
, you have m
Managemen
n red at point
by clicking the
dify an existin
c and subscr
any options to
nt Portal usin
t 1 in the follo
e New button
ng namespace
iption entities
o accomplish
ng the buttons
owing figure.

15
or the Modif

e to include
s required by
this task. Th
s in the

51
fy
e

152
You can use the navigation tree-view shown at point 2 to select an existing entity and display its
properties in the vertical bar highlighted at point 3. To remove the selected entity, you press the
Delete button in the Manage Entities command bar.
The current user interface supplied by the Windows Azure Management Portal allows a
user to create queues, topics, and subscriptions and define their properties, but not to
create or display rules for an existing subscription. At the moment, you can accomplish
this task by only using the .NET messaging API. In particular, to add a new rule to an
existing subscription you can use the AddRule(String, Filter) or the
AddRule(RuleDescription) methods exposed by the SubscriptionClient class, whereas to
enumerate the rules of an existing subscription, you can use the GetRules method of the
NamespaceManager class. The latter can be used for managing entities, such as
queues, topics, subscriptions, and rules, in a Service Bus namespace.In June, I created a
tool called Service Bus Explorer that allows a user to create, delete and test queues,
topics, subscriptions, and rules. My tool was able to manage entities in the AppFabric
Labs Beta environment. However, the new version of the Service Bus API introduced
some breaking changes, as you can read here, so I'm working at a new version of the
Service Bus Explorer tool that will introduce a significant amount of interesting and
hopefully useful features. So stay tuned and come back frequently on this site as I plan to
publish it soon.
For your convenience, I created a console application called Provisioning that uses the
functionality provided by the NamespaceManager class to create the queues, topics, and
subscriptions required by the solution. When it starts, the console applications prompt for service
namespace credentials. These are used to authenticate with the Access Control service, and
acquire an access token that proves to the Service Bus infrastructure that the application is
authorized to provision new messaging entities. Then the application prompts for the value to
assign to the properties of the entities to create such as EnabledBatchedOperations and
EnableDeadLetteringOnMessageExpiration for queues. The Provisioning application creates the
following entities in the specified Service Bus namespace:
A queue called requestqueue used by the client application to send request messages to
BizTalk Server.
A queue called responsequeue used by BizTalk Server to send response messages to the
client application.
A topic called requesttopic used by the client application to send request messages to
BizTalk Server.
A topic called responsetopic used by BizTalk Server to send response messages to the
client application.
A subscription called ItalyMilan for the requesttopic. The latter is used by BizTalk Server to
receive request messages from the requesttopic. The subscription in question has a single
rule defined as follows:
a. Filter: Country='Italy' and City='Milan'
b. Action: Set Area='Western Europe'
Note
A
ap
qu
a
b
The fo

The fo

#regio
using
using
using
#endre

namesp
{
st
{


A subscription
pplication to r
uestion has a
. Filter: Coun
. Action: Se
ollowing figur
ollowing exam
on Using Dir
System;
Microsoft.S
Microsoft.S
egion
pace Microso
tatic class
#region P
//*******
called ItalyM
receive respo
a single rule d
ntry='Italy'
et Area='West
e displays the
mple contains
ectives
erviceBus;
erviceBus.Me
ft.WindowsAz
Program
rivate Const
************
Milan for the r
onse message
defined as foll
and City='Mi
tern Europe'
e output of the
s the code of t
ssaging;
ure.CAT.Samp
ants
********
responsetopi
es from the re
ows:
ilan'
e Provisionin
the Provision
ples.ServiceB
ic. The latter
esponsetopic
ng console a
ning applicati
Bus.Provision
is used by cli
c. The subscr
pplication.
ion:
ning

15
ient
ription in

53

154
// Constants
//***************************

private const string RequestQueue = "requestqueue";
private const string ResponseQueue = "responsequeue";
private const string RequestTopic = "requesttopic";
private const string ResponseTopic = "responsetopic";
private const string RequestSubscription = "ItalyMilan";
private const string ResponseSubscription = "ItalyMilan";
#endregion
static void Main()
{
var defaultColor = Console.ForegroundColor;

try
{
// Set Window Size
Console.WindowWidth = 100;
Console.BufferWidth = 100;
Console.WindowHeight = 48;
Console.BufferHeight = 48;

// Print Header
Console.WriteLine("Read Credentials:");
Console.WriteLine("-----------------");

// Read Service Bus Namespace
Console.ForegroundColor = ConsoleColor.Green;
Console.Write("Service Bus Namespace: ");
Console.ForegroundColor = defaultColor;
var serviceNamespace = Console.ReadLine();

// Read Service Bus Issuer Name
Console.ForegroundColor = ConsoleColor.Green;
Console.Write("Service Bus Issuer Name: ");
Console.ForegroundColor = defaultColor;
var issuerName = Console.ReadLine();

// Read Service Bus Issuer Secret
Console.ForegroundColor = ConsoleColor.Green;
Console.Write("Service Bus Issuer Secret: ");
Console.ForegroundColor = defaultColor;

155
var issuerSecret = Console.ReadLine();

// Print Header
Console.WriteLine();
Console.WriteLine("Enter Queues Properties:");
Console.WriteLine("------------------------");

// Read Queue EnabledBatchedOperations
Console.ForegroundColor = ConsoleColor.Green;
Console.Write("Queues: ");
Console.ForegroundColor = defaultColor;
Console.Write("Set the ");
Console.ForegroundColor = ConsoleColor.Yellow;
Console.Write("EnabledBatchedOperations ");
Console.ForegroundColor = defaultColor;
Console.Write("to true [y=Yes, n=No]?");
var key = Console.ReadKey().KeyChar;
var queueEnabledBatchedOperations = key == 'y' || key == 'Y';
Console.WriteLine();

// Read Queue EnableDeadLetteringOnMessageExpiration
Console.ForegroundColor = ConsoleColor.Green;
Console.Write("Queues: ");
Console.ForegroundColor = defaultColor;
Console.Write("Set the ");
Console.ForegroundColor = ConsoleColor.Yellow;
Console.Write("EnableDeadLetteringOnMessageExpiration ");
Console.ForegroundColor = defaultColor;
Console.Write("to true [y=Yes, n=No]?");
key = Console.ReadKey().KeyChar;
var queueEnableDeadLetteringOnMessageExpiration = key == 'y' || key ==
'Y';
Console.WriteLine();

// Read Queue RequiresDuplicateDetection
Console.ForegroundColor = ConsoleColor.Green;
Console.Write("Queues: ");
Console.ForegroundColor = defaultColor;
Console.Write("Set the ");
Console.ForegroundColor = ConsoleColor.Yellow;
Console.Write("RequiresDuplicateDetection ");
Console.ForegroundColor = defaultColor;

156
Console.Write("to true [y=Yes, n=No]?");
key = Console.ReadKey().KeyChar;
var queueRequiresDuplicateDetection = key == 'y' || key == 'Y';
Console.WriteLine();

// Read Queue RequiresSession
Console.ForegroundColor = ConsoleColor.Green;
Console.Write("Queues: ");
Console.ForegroundColor = defaultColor;
Console.Write("Set the ");
Console.ForegroundColor = ConsoleColor.Yellow;
Console.Write("RequiresSession ");
Console.ForegroundColor = defaultColor;
Console.Write("to true [y=Yes, n=No]?");
key = Console.ReadKey().KeyChar;
var queueRequiresSession = key == 'y' || key == 'Y';
Console.WriteLine();

// Print Header
Console.WriteLine();
Console.WriteLine("Enter Topic Properties:");
Console.WriteLine("-----------------------");

// Read Topic EnabledBatchedOperations
Console.ForegroundColor = ConsoleColor.Green;
Console.Write("Topics: ");
Console.ForegroundColor = defaultColor;
Console.Write("Set the ");
Console.ForegroundColor = ConsoleColor.Yellow;
Console.Write("EnabledBatchedOperations ");
Console.ForegroundColor = defaultColor;
Console.Write("to true [y=Yes, n=No]?");
key = Console.ReadKey().KeyChar;
var topicEnabledBatchedOperations = key == 'y' || key == 'Y';
Console.WriteLine();

// Read Topic RequiresDuplicateDetection
Console.ForegroundColor = ConsoleColor.Green;
Console.Write("Topics: ");
Console.ForegroundColor = defaultColor;
Console.Write("Set the ");
Console.ForegroundColor = ConsoleColor.Yellow;

157
Console.Write("RequiresDuplicateDetection ");
Console.ForegroundColor = defaultColor;
Console.Write("to true [y=Yes, n=No]?");
key = Console.ReadKey().KeyChar;
var topicRequiresDuplicateDetection = key == 'y' || key == 'Y';
Console.WriteLine();

// Print Header
Console.WriteLine();
Console.WriteLine("Enter Subscriptions Properties: ");
Console.WriteLine("-------------------------------");

// Read Subscription EnabledBatchedOperations
Console.ForegroundColor = ConsoleColor.Green;
Console.Write("Subscriptions: ");
Console.ForegroundColor = defaultColor;
Console.Write("Set the ");
Console.ForegroundColor = ConsoleColor.Yellow;
Console.Write("EnabledBatchedOperations ");
Console.ForegroundColor = defaultColor;
Console.Write("to true [y=Yes, n=No]?");
key = Console.ReadKey().KeyChar;
var subscriptionnabledBatchedOperations = key == 'y' || key == 'Y';
Console.WriteLine();

// Read Subscription EnableDeadLetteringOnFilterEvaluationExceptions
Console.ForegroundColor = ConsoleColor.Green;
Console.Write("Subscriptions: ");
Console.ForegroundColor = defaultColor;
Console.Write("Set the ");
Console.ForegroundColor = ConsoleColor.Yellow;
Console.Write("EnableDeadLetteringOnFilterEvaluationExceptions ");
Console.ForegroundColor = defaultColor;
Console.Write("to true [y=Yes, n=No]?");
key = Console.ReadKey().KeyChar;
var subscriptionEnableDeadLetteringOnFilterEvaluationExceptions = key ==
'y' || key == 'Y';
Console.WriteLine();

// Read Subscription EnableDeadLetteringOnMessageExpiration
Console.ForegroundColor = ConsoleColor.Green;
Console.Write("Subscriptions: ");

158
Console.ForegroundColor = defaultColor;
Console.Write("Set the ");
Console.ForegroundColor = ConsoleColor.Yellow;
Console.Write("EnableDeadLetteringOnMessageExpiration ");
Console.ForegroundColor = defaultColor;
Console.Write("to true [y=Yes, n=No]?");
key = Console.ReadKey().KeyChar;
var subscriptionEnableDeadLetteringOnMessageExpiration = key == 'y' ||
key == 'Y';
Console.WriteLine();

// Read Subscription RequiresSession
Console.ForegroundColor = ConsoleColor.Green;
Console.Write("Subscriptions: ");
Console.ForegroundColor = defaultColor;
Console.Write("Set the ");
Console.ForegroundColor = ConsoleColor.Yellow;
Console.Write("RequiresSession ");
Console.ForegroundColor = defaultColor;
Console.Write("to true [y=Yes, n=No]?");
key = Console.ReadKey().KeyChar;
var subscriptionRequiresSession = key == 'y' || key == 'Y';
Console.WriteLine();

// Get ServiceBusNamespaceClient for management operations
var managementUri =
ServiceBusEnvironment.CreateServiceUri("https", serviceNamespace,
string.Empty);
var tokenProvider =
TokenProvider.CreateSharedSecretTokenProvider(issuerName,
issuerSecret);
var namespaceManager = new NamespaceManager(managementUri,
tokenProvider);

// Print Header
Console.WriteLine();
Console.WriteLine("Create Queues:");
Console.WriteLine("--------------");

// Create RequestQueue
Console.Write("Creating ");
Console.ForegroundColor = ConsoleColor.Yellow;

159
Console.Write(RequestQueue);
Console.ForegroundColor = defaultColor;
Console.WriteLine(" queue...");
if (namespaceManager.QueueExists(RequestQueue))
{
namespaceManager.DeleteQueue(RequestQueue);
}
namespaceManager.CreateQueue(new QueueDescription(RequestQueue)
{
EnableBatchedOperations =
queueEnabledBatchedOperations,
EnableDeadLetteringOnMessageExpiration =
queueEnableDeadLetteringOnMessageExpiration,
RequiresDuplicateDetection =
queueRequiresDuplicateDetection,
RequiresSession = queueRequiresSession
});
Console.ForegroundColor = ConsoleColor.Yellow;
Console.Write(RequestQueue);
Console.ForegroundColor = defaultColor;
Console.WriteLine(" queue successfully created.");

// Create ResponseQueue
Console.Write("Creating ");
Console.ForegroundColor = ConsoleColor.Yellow;
Console.Write(ResponseQueue);
Console.ForegroundColor = defaultColor;
Console.WriteLine(" queue...");
if (namespaceManager.QueueExists(ResponseQueue))
{
namespaceManager.DeleteQueue(ResponseQueue);
}
namespaceManager.CreateQueue(new QueueDescription(ResponseQueue)
{
EnableBatchedOperations = queueEnabledBatchedOperations,
EnableDeadLetteringOnMessageExpiration =
queueEnableDeadLetteringOnMessageExpiration,
RequiresDuplicateDetection =
queueRequiresDuplicateDetection,
RequiresSession = queueRequiresSession
});
Console.ForegroundColor = ConsoleColor.Yellow;

160
Console.Write(ResponseQueue);
Console.ForegroundColor = defaultColor;
Console.WriteLine(" queue successfully created.");

// Print Header
Console.WriteLine();
Console.WriteLine("Create Topics:");
Console.WriteLine("--------------");

// Create RequestTopic
Console.Write("Creating ");
Console.ForegroundColor = ConsoleColor.Yellow;
Console.Write(RequestTopic);
Console.ForegroundColor = defaultColor;
Console.WriteLine(" topic...");
if (namespaceManager.TopicExists(RequestTopic))
{
namespaceManager.DeleteTopic(RequestTopic);
}
namespaceManager.CreateTopic(new TopicDescription(RequestTopic)
{
EnableBatchedOperations =
topicEnabledBatchedOperations,
RequiresDuplicateDetection =
topicRequiresDuplicateDetection
});
Console.ForegroundColor = ConsoleColor.Yellow;
Console.Write(RequestTopic);
Console.ForegroundColor = defaultColor;
Console.WriteLine(" topic successfully created.");

// Create ResponseTopic
Console.Write("Creating ");
Console.ForegroundColor = ConsoleColor.Yellow;
Console.Write(ResponseTopic);
Console.ForegroundColor = defaultColor;
Console.WriteLine(" topic...");
if (namespaceManager.TopicExists(ResponseTopic))
{
namespaceManager.DeleteTopic(ResponseTopic);
}
namespaceManager.CreateTopic(new TopicDescription(ResponseTopic)

161
{
EnableBatchedOperations =
topicEnabledBatchedOperations,
RequiresDuplicateDetection =
topicRequiresDuplicateDetection
});
Console.ForegroundColor = ConsoleColor.Yellow;
Console.Write(ResponseTopic);
Console.ForegroundColor = defaultColor;
Console.WriteLine(" topic successfully created.");

// Print Header
Console.WriteLine();
Console.WriteLine("Create Subscriptions:");
Console.WriteLine("--------------");

// Create Request Subscription
Console.Write("Creating ");
Console.ForegroundColor = ConsoleColor.Yellow;
Console.Write(RequestSubscription);
Console.ForegroundColor = defaultColor;
Console.Write(" subscription for the ");
Console.ForegroundColor = ConsoleColor.Yellow;
Console.Write(RequestTopic);
Console.ForegroundColor = defaultColor;
Console.WriteLine(" topic...");
var ruleDescription = new RuleDescription(new SqlFilter("Country='Italy'
and City='Milan'"))
{
Name = "$Default",
Action = new SqlRuleAction("Set
Area='Western Europe'")
};
var subscriptionDescription = new SubscriptionDescription(RequestTopic,
RequestSubscription)
{
EnableBatchedOperations = subscriptionnabledBatchedOperations,
EnableDeadLetteringOnFilterEvaluationExceptions =
subscriptionEnableDeadLetteringOnFilterEvaluationExceptions,
EnableDeadLetteringOnMessageExpiration =
subscriptionEnableDeadLetteringOnMessageExpiration,
RequiresSession = subscriptionRequiresSession

162
};
namespaceManager.CreateSubscription(subscriptionDescription,
ruleDescription);
Console.ForegroundColor = ConsoleColor.Yellow;
Console.Write(RequestSubscription);
Console.ForegroundColor = defaultColor;
Console.Write(" subscription for the ");
Console.ForegroundColor = ConsoleColor.Yellow;
Console.Write(RequestTopic);
Console.ForegroundColor = defaultColor;
Console.WriteLine(" topic successfully created.");

// Create Response Subscription
Console.Write("Creating ");
Console.ForegroundColor = ConsoleColor.Yellow;
Console.Write(ResponseSubscription);
Console.ForegroundColor = defaultColor;
Console.Write(" subscription for the ");
Console.ForegroundColor = ConsoleColor.Yellow;
Console.Write(ResponseTopic);
Console.ForegroundColor = defaultColor;
Console.WriteLine(" topic...");
ruleDescription = new RuleDescription(new SqlFilter("Country='Italy' and
City='Milan'"))
{
Action = new SqlRuleAction("Set Area='Western
Europe'")
};
subscriptionDescription = new SubscriptionDescription(ResponseTopic,
ResponseSubscription)
{
EnableBatchedOperations = subscriptionnabledBatchedOperations,
EnableDeadLetteringOnFilterEvaluationExceptions =
subscriptionEnableDeadLetteringOnFilterEvaluationExceptions,
EnableDeadLetteringOnMessageExpiration =
subscriptionEnableDeadLetteringOnMessageExpiration,
RequiresSession = subscriptionRequiresSession
};
namespaceManager.CreateSubscription(subscriptionDescription,
ruleDescription);
Console.ForegroundColor = ConsoleColor.Yellow;
Console.Write(ResponseSubscription);

163
Console.ForegroundColor = defaultColor;
Console.Write(" subscription for the ");
Console.ForegroundColor = ConsoleColor.Yellow;
Console.Write(ResponseTopic);
Console.ForegroundColor = defaultColor;
Console.WriteLine(" topic successfully created.");
Console.WriteLine();

// Close the application
Console.WriteLine("Press any key to continue ...");
Console.ReadLine();
}
catch (Exception ex)
{
Console.ForegroundColor = ConsoleColor.Yellow;
Console.Write("Exception: ");
Console.ForegroundColor = defaultColor;
Console.Write(ex.Message);
}
}
}
}
I did not test all the possible combinations of properties for queues and topics so the
demo may not work as expected with all the configurations.
Data and Message Contracts
I started defining the data and message contracts for the request and response messages. These
two types of contracts have different roles in WCF:
1. Data contracts provide a mechanism to map .NET CLR types that are defined in code and
XML Schemas (XSD) defined by the W3C organization (www.w3c.org). Data contracts are
published in the services metadata, allowing clients to convert the neutral, technology-
agnostic representation of the data types to their native representations.
2. Message contracts describe the structure of SOAP messages sent to and from a service
and enable you to inspect and control most of the details in the SOAP header and body.
Whereas data contracts enable interoperability through the XML Schema Definition (XSD)
standard, message contracts enable you to interoperate with any system that communicates
through SOAP. Using message contracts gives you complete control over the SOAP
message sent to and from a service by providing access to the SOAP headers and bodies
directly. This allows the use of simple or complex types to define the exact content of the
SOAP parts.
Note

164
In my solution, I created two separate projects called DataContracts and MessageContracts
respectively. For your convenience, I included below the code of the classes used to define the
Data and Message Contract of the request and response messages.
CalculatorRequest Class
[Serializable]
[XmlType(TypeName = "CalculatorRequest",
Namespace = "http://windowsazure.cat.microsoft.com/samples/servicebus")]
[XmlRoot(ElementName = "CalculatorRequest",
Namespace = http://windowsazure.cat.microsoft.com/samples/servicebus,
IsNullable = false)]
[DataContract(Name = "CalculatorRequest",
Namespace = "http://windowsazure.cat.microsoft.com/samples/servicebus")]
public class CalculatorRequest
{
#region Private Fields
private OperationList operationList;
#endregion

#region Public Constructors
public CalculatorRequest()
{
operationList = new OperationList();
}

public CalculatorRequest(OperationList operationList)
{
this.operationList = operationList;
}
#endregion

#region Public Properties
[XmlArrayItem("Operation", Type=typeof(Operation), IsNullable = false)]
[DataMember(Order = 1)]
public OperationList Operations
{
get
{
return operationList;
}
set
{
operationList = value;

165
}
}
#endregion
}

[CollectionDataContract(Name = "OperationList",
Namespace =
http://windowsazure.cat.microsoft.com/samples/servicebus,
ItemName = "Operation")]
public class OperationList : List<Operation>
{
}

[Serializable]
[XmlType(TypeName = "Operation",
AnonymousType = true,
Namespace = "http://windowsazure.cat.microsoft.com/samples/servicebus")]
[DataContract(Name = "Operation",
Namespace = "http://windowsazure.cat.microsoft.com/samples/servicebus")]
public class Operation
{
#region Private Fields
private string op;
private double operand1;
private double operand2;
#endregion

#region Public Constructors
public Operation()
{
}

public Operation(string op,
double operand1,
double operand2)
{
this.op = op;
this.operand1 = operand1;
this.operand2 = operand2;
}
#endregion


166
#region Public Properties
[XmlElement]
[DataMember(Order = 1)]
public string Operator
{
get
{
return op;
}
set
{
op = value;
}
}

[XmlElement]
[DataMember(Order = 2)]
public double Operand1
{
get
{
return operand1;
}
set
{
operand1 = value;
}
}

[XmlElement]
[DataMember(Order = 3)]
public double Operand2
{
get
{
return operand2;
}
set
{
operand2 = value;
}
}

167
#endregion
}

CalculatorResponse Class

[Serializable]
[XmlType(TypeName = "CalculatorResponse",
Namespace = "http://windowsazure.cat.microsoft.com/samples/servicebus")]
[XmlRoot(ElementName = "CalculatorResponse",
Namespace = http://windowsazure.cat.microsoft.com/samples/servicebus,
IsNullable = false)]
[DataContract(Name = "CalculatorResponse",
Namespace = "http://windowsazure.cat.microsoft.com/samples/servicebus")]
public class CalculatorResponse
{
#region Private Fields
private string status;
private ResultList resultList;
#endregion

#region Public Constructors
public CalculatorResponse()
{
status = default(string);
resultList = new ResultList();
}

public CalculatorResponse(string status)
{
this.status = status;
resultList = new ResultList();
}

public CalculatorResponse(string status, ResultList resultList)
{
this.status = status;
this.resultList = resultList;
}
#endregion

#region Public Properties

168
[XmlElement]
[DataMember(Order = 1)]
public string Status
{
get
{
return status;
}
set
{
status = value;
}
}

[XmlArrayItem("Result", Type=typeof(Result), IsNullable=false)]
[DataMember(Order = 2)]
public ResultList Results
{
get
{
return resultList;
}
set
{
resultList = value;
}
}
#endregion
}

[CollectionDataContract(Name = "ResultList",
Namespace =
http://windowsazure.cat.microsoft.com/samples/servicebus,
ItemName = "Result")]
public class ResultList : List<Result>
{
}

[Serializable]
[XmlType(TypeName = "Result",
AnonymousType = true,
Namespace = "http://windowsazure.cat.microsoft.com/samples/servicebus")]

169
[DataContract(Name = "Result",
Namespace = "http://windowsazure.cat.microsoft.com/samples/servicebus")]
public class Result
{
#region Private Fields
private double value;
private string error;
#endregion

#region Public Constructors
public Result()
{
value = default(double);
error = default(string);
}

public Result(double value, string error)
{
this.value = value;
this.error = error;
}
#endregion

#region Public Properties
[XmlElement]
[DataMember(Order = 1)]
public double Value
{
get
{
return value;
}
set
{
this.value = value;
}
}

[XmlElement]
[DataMember(Order = 2)]
public string Error
{

170
get
{
return error;
}
set
{
error = value;
}
}
#endregion
}
CalculatorRequestMessage Class
[MessageContract(IsWrapped=false)]
public class CalculatorRequestMessage
{
#region Private Fields
private CalculatorRequest calculatorRequest;
#endregion

#region Public Constructors
public CalculatorRequestMessage()
{
this.calculatorRequest = null;
}

public CalculatorRequestMessage(CalculatorRequest calculatorRequest)
{
this.calculatorRequest = calculatorRequest;
}
#endregion

#region Public Properties
[MessageBodyMember(Namespace =
"http://windowsazure.cat.microsoft.com/samples/servicebus")]
public CalculatorRequest CalculatorRequest
{
get
{
return this.calculatorRequest;
}
set
{

171
this.calculatorRequest = value;
}
}
#endregion
}
CalculatorResponseMessage Class
[MessageContract(IsWrapped = false)]
public class CalculatorResponseMessage
{
#region Private Fields
private CalculatorResponse calculatorResponse;
#endregion

#region Public Constructors
public CalculatorResponseMessage()
{
this.calculatorResponse = null;
}

public CalculatorResponseMessage(CalculatorResponse calculatorResponse)
{
this.calculatorResponse = calculatorResponse;
}
#endregion

#region Public Properties
[MessageBodyMember(Namespace =
"http://windowsazure.cat.microsoft.com/samples/servicebus")]
public CalculatorResponse CalculatorResponse
{
get
{
return this.calculatorResponse;
}
set
{
this.calculatorResponse = value;
}
}
#endregion
}

172
Indeed, I could have used just data contracts to model messages as message contracts add a
degree of complexity. However, by assigning false to the IsWrapped property exposed by the
MessageContractAttribute, you specify that the message body wont be contained in a wrapper
element. Typically, the wrapper element of a request message is the name of the operation
invoked and its defined in the WSDL. Setting the value of the IsWrapped property to false, you
can simply select the Body option in both the Inbound BizTalk message body and Outbound
WCF message body sections on the Messages tab when configuring a WCF receive location.
Otherwise you should define a Path in the Inbound BizTalk message body section to extract
the payload from the inbound message, and specify a template in the Outbound WCF message
body section to include the outgoing response message within a wrapper element.



173
The namespace of the message and data contract classes match those defined by the
XML schemas that model the request and response messages in the BizTalk Server
application.




See Also
"Using Data Contracts" topic on MSDN.
Using Message Contracts topic on MSDN
Service Contracts
The next step was to define the ServiceContracts used by the client application to exchange
messages with the Service Bus. To this purpose, I created a new project in my solution called
ServiceContracts, and I defined two service contract interfaces used by the client application
respectively to send and receive messages from the Service Bus messaging entities. Indeed, I
created two different versions of the service contract used to receive response messages:
The ICalculatorResponse interface is meant to be used to receive response messages from
a non-sessionful queue or subscription
The ICalculatorResponseSessionful interface inherits from the ICalculatorResponse
service contract and is marked with the [ServiceContract(SessionMode =
SessionMode.Required)] attribute. This service contract is meant to be used to receive
response messages from a non-sessionful queue or subscription.
Note that the methods defined by all the contracts must be one-way.
ICalculatorRequest, ICalculatorResponse, ICalculatorResponseSessionful Interfaces

#region Using Directives
using System.ServiceModel;
using Microsoft.WindowsAzure.CAT.Samples.ServiceBus.MessageContracts;
#endregion

namespace Microsoft.WindowsAzure.CAT.Samples.ServiceBus.ServiceContracts
{
[ServiceContract(Namespace =
"http://windowsazure.cat.microsoft.com/samples/servicebus",
SessionMode = SessionMode.Allowed)]
public interface ICalculatorRequest
Note

174
{
[OperationContract(Action = "SendRequest", IsOneWay = true)]
[ReceiveContextEnabled(ManualControl = true)]
void SendRequest(CalculatorRequestMessage calculatorRequestMessage);
}

[ServiceContract(Namespace =
"http://windowsazure.cat.microsoft.com/samples/servicebus",
SessionMode = SessionMode.Allowed)]
public interface ICalculatorResponse
{
[OperationContract(Action = "ReceiveResponse", IsOneWay = true)]
[ReceiveContextEnabled(ManualControl = true)]
void ReceiveResponse(CalculatorResponseMessage calculatorResponseMessage);
}

[ServiceContract(Namespace =
"http://windowsazure.cat.microsoft.com/samples/servicebus",
SessionMode = SessionMode.Required)]
public interface ICalculatorResponseSessionful : ICalculatorResponse
{
}
}
We are now ready to look at the code of the client application.




Client Application
Since the Windows Forms application exchanges messages with the ServiceBusSample BizTalk
application asynchronously through Service Bus messaging entities, it acts as a client and a
service application at the same time. Windows Forms uses WCF and the NetMessagingBinding
to perform the following actions:
1. Send request messages to the requestqueue.
2. Send request messages to the requesttopic.
3. Receive response messages from the responsequeue.
4. Receive response messages from the ItalyMilan subscription of the responsetopic.

175
Lets start reviewing the configuration file of the client application that plays a central role in the
definition of the WCF client and service endpoints used to communicate with the Service Bus.
App.Config
1: <?xml version="1.0"?>
2: <configuration>
3: <system.diagnostics>
4: <sources>
5: <source name="System.ServiceModel.MessageLogging" switchValue="Warning,
ActivityTracing">
6: <listeners>
7: <add type="System.Diagnostics.DefaultTraceListener" name="Default">
8: <filter type="" />
9: </add>
10: <add name="ServiceModelMessageLoggingListener">
11: <filter type="" />
12: </add>
13: </listeners>
14: </source>
15: </sources>
16: <sharedListeners>
17: <add initializeData="C:\ServiceBusQueueClient.svclog"
18: type="System.Diagnostics.XmlWriterTraceListener, System,
Version=2.0.0.0, Culture=neutral,
PublicKeyToken=b77a5c561934e089"
19: name="ServiceModelMessageLoggingListener"
20: traceOutputOptions="Timestamp">
21: <filter type="" />
22: </add>
23: </sharedListeners>
24: <trace autoflush="true" indentsize="4">
25: <listeners>
26: <clear/>
27: <add name="LogTraceListener"
28:
type="Microsoft.WindowsAzure.CAT.Samples.ServiceBus.Client.LogTraceListener, Client"
29: initializeData="" />
30: </listeners>
31: </trace>
32: </system.diagnostics>
33: <startup>
34: <supportedRuntime version="v4.0" sku=".NETFramework,Version=v4.0"/>
35: </startup>

176
36: <system.serviceModel>
37: <diagnostics>
38: <messageLogging logEntireMessage="true"
39: logMalformedMessages="false"
40: logMessagesAtServiceLevel="true"
41: logMessagesAtTransportLevel="false" />
42: </diagnostics>
43: <behaviors>
44: <endpointBehaviors>
45: <behavior name="securityBehavior">
46: <transportClientEndpointBehavior>
47: <tokenProvider>
48: <sharedSecret issuerName="owner"
49: issuerSecret="[ISSUER SECRET RETRIEVED FROM THE AZURE
MGMT PORTAL]" />
50: </tokenProvider>
51: </transportClientEndpointBehavior>
52: </behavior>
53: </endpointBehaviors>
54: </behaviors>
55: <bindings>
56: <basicHttpBinding>
57: <binding name="basicHttpBinding"
58: closeTimeout="00:10:00"
59: openTimeout="00:10:00"
60: receiveTimeout="00:10:00"
61: sendTimeout="00:10:00">
62: <security mode="None">
63: <transport clientCredentialType="None" proxyCredentialType="None"
64: realm="" />
65: <message clientCredentialType="UserName" algorithmSuite="Default" />
66: </security>
67: </binding>
68: </basicHttpBinding>
69: <basicHttpRelayBinding>
70: <binding name="basicHttpRelayBinding"
71: closeTimeout="00:10:00"
72: openTimeout="00:10:00"
73: receiveTimeout="00:10:00"
74: sendTimeout="00:10:00">
75: <security mode="Transport"
relayClientAuthenticationType="RelayAccessToken" />

177
76: </binding>
77: </basicHttpRelayBinding>
78: <netTcpRelayBinding>
79: <binding name="netTcpRelayBinding"
80: closeTimeout="00:10:00"
81: openTimeout="00:10:00"
82: receiveTimeout="00:10:00"
83: sendTimeout="00:10:00">
84: <security mode="Transport"
relayClientAuthenticationType="RelayAccessToken" />
85: </binding>
86: </netTcpRelayBinding>
87: <netMessagingBinding>
88: <binding name="netMessagingBinding"
89: sendTimeout="00:03:00"
90: receiveTimeout="00:03:00"
91: openTimeout="00:03:00"
92: closeTimeout="00:03:00"
93: sessionIdleTimeout="00:01:00"
94: prefetchCount="-1">
95: <transportSettings batchFlushInterval="00:00:01" />
96: </binding>
97: </netMessagingBinding>
98: </bindings>
99: <client>
100: <!-- Invoke BizTalk via Service Bus Queue -->
101: <endpoint address="sb://paolosalvatori.servicebus.windows.net/requestqueue"
102: behaviorConfiguration="securityBehavior"
103: binding="netMessagingBinding"
104: bindingConfiguration="netMessagingBinding"
105:
contract="Microsoft.WindowsAzure.CAT.Samples.ServiceBus.ServiceContracts.ICalculatorReque
st"
106: name="requestQueueClientEndpoint" />
107: <!-- Invoke BizTalk via Service Bus Topic -->
108: <endpoint address="sb://paolosalvatori.servicebus.windows.net/requesttopic"
109: behaviorConfiguration="securityBehavior"
110: binding="netMessagingBinding"
111: bindingConfiguration="netMessagingBinding"
112:
contract="Microsoft.WindowsAzure.CAT.Samples.ServiceBus.ServiceContracts.ICalculatorReque
st"

178
113: name="BizTalkServiceBusTopic" />
114: <!-- Invoke BizTalk via Service Bus Relay Service -->
115: <endpoint
address="sb://paolosalvatori.servicebus.windows.net/nettcp/calculatorservice"
116: behaviorConfiguration="securityBehavior"
117: binding="netTcpRelayBinding"
118: bindingConfiguration="netTcpRelayBinding"
119:
contract="Microsoft.WindowsAzure.CAT.Samples.ServiceBus.ServiceContracts.ICalculatorServi
ce"
120: name="netTcpRelayBindingClientEndpoint" />
121: <!-- Invoke BizTalk directly via WCF Receive Location -->
122: <endpoint
address="http://localhost/newcalculatorservice/calculatorservice.svc"
123: binding="basicHttpBinding"
124: bindingConfiguration="basicHttpBinding"
125:
contract="Microsoft.WindowsAzure.CAT.Samples.ServiceBus.ServiceContracts.ICalculatorServi
ce"
126: name="basicHttpBindingClientEndpoint" />
127: </client>
128: <services>
129: <service
name="Microsoft.WindowsAzure.CAT.Samples.ServiceBus.Service.ResponseHandlerService">
130: <endpoint
address="sb://paolosalvatori.servicebus.windows.net/responsequeue"
131: behaviorConfiguration="securityBehavior"
132: binding="netMessagingBinding"
133: bindingConfiguration="netMessagingBinding"
134: name="responseQueueServiceEndpoint"
135:
contract="Microsoft.WindowsAzure.CAT.Samples.ServiceBus.ServiceContracts.ICalculatorRespo
nse" />
136: <endpoint
address="sb://paolosalvatori.servicebus.windows.net/responsetopic"
137:
listenUri="sb://paolosalvatori.servicebus.windows.net/responsetopic/Subscriptions/ItalyMi
lan"
138: behaviorConfiguration="securityBehavior"
139: binding="netMessagingBinding"
140: bindingConfiguration="netMessagingBinding"
141: name="responseSubscriptionServiceEndpoint"

179
142:
contract="Microsoft.WindowsAzure.CAT.Samples.ServiceBus.ServiceContracts.ICalculatorRespo
nse" />
143: </service>
144: </services>
145: </system.serviceModel>
146: </configuration>
Please find below a brief description of the main elements and sections of the configuration file:
Lines [3-32] define a custom trace listener called LogTraceListener used by the
ResponseHandlerService to write response message to the log control of the Windows
Forms application.
Lines [33-35] the startup section specifies which versions of the common language runtime
the application supports.
Lines [45-52] contain the definition of the securityBehavior used by client and service
endpoint to authenticate with the Access Control Service. In particular, the
TransportClientEndpointBehavior is used to define shared secret credentials. For more
information on how to retrieve credentials from the Windows Azure Management Portal,
see the box below.
Lines [87-97] contain the configuration of the NetMessagingBinding used client and service
endpoints to exchange messages with the Service Bus.
Lines [101-106] contain the definition of the requestQueueClientEndpoint used by the
application to send request messages to the requestqueue. The address of the client
endpoint is given by the concatenation of the URL of the service namespace and the name of
the queue.
Lines [108-113] contain the definition of the requestTopicClientEndpoint used by the
application to send request messages to the requesttopic. The address of the client
endpoint is given by the concatenation of the URL of the service namespace and the name of
the topic
Lines [130-135] contain the definition of the responseQueueServiceEndpoint used by the
application to receive response messages from the responsequeue. The address of the
service endpoint is given by the concatenation of the URL of the service namespace and the
name of the queue.
Lines [108-113] contain the definition of the responseSubscriptionServiceEndpoint used
by the application to send receive response messages from the ItalyMilan subscription for
the responsetopic. When you define a WCF service endpoint that uses the
NetMessagingBinding to receive messages from a subscription, you have to proceed as
follows (for more information on this, see the box below):
As value of the address attribute, specify the URL of the topic which the subscription
belongs to. The URL of the topic is given by the concatenation of the URL of the service
namespace and the name of the topic.
As value of the listenUri attribute, specify the URL of the subscription. The URL of the
subscription is defined by the concatenation of the topic URL, the string /Subscriptions/,
and the name of the subscription.
Assign the value Explicit to the listenUriMode attribute. The default value for the
listenUriMode is Explicit, so this setting is optional.
W
qu
sa
re
se
se
F
T
S
on
ot
A
be

This o
clickin
N
N
When you con
ueue or subs
ample, when
esponseSub
essionful que
ervice contrac
or more inform
he Service Bu
ecret, and Si
nly Shared S
ther credentia
Azure Manage
elow after sel
opens up the
ng the Copy t
Note
Note
figure a WCF
cription, the s
you configure
scriptionSer
ue and subsc
ct with the se
mation, see th
us supports th
imple Web T
Secret creden
al schemes. Y
ement Portal
lecting a certa
modal dialog
to Clipboard
F service endp
service contra
e the respons
rviceEndpoin
cription, you h
ssionful ICalc
he Service Co
hree different
Token, but this
ntials. Howeve
You can retrie
l by clicking th
ain namespac
shown in the
button highli
point to consu
act needs to s
seQueueSer
nt endpoints t
have to replac
culatorRespo
ontracts secti
t types of cred
s version of th
er, you can ea
eve the issuer
he View butto
ce in the Serv
e picture below
ghted in red.
ume message
support sessio
rviceEndpoin
to receive, res
ce the ICalcu
onseSession
ion later in the
dential schem
he Service B
asily extend m
r-secret key fr
on highlighted
vice Bus sec
w where you
es from a ses
ons. Therefor
nt or
spectively, fro
ulatorRespon
nful contract i
e article.
mes: SAML, S
Bus Explorer
my code to su
rom the Wind
d in red in the
ction.
can retrieve t

18
ssionful
re, in our
om a
nse
interface.
Shared
supports
upport
dows
e picture

the key by
80

181


By convention, the name of the Default Issuer is always owner.
When you define a WCF service endpoint that uses the NetMessagingBinding to receive
messages from a subscription, if you make the mistake to assign the URL of the
subscription to the address attribute of the service endpoint (as reported in configuration
below), at runtime an FaultException like the following will occur:

The message with To 'sb://paolosalvatori.servicebus.windows.net/responsetopic' cannot be
processed at
the receiver, due to an AddressFilter mismatch at the EndpointDispatcher.
Check that the sender and receiver's EndpointAddresses agree."}

Wrong Configuration
<?xml version="1.0"?>
<configuration>
...
<system.serviceModel>
...
<services>
<service
name="Microsoft.WindowsAzure.CAT.Samples.ServiceBus.Service.ResponseHandlerService">
<endpoint
address="sb://paolosalvatori.servicebus.windows.net/responsetopic/Subscriptions/ItalyMila
n"
behaviorConfiguration="securityBehavior"
binding="netMessagingBinding"
bindingConfiguration="netMessagingBinding"
Note
Note

182
name="responseSubscriptionServiceEndpoint"

contract="Microsoft.WindowsAzure.CAT.Samples.ServiceBus.ServiceContracts.ICalculatorRespo
nse" />
</service>
</services>
</system.serviceModel>
</configuration>

The error is due to the fact the WS-Addressing To header of the message contains the address of
the topic and not the address of the subscription:
<:Envelope xmlns:s="http://www.w3.org/2003/05/soap-envelope"
xmlns:a="http://www.w3.org/2005/08/addressing">
<s:Header>
<a:Action s:mustUnderstand="1">ReceiveResponse</a:Action>
<a:MessageID>urn:uuid:64cb0b06-1622-4920-a035-c27b610cfcaf</a:MessageID>
<a:To
s:mustUnderstand="1">sb://paolosalvatori.servicebus.windows.net/responsetopic</a:To>
</s:Header>
<s:Body>... stream ...</s:Body>
</s:Envelope>

To correctly configure the service endpoint to receive messages from a subscription, you have to
proceed as follows:
As value of the address attribute, you have specify the URL of the topic which the
subscription belongs to. The URL of the topic is given by the concatenation of the URL of the
service namespace and the name of the topic.
As value of the listenUri attribute, you have to specify the URL of the subscription. The URL
of the subscription is defined by the concatenation of the topic URL, the string
/Subscriptions/ and the name of the subscription.
Assign the value Explicit to the listenUriMode attribute. The default value for the
listenUriMode is Explicit, so this setting is optional.
See the following page on MSDN for a description of the address, listenUri and listenUriMode
attributes.
Correct Configuration
<?xml version="1.0"?>
<configuration>
...
<system.serviceModel>
...
<services>

183
<service
name="Microsoft.WindowsAzure.CAT.Samples.ServiceBus.Service.ResponseHandlerService">
<endpoint address="sb://paolosalvatori.servicebus.windows.net/responsetopic"

listenUri="sb://paolosalvatori.servicebus.windows.net/responsetopic/Subscriptions/ItalyMi
lan"
behaviorConfiguration="securityBehavior"
binding="netMessagingBinding"
bindingConfiguration="netMessagingBinding"
name="subscriptionEndpoint"

contract="Microsoft.WindowsAzure.CAT.Samples.ServiceBus.ServiceContracts.ICalculatorRespo
nse" />
</service>
</services>
</system.serviceModel>
</configuration>

To accomplish the same task via API, you have to properly set the value of the Address, ListenUri
and ListenUriMode properties of your ServiceEndpoint instance as indicated in this note.
The following example shows the code used by the client application to start the
ResponseHandlerService used to read response messages from the responsequeue and the
ItalyMilan subscription of the responsetopic. Well examine the code of the service in the next
section.
StartServiceHost Method
private void StartServiceHost()
{
try
{
// Creating the service host object as defined in config
var serviceHost = new ServiceHost(typeof(ResponseHandlerService));

// Add ErrorServiceBehavior for handling errors encounter by servicehost during
execution.
serviceHost.Description.Behaviors.Add(new ErrorServiceBehavior());


foreach (var serviceEndpoint in serviceHost.Description.Endpoints)
{
if (serviceEndpoint.Name == "responseQueueServiceEndpoint")
{
responseQueueUri = serviceEndpoint.Address.Uri.AbsoluteUri;

184
WriteToLog(string.Format(ServiceHostListeningOnQueue,
serviceEndpoint.Address.Uri.AbsoluteUri));
}
if (serviceEndpoint.Name == "responseSubscriptionServiceEndpoint")
{
responseTopicUri = serviceEndpoint.Address.Uri.AbsoluteUri;
WriteToLog(string.Format(ServiceHostListeningOnSubscription,
serviceEndpoint.ListenUri.AbsoluteUri));
}
}

// Start the service host
serviceHost.Open();
WriteToLog(ServiceHostSuccessfullyOpened);
}
catch (Exception ex)
{
mainForm.HandleException(ex);
}
}
The following picture shows the user interface of the client application.

The ra
the re
conta
from t
comm
Broke
respo
of the
to the
Send
privat
{
tr
{

adio buttons c
equest messa
ined in the Re
the response
municate the s
eredMessage
onseTopicUr
e method used
e code to facil
RequestMes
te void Send
ry
if (strin
contained in t
age to the req
esponse Met
equeue or fro
selection to th
Property obje
ri private field
d by the clien
itate its under
ssageUsingW
RequestMessa
g.IsNullOrEm
the Request
questqueue o
thod group a
om the ItalyM
he underlying
ect to assign t
s to the Reply
t application.
rstanding.
WCF Method
geUsingWCF(s
mpty(endpoint
Method grou
or the reques
llows you to s
ilan subscript
BizTalk appl
the value of th
yTo property.
For your con
string endpoi
tConfiguratio
up allow you t
ttopic, where
select whethe
tion of the res
ication, the ap
he responseQ
. The followin
nvenience, co
intConfigurat
onName))
o choose whe
eas the radio
er to receive t
sponsetopic
pplication use
QueueUri or
g table conta
mments have
tionName)

18

ether to send
buttons
he response
c. To
es a
ins the code
e been added
85

d

186
{
WriteToLog(EndpointConfigurationNameCannotBeNull);
return;
}

// Set the wait cursor
Cursor = Cursors.WaitCursor;

// Make sure that the request message contains at least an operation
if (operationList == null ||
operationList.Count == 0)
{
WriteToLog(OperationListCannotBeNull);
return;
}

// Create warning collection
var warningCollection = new List<string>();

// Create request message
var calculatorRequest = new CalculatorRequest(operationList);
var calculatorRequestMessage = new CalculatorRequestMessage(calculatorRequest);

// Create the channel factory for the currennt client endpoint
// and cache it in the channelFactoryDictionary
if (!channelFactoryDictionary.ContainsKey(endpointConfigurationName))
{
channelFactoryDictionary[endpointConfigurationName] =
new ChannelFactory<ICalculatorRequest>(endpointConfigurationName);
}

// Create the channel for the currennt client endpoint
// and cache it in the channelDictionary
if (!channelDictionary.ContainsKey(endpointConfigurationName))
{
channelDictionary[endpointConfigurationName] =
channelFactoryDictionary[endpointConfigurationName].CreateChannel();
}

// Use the OperationContextScope to create a block within which to access the
current OperationScope

187
using (new
OperationContextScope((IContextChannel)channelDictionary[endpointConfigurationName]))
{
// Create a new BrokeredMessageProperty object
var brokeredMessageProperty = new BrokeredMessageProperty();

// Read the user defined properties and add them to the
// Properties collection of the BrokeredMessageProperty object
foreach (var e in propertiesBindingSource.Cast<PropertyInfo>())
{
try
{
e.Key = e.Key.Trim();
if (e.Type != StringType && e.Value == null)
{
warningCollection.Add(string.Format(CultureInfo.CurrentUICulture,
PropertyValueCannotBeNull,
e.Key));
}
else
{
if (brokeredMessageProperty.Properties.ContainsKey(e.Key))
{
brokeredMessageProperty.Properties[e.Key] =
ConversionHelper.MapStringTypeToCLRType(e.Type, e.Value);
}
else
{
brokeredMessageProperty.Properties.Add(e.Key,
ConversionHelper.MapStringTypeToCLRType(e.Type,
e.Value));
}
}
}
catch (Exception ex)
{
warningCollection.Add(string.Format(CultureInfo.CurrentUICulture,
PropertyConversionError, e.Key, ex.Message));
}
}

// if the warning collection contains at least one or more items,

188
// write them to the log and return immediately
StringBuilder builder;
if (warningCollection.Count > 0)
{
builder = new StringBuilder(WarningHeader);
var warnings = warningCollection.ToArray<string>();
for (var i = 0; i < warningCollection.Count; i++)
{
builder.AppendFormat(WarningFormat, warnings[i]);
}
mainForm.WriteToLog(builder.ToString());
return;
}

// Set the BrokeredMessageProperty properties
brokeredMessageProperty.Label = txtLabel.Text;
brokeredMessageProperty.MessageId = Guid.NewGuid().ToString();
brokeredMessageProperty.SessionId = sessionId;
brokeredMessageProperty.ReplyToSessionId = sessionId;
brokeredMessageProperty.ReplyTo = responseQueueRadioButton.Checked
? responseQueueUri
: responseTopicUri;

OperationContext.Current.OutgoingMessageProperties.Add(BrokeredMessageProperty.Name,

brokeredMessageProperty);

// Send the request message to the requestqueue or requesttopic
var stopwatch = new Stopwatch();
try
{
stopwatch.Start();

channelDictionary[endpointConfigurationName].SendRequest(calculatorRequestMessage);
}
catch (CommunicationException ex)
{
if (channelFactoryDictionary[endpointConfigurationName] != null)
{
channelFactoryDictionary[endpointConfigurationName].Abort();
channelFactoryDictionary.Remove(endpointConfigurationName);
channelDictionary.Remove(endpointConfigurationName);

189
}
HandleException(ex);
}
catch (Exception ex)
{
if (channelFactoryDictionary[endpointConfigurationName] != null)
{
channelFactoryDictionary[endpointConfigurationName].Abort();
channelFactoryDictionary.Remove(endpointConfigurationName);
channelDictionary.Remove(endpointConfigurationName);
}
HandleException(ex);
}
finally
{
stopwatch.Stop();
}
// Log the request message and its properties
builder = new StringBuilder();
builder.AppendLine(string.Format(CultureInfo.CurrentCulture,
MessageSuccessfullySent,

channelFactoryDictionary[endpointConfigurationName].Endpoint.Address.Uri.AbsoluteUri,
brokeredMessageProperty.MessageId,
brokeredMessageProperty.SessionId,
brokeredMessageProperty.Label,
stopwatch.ElapsedMilliseconds));
builder.AppendLine(PayloadFormat);
for (var i = 0; i < calculatorRequest.Operations.Count; i++)
{
builder.AppendLine(string.Format(RequestFormat,
i + 1,

calculatorRequest.Operations[i].Operand1,

calculatorRequest.Operations[i].Operator,

calculatorRequest.Operations[i].Operand2));
}
builder.AppendLine(SentMessagePropertiesHeader);
foreach (var p in brokeredMessageProperty.Properties)
{

190
builder.AppendLine(string.Format(MessagePropertyFormat,
p.Key,
p.Value));
}
var traceMessage = builder.ToString();
WriteToLog(traceMessage.Substring(0, traceMessage.Length - 1));
}
}
catch (Exception ex)
{
// Handle the exception
HandleException(ex);
}
finally
{
// Restoire the defaulf cursor
Cursor = Cursors.Default;
}
}




Response Handler Service
The following example contains the code of the WCF service used by the client application to
retrieve and log response messages from the responsequeue and ItalyMilan subscription of the
responsetopic. To accomplish this result, the service exposes two different endpoints each of
which uses the NetMessagingBinding and receives messages from one of the two queues.
Indeed, each subscription can be seen as a virtual queue getting copies of messages published
to the topic they belong to. The example below shows the code of the ResponseHandlerService
class. As you can notice, the service retrieves the BrokeredMessageProperty from the Properties
collection of the incoming WCF message and uses this object to access to the properties of the
response message. Since in the ICalculatorResponse service contract the ReceiveResponse
method is decorated with the [ReceiveContextEnabled(ManualControl = true)] attribute, the
service method must explicitly signal the receive. This signaling requires the service to explicitly
invoke the ReceiveContext.Complete method to commit the receive operation. In fact, as we said
at the beginning of the article, when the ManualControl property is set to true, the message
received from the channel is delivered to the service operation with a lock for the message. The
service implementation must call either Complete(TimeSpan) or Abandon(TimeSpan) to signal
the receive completion of the message. Failure to call either of these methods causes the lock to

191
be held on the message until the lock timeout interval elapses. Once the lock is released (either
through a call to Abandon(TimeSpan)) or lock timeout) the message is redispatched from the
channel to the service. Calling Complete(TimeSpan) marks the message as successfully
received.
ResponseHandlerService Class
[ServiceBehavior(Namespace = "http://windowsazure.cat.microsoft.com/samples/servicebus")]
public class ResponseHandlerService : ICalculatorResponseSessionful
{
#region Private Constants
//***************************
// Formats
//***************************
private const string MessageSuccessfullyReceived = "Response Message Received:\n -
EndpointUrl:[{0}]" +
"\n - CorrelationId=[{1}]\n -
SessionId=[{2}]\n - Label=[{3}]";
private const string ReceivedMessagePropertiesHeader = "Properties:";
private const string PayloadFormat = "Payload:";
private const string StatusFormat = " - Status=[{0}]";
private const string ResultFormat = " - Result[{0}]: Value=[{1}] Error=[{2}]";
private const string MessagePropertyFormat = " - Key=[{0}] Value=[{1}]";

//***************************
// Constants
//***************************
private const string Empty = "EMPTY";
#endregion

#region Public Operations
[OperationBehavior]
public void ReceiveResponse(CalculatorResponseMessage calculatorResponseMessage)
{
try
{
if (calculatorResponseMessage != null &&
calculatorResponseMessage.CalculatorResponse != null)
{
// Initialize calculatorResponse var
var calculatorResponse = calculatorResponseMessage.CalculatorResponse;

// Get the message properties

192
var incomingProperties =
OperationContext.Current.IncomingMessageProperties;
var brokeredMessageProperty =
incomingProperties[BrokeredMessageProperty.Name] as
BrokeredMessageProperty;

// Trace the response message
var builder = new StringBuilder();
if (brokeredMessageProperty != null)
builder.AppendLine(string.Format(MessageSuccessfullyReceived,

OperationContext.Current.Channel.LocalAddress.Uri.AbsoluteUri,

brokeredMessageProperty.CorrelationId ?? Empty,
brokeredMessageProperty.SessionId ??
Empty,
brokeredMessageProperty.Label ??
Empty));
builder.AppendLine(PayloadFormat);
builder.AppendLine(string.Format(StatusFormat,
calculatorResponse.Status));
if (calculatorResponse.Results != null &&
calculatorResponse.Results.Count > 0)
{
for (int i = 0; i < calculatorResponse.Results.Count; i++)
{
builder.AppendLine(string.Format(ResultFormat,
i + 1,

calculatorResponse.Results[i].Value,

calculatorResponse.Results[i].Error));
}
}
builder.AppendLine(ReceivedMessagePropertiesHeader);
if (brokeredMessageProperty != null)
{
foreach (var property in brokeredMessageProperty.Properties)
{
builder.AppendLine(string.Format(MessagePropertyFormat,
property.Key,
property.Value));

193
}
}
var traceMessage = builder.ToString();
Trace.WriteLine(traceMessage.Substring(0, traceMessage.Length - 1));

//Complete the Message
ReceiveContext receiveContext;
if (ReceiveContext.TryGet(incomingProperties, out receiveContext))
{
receiveContext.Complete(TimeSpan.FromSeconds(10.0d));
}
else
{
throw new InvalidOperationException("An exception occurred.");
}
}
}
catch (Exception ex)
{
Trace.WriteLine(ex.Message);
}
}
#endregion
}
ListenUri Endpoint Behavior
As you will see ahead in the article, its easy to define a WCF-Custom receive location to read
messages from a Service Bus queue or from a subscription. Nevertheless, in the previous section
we have seen that when defining a WCF service endpoint to consume messages from a
subscription, you have to you have to specify the URL of the topic as the address of the service
endpoint and the URL of the subscription as its listenUri. Now, when configuring a WCF receive
location using the WCF-Custom adapter, there is a textbox for specifying the service endpoint
address on the General tab of the configuration dialog, however there's no field to specify a value
of the listenUri and ListenUriMode properties.
For this reason, I decided to create a custom endpoint behavior that at runtime can set these
values for a WCF-Custom receive location. The first attempt was to use the
AddBindingParameters and ApplyDispatchBehavior methods exposed by the custom endpoint
behavior to set the listenUri and ListenUriMode of the service endpoint passed as a parameter
to both methods. However, this technique didn't work as expected.
Finally I solved the problem in the following way: I created a binding extension element class
called ListenUriBehaviorExtensionElement to register the custom endpoint behavior in the

194
machine.config file. This component exposes a property called listenUri that allows a user to
specify the URL of a subscription when configuring a WCF-Custom receive location.
At runtime, the ListenUriBehaviorExtensionElement component creates an instance of the
ListenUriEndpointBehavior class. The AddBindingParameters method of the custom endpoint
behavior replaces the original binding with a CustomBinding that contains the same binding
elements and injects an instance of the ListenUriBindingElement at the top of the binding. This
way, at runtime, the custom binding will be the first to execute. Finally, the BuildChannelListener
method of the ListenUriBindingElement assigns the URL specified in the configuration of the
receive location to the ListenUriBaseAddress property of the BindingContext and sets the value of
its ListenUriMode property to Explicit. For you convenience, I included the code for the three
classes below. Later in the article I'll show you how to use this component when defining a WCF-
Custom receive location that receives messages from a Service Bus subscription.
ListenUriBehaviorExtensionElement Class
public class ListenUriBehaviorExtensionElement : IEndpointBehavior
{
#region Private Constants
//***************************
// Constants
//***************************

private const string ListenUriName = "listenUri";
private const string IsTrackingEnabledName = "isTrackingEnabled";
private const string ListenUriDescription = "Gets or sets the URI at which the
service endpoint listens.";
private const string IsTrackingEnabledDescription = "Gets or sets a value indicating
whether tracking is enabled.";
#endregion

#region BehaviorExtensionElement Members
//***************************
// Protected Methods
//***************************

/// <summary>
/// Creates a behavior extension based on the current configuration settings.
/// </summary>
/// <returns>The behavior extension.</returns>
protected override object CreateBehavior()
{
return new ListenUriEndpointBehavior(ListenUri, IsTrackingEnabled);
}


195
/// <summary>
/// Gets the type of behavior.
/// </summary>
public override Type BehaviorType
{
get
{
return typeof(ListenUriEndpointBehavior);
}
}
#endregion

#region Public Properties
/// <summary>
/// Gets or sets the URI at which the service endpoint listens.
/// </summary>
[ConfigurationProperty(ListenUriName, IsRequired = true)]
[SettingsDescription(ListenUriDescription)]
public string ListenUri
{
get
{
return (string)base[ListenUriName];
}
set
{
base[ListenUriName] = value;
}
}

/// <summary>
/// Gets or sets a value indicating whether the message inspector is enabled.
/// </summary>
[ConfigurationProperty(IsTrackingEnabledName, DefaultValue = true, IsRequired =
false)]
[SettingsDescription(IsTrackingEnabledDescription)]
public bool IsTrackingEnabled
{
get
{
return (bool)base[IsTrackingEnabledName];
}

196
set
{
base[IsTrackingEnabledName] = value;
}
}
#endregion
}
ListenUriEndpointBehavior Class
public class ListenUriEndpointBehavior : IEndpointBehavior
{
#region Private Constants
//***************************
// Constants
//***************************

private const string ListerUriMessageFormat = "[ListenUriEndpointBehavior] ListenUri
= [{0}].";
#endregion

#region Public Constructors
private readonly string listenUri;
private readonly bool isTrackingEnabled;
#endregion

#region Public Constructors
/// <summary>
/// Initializes a new instance of the ListenUriEndpointBehavior class.
/// </summary>
/// <param name="listenUri">The URI at which the service endpoint listens</param>
/// <param name="isTrackingEnabled">A boolean value indicating whether tracking is
enabled</param>
public ListenUriEndpointBehavior(string listenUri, bool isTrackingEnabled)
{
this.listenUri = listenUri;
this.isTrackingEnabled = isTrackingEnabled;
}
#endregion

#region IEndpointBehavior Members
/// <summary>
/// Implement to pass data at runtime to bindings to support custom behavior.
/// </summary>

197
/// <param name="endpoint">The endpoint to modify.</param>
/// <param name="bindingParameters">The objects that binding elements require to
support the behavior.</param>
void IEndpointBehavior.AddBindingParameters(ServiceEndpoint endpoint,
BindingParameterCollection bindingParameters)
{
if (endpoint != null &&
!String.IsNullOrEmpty(listenUri))
{
// Read the binding elements from the original binding
var bindingElementCollection = endpoint.Binding.CreateBindingElements();
// Create an array of binding elements
var bindingElementArray = new BindingElement[bindingElementCollection.Count +
1];
// Add an instance of the ListenUriBindingElement as first binding element of
the array
bindingElementArray[0] = new ListenUriBindingElement(listenUri);
// Copy the binding elements of the original binding to the array
bindingElementCollection.CopyTo(bindingElementArray, 1);
// Create a custom binding with the same binding elements as the original
// binding with the addition of the custom binding as first item
var customBinding = new CustomBinding(bindingElementArray)
{
CloseTimeout = endpoint.Binding.CloseTimeout,
OpenTimeout = endpoint.Binding.OpenTimeout,
ReceiveTimeout = endpoint.Binding.ReceiveTimeout,
SendTimeout = endpoint.Binding.SendTimeout,
Name = endpoint.Binding.Name,
Namespace = endpoint.Binding.Namespace
};
//Replace the original binding with the newly created binding
endpoint.Binding = customBinding;
Trace.WriteLineIf(isTrackingEnabled,
string.Format(ListerUriMessageFormat,
listenUri));
}
}

/// <summary>
/// Implements a modification or extension of the client across an endpoint.
/// </summary>
/// <param name="endpoint">The endpoint that is to be customized.</param>

198
/// <param name="clientRuntime">The client runtime to be customized.</param>
void IEndpointBehavior.ApplyClientBehavior(ServiceEndpoint endpoint, ClientRuntime
clientRuntime)
{
}

/// <summary>
/// Implements a modification or extension of the service across an endpoint.
/// </summary>
/// <param name="endpoint">The endpoint that exposes the contract.</param>
/// <param name="endpointDispatcher">The endpoint dispatcher to be modified or
extended.</param>
void IEndpointBehavior.ApplyDispatchBehavior(ServiceEndpoint endpoint,
EndpointDispatcher endpointDispatcher)
{
}

/// <summary>
/// Implement to confirm that the endpoint meets some intended criteria.
/// </summary>
/// <param name="endpoint">The endpoint to validate.</param>
void IEndpointBehavior.Validate(ServiceEndpoint endpoint)
{
}
#endregion
}
ListenUriBindingElement Class
public class ListenUriBindingElement : BindingElement
{
#region Private Fields
private readonly string listenUri;
#endregion

#region Public Constructor
/// <summary>
/// Initializes a new instance of the ListenUriBindingElement class.
/// </summary>
/// <param name="listenUri">A BindingElement object that is a deep clone of the
original.</param>
public ListenUriBindingElement(string listenUri)
{
this.listenUri = listenUri;

199
}
#endregion

#region BindingElement Members
/// <summary>
/// returns a copy of the binding element object.
/// </summary>
/// <returns></returns>
public override BindingElement Clone()
{
return new ListenUriBindingElement(listenUri);
}

/// <summary>
/// Returns a typed object requested, if present, from the appropriate layer in the
binding stack.
/// </summary>
/// <typeparam name="T">The typed object for which the method is
querying.</typeparam>
/// <param name="context">The BindingContext for the binding element.</param>
/// <returns>The typed object T requested if it is present or nullif it is not
present.</returns>
public override T GetProperty<T>(BindingContext context)
{
return context.GetInnerProperty<T>();
}

/// <summary>
/// Returns a value that indicates whether the binding element can build a
/// channel factory for a specific type of channel.
/// </summary>
/// <typeparam name="TChannel">The type of channel the channel factory
produces.</typeparam>
/// <param name="context">The BindingContext that provides context for the binding
element.</param>
/// <returns>true if the IChannelFactory<TChannel/>of type TChannel can be built by
/// the binding element; otherwise, false.</returns>
public override bool CanBuildChannelFactory<TChannel>(BindingContext context)
{
return false;
}


200
/// <summary>
/// Initializes a channel factory for producing channels of a specified type from the
binding context.
/// </summary>
/// <typeparam name="TChannel">The type of channel the factory builds.</typeparam>
/// <param name="context">The BindingContext that provides context for the binding
element.</param>
/// <returns>The IChannelFactory<TChannel/>of type TChannel initialized from the
context. </returns>
public override IChannelFactory<TChannel>
BuildChannelFactory<TChannel>(BindingContext context)
{
throw new NotSupportedException();
}

/// <summary>
/// Returns a value that indicates whether the binding element can build a channel
/// listener for a specific type of channel.
/// </summary>
/// <typeparam name="TChannel">The type of channel listener to build.</typeparam>
/// <param name="context">The BindingContext for the binding element.</param>
/// <returns>true if a channel listener of the specified type can be built;
/// otherwise, false. The default is false. </returns>
public override bool CanBuildChannelListener<TChannel>(BindingContext context)
{
return context.CanBuildInnerChannelListener<TChannel>();
}

/// <summary>
/// Initializes a channel listener for producing channels of a specified type from
the binding context.
/// </summary>
/// <typeparam name="TChannel">The type of channel that the listener is built to
accept.</typeparam>
/// <param name="context">The BindingContext for the binding element.</param>
/// <returns>The IChannelListener<TChannel/>of type IChannel initialized from the
context.</returns>
public override IChannelListener<TChannel>
BuildChannelListener<TChannel>(BindingContext context)
{
if (!string.IsNullOrEmpty(listenUri))
{




}
#e
}

Comp
You c
monit





Ses
In the
subsc
IOrde
// Ser
// All
enforc
// Ses
Broker
[Servi
public
{
[O
[R
vo
}

// Ser
conte
conte
}
return co

endregion
ponent Track
can enable the
tor its runtime
sionChan
e first part of th
cription, a WC
erServiceSes
rviceBus doe
l senders se
ce SessionMo
ssionful mes
redMessagePr
iceContract]
c interface
OperationCon
ReceiveConte
oid ReceiveO
rviceBus sup
xt.ListenUri
xt.ListenUri
ntext.BuildI
king
e component
e behavior.
nnel End
he article we
CF service ne
ssionful in the
s not suppor
nding messag
de.Required.
sages are se
operty objec

IOrderServic
tract(IsOneW
xtEnabled(Ma
rder(Order o
ports both I
BaseAddress
Mode = Liste
nnerChannelL
tracking and
point Be
have seen th
eds to implem
e following co
rt IOutputSes
ges to sessio

nt by settin
t.
e
Way = true)]
nualControl
rder);
InputChannel
= new Uri(li
enUriMode.Exp
Listener<TCha
use DebugV
havior
hat to receive
ment a sessio
ode snippet:
ssionChannel.
onful queue m
ng the Sessio
= true)]
l and IInputS
istenUri);
plicit;
annel>();
View, as show
messages fro
on-aware cont

must use a co
onId property
SessionChanne
wn in the pictu
om a session
tract interface
ontract which
y of the
el.

20
re below, to

ful queue or
e like
h does not
01

202
// A sessionful service listening to a sessionful queue must have SessionMode.Required in
its contract.
[ServiceContract(SessionMode = SessionMode.Required)]
public interface IOrderServiceSessionful : IOrderService
{
}

[ServiceBehavior(InstanceContextMode = InstanceContextMode.PerSession, ConcurrencyMode =
ConcurrencyMode.Single)]
public class OrderService : IOrderServiceSessionful
{
[OperationBehavior]
public void ReceiveOrder(Order order)
{
...
}
}
In the section on BizTalk Server WCF adapters we have noticed that WCF receive locations are
implemented by a singleton instance of a WCF service class called BizTalkServiceInstance which
implements multiple untyped, generic service contracts.
IOneWayAsync
IOneWayAsyncTxn
ITwoWayAsync
ITwoWayAsyncVoid
ITwoWayAsyncVoidTxn
All of these interfaces are marked with the following attribute:
[ServiceContract(Namespace = "http://www.microsoft.com/biztalk/2006/r2/wcf-adapter",
SessionMode = SessionMode.Allowed)]
In particular, the SessionMode property of the ServiceContract attribute is set to
SessionMode.Allowed. This setting specifies that the contract supports sessions if the incoming
binding supports them. In other words, all the service contracts implemented by the
BizTalkServiceInstance support, but not require, a session.
When the following conditions are satisfied, the WCF runtime always selects the IInputChannel
over the IInputSessionChannel when it selects the channel to use:
The service contract exposed by a service endpoint allows, but not requires sessions
(SessionMode = SessionMode.Allowed).
The binding used by the service endpoint supports both sessionless and session-based
communication (like the NetMessagingBinding).
This rule implies that when you define a WCF-Custom receive location to consume messages
from a sessionful queue or subscription, when you enable it, you will see an error like the
following in the Application Log:

203
The adapter "WCF-Custom" raised an error message. Details
"System.InvalidOperationException: It is not
possible for an entity that requires sessions to create a non-sessionful message
receiver..
TrackingId:9163a41f-d792-406d-acbe-eb93ab2defb8_1_1,TimeStamp:10/3/2011 12:39:02 PM --->
System.ServiceModel.FaultException`1[System.ServiceModel.ExceptionDetail]: It is not
possible for an entity
that requires sessions to create a non-sessionful message receiver..
TrackingId:9163a41f-d792-406d-acbe-eb93ab2defb8_1_1,TimeStamp:10/3/2011 12:39:02 PM
To avoid this problem, when we control the definition of the service contract, we can simply mark
its ServiceContract attribute as requiring sessions. Unfortunately, we cannot change the definition
of the service contracts implemented by the BizTalkServiceInstance class. So, how can we force
a WCF receive location to use an IInputSessionChannel instead of an IInputChannel when
consuming messages from a sessionful queue or subscription? Even in this case I solved the
problem by using a WCF extension.
I created a binding extension element class called SessionChannelBehaviorExtensionElement
to register the custom endpoint behavior in the machine.config. At runtime, this component
creates an instance of the SessionChannelEndpointBehavior class. The
AddBindingParameters method of the custom endpoint behavior replaces the original binding (in
our case the NetMessagingBinding) with a CustomBinding that contains the same binding
elements and injects an instance of the SessionChannelBindingElement at the top of the new
binding. This way, at runtime, the custom binding will be the first to execute. Finally, I customized
the BindingElement.CanBuildChannelListener<TChannel> method to return false when the type
of channel is IInputChannel. In a nutshell, this component forces the WCF runtime to skip the
IInputChannel. Hence, you can use it in a WCF-Custom receive location to force the WCF
adapter to use the IInputSessionChannel. For your convenience, I included the code of the
three classes below. Later in the article I'll show you how to use this component when defining a
WCF-Custom receive location that consumes messages from a sessionful queue or subscription.
SessionChannelBehaviorExtensionElement Class
public class SessionChannelBehaviorExtensionElement : BehaviorExtensionElement
{
#region Private Constants
//***************************
// Constants
//***************************

private const string IsTrackingEnabledName = "isTrackingEnabled";
private const string IsTrackingEnabledDescription =
"Gets or sets a value indicating whether
tracking is enabled.";
#endregion


204
#region BehaviorExtensionElement Members
//***************************
// Protected Methods
//***************************

/// <summary>
/// Creates a behavior extension based on the current configuration settings.
/// </summary>
/// <returns>The behavior extension.</returns>
protected override object CreateBehavior()
{
return new SessionChannelEndpointBehavior(IsTrackingEnabled);
}

/// <summary>
/// Gets the type of behavior.
/// </summary>
public override Type BehaviorType
{
get
{
return typeof(SessionChannelEndpointBehavior);
}
}
#endregion

#region Public Properties
/// <summary>
/// Gets or sets a value indicating whether the message inspector is enabled.
/// </summary>
[ConfigurationProperty(IsTrackingEnabledName, DefaultValue = true, IsRequired =
false)]
[SettingsDescription(IsTrackingEnabledDescription)]
public bool IsTrackingEnabled
{
get
{
return (bool)base[IsTrackingEnabledName];
}
set
{
base[IsTrackingEnabledName] = value;

205
}
}
#endregion
}
SessionChannelBindingElement Class
public class SessionChannelBindingElement : BindingElement
{
#region Private Constants
//***************************
// Constants
//***************************

private const string CanBuildIInputSessionChannel =
"[SessionChannelBindingElement] CanBuildChannelListener returned
true for InputSessionChannel.";
private const string CannotBuildIInputChannel =
"[SessionChannelBindingElement] CanBuildChannelListener returned
false for IInputChannel.";
#endregion

#region Private Fields
private readonly bool isTrackingEnabled;
#endregion

#region Public Constructor
/// <summary>
/// Initializes a new instance of the SessionChannelBindingElement class.
/// </summary>
/// <param name="isTrackingEnabled">A boolean value indicating whether tracking is
enabled</param>
public SessionChannelBindingElement(bool isTrackingEnabled)
{
this.isTrackingEnabled = isTrackingEnabled;
}
#endregion

#region BindingElement Members
/// <summary>
/// returns a copy of the binding element object.
/// </summary>
/// <returns></returns>
public override BindingElement Clone()

206
{
return new SessionChannelBindingElement(isTrackingEnabled);
}

/// <summary>
/// Returns a typed object requested, if present, from the appropriate layer in the
binding stack.
/// </summary>
/// <typeparam name="T">The typed object for which the method is
querying.</typeparam>
/// <param name="context">The BindingContext for the binding element.</param>
/// <returns>The typed object T requested if it is present or nullif it is not
present.</returns>
public override T GetProperty<T>(BindingContext context)
{
return context.GetInnerProperty<T>();
}

/// <summary>
/// Returns a value that indicates whether the binding element can build
/// a channel factory for a specific type of channel.
/// </summary>
/// <typeparam name="TChannel">The type of channel the channel factory
produces.</typeparam>
/// <param name="context">The BindingContext that provides context for the binding
element.</param>
/// <returns>true if the IChannelFactory<TChannel/>of type TChannel can be built by
/// the binding element; otherwise, false.</returns>
public override bool CanBuildChannelFactory<TChannel>(BindingContext context)
{
return false;
}

/// <summary>
/// Initializes a channel factory for producing channels of a specified type from the
binding context.
/// </summary>
/// <typeparam name="TChannel">The type of channel the factory builds.</typeparam>
/// <param name="context">The BindingContext that provides context for the binding
element.</param>
/// <returns>The IChannelFactory<TChannel/>of type TChannel initialized from the
context. </returns>

207
public override IChannelFactory<TChannel>
BuildChannelFactory<TChannel>(BindingContext context)
{
throw new NotSupportedException();
}

/// <summary>
/// Returns a value that indicates whether the binding element can build a
/// channel listener for a specific type of channel.
/// </summary>
/// <typeparam name="TChannel">The type of channel listener to build.</typeparam>
/// <param name="context">The BindingContext for the binding element.</param>
/// <returns>true if a channel listener of the specified type can be built;
/// otherwise, false. The default is false. </returns>
public override bool CanBuildChannelListener<TChannel>(BindingContext context)
{
var ok = typeof(TChannel) != typeof(IInputChannel) &&
context.CanBuildInnerChannelListener<TChannel>();
Trace.WriteLineIf(isTrackingEnabled, ok ? CanBuildIInputSessionChannel :
CannotBuildIInputChannel);
return ok;
}

/// <summary>
/// Initializes a channel listener for producing channels of a specified type from
the binding context.
/// </summary>
/// <typeparam name="TChannel">The type of channel that the listener is built to
accept.</typeparam>
/// <param name="context">The BindingContext for the binding element.</param>
/// <returns>The IChannelListener<TChannel/>of type IChannel initialized from the
context.</returns>
public override IChannelListener<TChannel>
BuildChannelListener<TChannel>(BindingContext context)
{
return context.BuildInnerChannelListener<TChannel>();
}
#endregion
}
Component Tracking
You can enable the component tracking and use DebugView, as shown in the picture below, to
monitor its runtime behavior.





Prop
To ac
transl
mess
schem
T
el
in

T
co
perty Sch
chieve a full in
ate the prope
age and vice
mas:
he first prope
lement for ea
ndependent fr
he second pr
ontext proper
hemas
ntegration bet
erties exposed
versa. The fi
erty schema is
ch property o
rom a specific
roperty schem
rties used by t
tween the Ser
d by a Broker
rst step in this
s called Brok
of the Brokere
c application a
ma is called ju
the ServiceB
rvice Bus and
redMessage i
s direction ha
eredMessag
edMessage cl
and is fully re

ust PropertyS
BusSample a
d BizTalk Serv
nto context p
as been the de
ePropertySc
ass. This pro
usable.
Schema and c
pplication.
ver, its neces
properties of a
efinition of tw
chema and co
operty schema
contains the s

20

ssary to
a BizTalk
wo property
ontains an
a is
specific
08

209


All of the context properties defined in both property schemas inherit from the
MessageContextPropertyBase base class. This inheritance allows you to define any of
these properties independently from the schema of the received message when you
define the filter expression of a direct bound orchestration.




Schemas
The request and response messages exchanged by the client and BizTalk applications have the
following formats:
CalculatorRequest Message
<CalculatorRequest xmlns="http://windowsazure.cat.microsoft.com/samples/servicebus">
<Method>DynamicSendPortOrchestration</Method>
<Operations>
<Operation>
<Operator>+</Operator>
<Operand1>82</Operand1>
<Operand2>18</Operand2>
</Operation>
<Operation>
<Operator>-</Operator>
<Operand1>30</Operand1>
<Operand2>12</Operand2>
</Operation>
<Operation>
<Operator>*</Operator>
<Operand1>25</Operand1>
Note

210
<Operand2>8</Operand2>
</Operation>
<Operation>
<Operator>\</Operator>
<Operand1>100</Operand1>
<Operand2>25</Operand2>
</Operation>
<Operation>
<Operator>+</Operator>
<Operand1>100</Operand1>
<Operand2>32</Operand2>
</Operation>
</Operations>
</CalculatorRequest>
CalculatorResponse Message
<CalculatorResponse xmlns="http://windowsazure.cat.microsoft.com/samples/servicebus">
<Status>Ok</Status>
<Results>
<Result>
<Value>100</Value>
<Error>None</Error>
</Result>
<Result>
<Value>18</Value>
<Error>None</Error>
</Result>
<Result>
<Value>200</Value>
<Error>None</Error>
</Result>
<Result>
<Value>4</Value>
<Error>None</Error>
</Result>
<Result>
<Value>132</Value>
<Error>None</Error>
</Result>
</Results>
</CalculatorResponse>
Both message types are defined by an XSD file contained in the Schemas project that you can find
in the companion code for this article.

211




Service Bus Message Inspector
The final step to bridge the gap between the Service Bus and BizTalk Server has been to create
a component to translate the properties of a BrokeredMessage into context properties of a
BizTalk message and vice versa. Indeed, a WCF-Custom receive location can use the
NetMessagingBinding to receive messages from a queue or subscription without the need to use
any custom component to extend the normal behavior of the WCF adapter runtime. However,
using this approach, a BizTalk application can read only the body of the message retrieved from
the Service Bus, not the value of the properties specified by the sender. At the beginning of the
article, I highlighted that its quite common to have a message with an empty body and the entire
payload specified in the user defined properties. For this reason, its extremely important, not to
say mandatory, to create a component that turns the properties of a BrokeredMessage into
context properties of a BizTalk message and vice versa.
In the first part of the article we have seen that a WCF service which uses the
NetMessagingBinding to read messages from a queue or a subscription, can retrieve a
BrokeredMessageProperty object from the Properties collection of the incoming WCF message.
This fact gives us the possibility to write a custom message inspector to intercept the incoming
WCF message and read the BrokeredMessageProperty from its properties. For obvious reasons,
this operation cannot be done using a custom pipeline component because a custom pipeline
does not provide access to the WCF message. In the previous step we created the property
schemas which define, respectively, the object specific and user defined properties for Service
Bus messages. The last piece to complete the puzzle is to find a way to promote or write a
property to the BizTalk message context. Fortunately, the WCF adapter runtime provides an easy
way to accomplish this task.
As explained in the BizTalk Server documentation on MSDN, to promote a set of properties to the
BizTalk message context you need to add their xml qualified name and value to a collection of
key/value pairs. Next you must add a new property to the inbound WCF message whose key
must be equal to:
http://schemas.microsoft.com/BizTalk/2006/01/Adapters/WCF-properties/Promote
and specify the collection of key/value pairs as its value. The WCF adapter looks for this property
in the WCF message and will promote all the properties contained in the collection.
Likewise, to write but not promote a set of properties to the BizTalk message context, you need to
add their XML qualified name and value to another collection of key/value pairs. Next you must
add another property to the inbound WCF message whose key must be equal to:
http://schemas.microsoft.com/BizTalk/2006/01/Adapters/WCF-properties/WriteToContext
and specify the second collection of key/value pairs as its value. The WCF adapter looks for this
property in the WCF message and writes all the properties contained in the collection to the
BizTalk message context.

212
The ServiceBusMessageInspector component exploits this feature to promote or write the
properties of the BrokeredMessageProperty to the context of the BizTalk message when a WCF-
Custom receive location reads a message from a Service Bus queue or subscription. Conversely,
the component uses another characteristic of the WCF adapter to retrieve the context property
from an outbound WCF message and translate them into properties of a
BrokeredMessageProperty. In fact, the send handler of the WCF adapter copies all the message
context properties into the Properties collection when it creates a WCF message from a BizTalk
message. Therefore, when using a WCF-Custom send port, this allows access to any message
context properties within a custom message inspector or channel component at runtime.
The ServiceBusMessageInspectorBehaviorExtensionElement can be used to register the
message inspector in the machine.config file. It exposes the following properties that you can
configure when creating a WCF-Custom receive location or send port:
IsComponentEnabled: gets or sets a value indicating whether the message inspector is
enabled.
IsTrackingEnabled: Gets or sets a value indicating whether the message inspector is
enabled.
PropertiesToPromote: defines which user defined properties must be extracted from the
Properties collection of the BrokeredMessageProperty and promoted to the BizTalk message
context. This property is intended to be used only when the message inspector is used in a
WCF-Custom receive location
PropertiesToWrite: defines which user defined properties must be extracted from the
Properties collection of the BrokeredMessageProperty and written but not promoted to the
BizTalk message context. This property is intended to be used only when the message
inspector is used in a WCF-Custom receive location.
PropertiesToSend: defines which context properties must be extracted from the Properties
collection of the WCF message and added to the Properties collection of the
BrokeredMessageProperty. This property is intended to be used only when the message
inspector is used in a WCF-Custom send port.
For you convenience, I included the code of the custom message inspector in the following
example.
ServiceBusMessageInspector Class
/// <summary>
/// This class is used to promote\write properties to the message context.
/// </summary>
public class ServiceBusMessageInspector : IDispatchMessageInspector,
IClientMessageInspector
{
#region Private Constants
//***************************
// Constants
//***************************

private const string Source = "ServiceBusMessageInspector";

213
private const string BrokeredMessagePropertySchemaNamespace =
"http://schemas.microsoft.com/servicebus/2011/brokered-message-property";
private const string PropertiesToPromoteKey =
"http://schemas.microsoft.com/BizTalk/2006/01/Adapters/WCF-properties/Promote";
private const string PropertiesToWriteKey =
"http://schemas.microsoft.com/BizTalk/2006/01/Adapters/WCF-
properties/WriteToContext";
private const string ContentTypeProperty =
"http://schemas.microsoft.com/servicebus/2011/brokered-message-
property#ContentType";
private const string CorrelationIdProperty =
"http://schemas.microsoft.com/servicebus/2011/brokered-message-
property#CorrelationId";
private const string LabelProperty =
"http://schemas.microsoft.com/servicebus/2011/brokered-message-property#Label";
private const string ReplyToProperty =
"http://schemas.microsoft.com/servicebus/2011/brokered-message-property#ReplyTo";
private const string ReplyToSessionIdProperty =
"http://schemas.microsoft.com/servicebus/2011/brokered-message-
property#ReplyToSessionId";
private const string SessionIdProperty =
"http://schemas.microsoft.com/servicebus/2011/brokered-message-
property#SessionId";
private const string ToProperty =
"http://schemas.microsoft.com/servicebus/2011/brokered-message-property#To";
private const string ContentType = "ContentType";
private const string CorrelationId = "CorrelationId";
private const string DeliveryCount = "DeliveryCount";
private const string EnqueuedTimeUtc = "EnqueuedTimeUtc";
private const string ExpiresAtUtc = "ExpiresAtUtc";
private const string Label = "Label";
private const string MessageId = "MessageId";
private const string ReplyTo = "ReplyTo";
private const string ReplyToSessionId = "ReplyToSessionId";
private const string ScheduledEnqueueTimeUtc = "ScheduledEnqueueTimeUtc";
private const string SequenceNumber = "SequenceNumber";
private const string SessionId = "SessionId";
private const string TimeToLive = "TimeToLive";
private const string To = "To";
private const string PropertiesToPromote = "PropertiesToPromote";
private const string PropertiesToWrite = "PropertiesToWrite";
private const string PropertiesToSend = "PropertiesToSend";

214

//***************************
// Messages
//***************************
private const string PropertiesElementCannotBeNull =
"[ServiceBusMessageInspector] The PropertiesElement object cannot be null.";
private const string NameAndNamespaceNumberAreDifferent =
"[ServiceBusMessageInspector] The number of items in the name and namespace
comma-separated " +
"lists is different in the {0} property.";
private const string PropertyAddedToList = "[ServiceBusMessageInspector] The
following property was added to " +
"the [{0}] list:\n\r - Name=[{1}]\n\r - Namespace=[{2}]\n\r - Value=[{3}]";
private const string ExceptionFormat = "[ServiceBusMessageInspector] The following
exception occurred=[{0}].";
#endregion

#region Private Fields
//***************************
// Private Fields
//***************************

private readonly bool isComponentEnabled;
private readonly bool isTrackingEnabled;
private readonly Dictionary<string, string> propertiesToPromoteDictionary = new
Dictionary<string, string>();
private readonly Dictionary<string, string> propertiesToWriteDictionary = new
Dictionary<string, string>();
private readonly Dictionary<string, string> propertiesToSendDictionary = new
Dictionary<string, string>();
#endregion

#region Public Constructors
//***************************
// Public Constructors
//***************************

/// <summary>
/// Initializes a new instance of the ServiceBusMessageInspector class.
/// </summary>
public ServiceBusMessageInspector()
{

215
isComponentEnabled = true;
}

/// <summary>
/// Initializes a new instance of the ServiceBusMessageInspector class.
/// </summary>
/// <param name="isComponentEnabled">This value indicates whether the message
inspector is enabled.</param>
/// <param name="isTrackingEnabled">This value indicates whether tracking is
enabled.</param>
/// <param name="propertiesToPromote">This object defines the properties to
promote.</param>
/// <param name="propertiesToWrite">This object defines the properties to
write.</param>
/// <param name="propertiesToSend">This object defines the properties to
send.</param>
public ServiceBusMessageInspector(bool isComponentEnabled,
bool isTrackingEnabled,
PropertiesElement propertiesToPromote,
PropertiesElement propertiesToWrite,
PropertiesElement propertiesToSend)
{
this.isComponentEnabled = isComponentEnabled;
this.isTrackingEnabled = isTrackingEnabled;

// Propulate the dictionary that contains the properties to promote
InitializeDictionary(PropertiesToPromote, propertiesToPromote,
propertiesToPromoteDictionary);

// Propulate the dictionary that contains the properties to write
InitializeDictionary(PropertiesToWrite, propertiesToWrite,
propertiesToWriteDictionary);

// Propulate the dictionary that contains the properties to send out
InitializeDictionary(PropertiesToSend, propertiesToSend,
propertiesToSendDictionary);
}
#endregion

#region IDispatchMessageInspector Members
//**********************************
// IDispatchMessageInspector Members

216
//**********************************

/// <summary>
/// Called after an inbound message has been received but
/// before the message is dispatched to the intended operation.
/// This method extracts the properties from the BrokeredMessageProperty
/// object contained in the WCF message properties and write or promote
/// them in the message context.
/// </summary>
/// <param name="request">The request message.</param>
/// <param name="channel">The incoming channel.</param>
/// <param name="instanceContext">The current service instance.</param>
/// <returns>The object used to correlate state. This object is passed back
/// in the BeforeSendReply method.</returns>
public object AfterReceiveRequest(ref Message request, IClientChannel channel,
InstanceContext instanceContext)
{
try
{
if (!isComponentEnabled)
{
return Guid.NewGuid();
}

var propertiesToPromoteList = new List<KeyValuePair<XmlQualifiedName,
object>>();
var propertiesToWriteList = new List<KeyValuePair<XmlQualifiedName,
object>>();

// Retreive the BrokeredMessageProperty from the current operation context
var incomingProperties = OperationContext.Current.IncomingMessageProperties;
var brokeredMessageProperty =
(BrokeredMessageProperty)incomingProperties[BrokeredMessageProperty.Name];

if (brokeredMessageProperty != null)
{
// Include BrokeredMessageProperty explicit properties
// in the list of properties to write to the BizTalk message context
if (!string.IsNullOrEmpty(brokeredMessageProperty.ContentType))
{
AddItemToList(ContentType,
BrokeredMessagePropertySchemaNamespace,

217
brokeredMessageProperty.ContentType,
PropertiesToWrite,
propertiesToWriteList);
}
if (!string.IsNullOrEmpty(brokeredMessageProperty.CorrelationId))
{
AddItemToList(CorrelationId,
BrokeredMessagePropertySchemaNamespace,
brokeredMessageProperty.CorrelationId,
PropertiesToWrite,
propertiesToWriteList);
}
AddItemToList(DeliveryCount,
BrokeredMessagePropertySchemaNamespace,
brokeredMessageProperty.DeliveryCount,
PropertiesToWrite,
propertiesToWriteList);
AddItemToList(EnqueuedTimeUtc,
BrokeredMessagePropertySchemaNamespace,
brokeredMessageProperty.EnqueuedTimeUtc,
PropertiesToWrite,
propertiesToWriteList);
AddItemToList(ExpiresAtUtc,
BrokeredMessagePropertySchemaNamespace,
brokeredMessageProperty.ExpiresAtUtc,
PropertiesToWrite,
propertiesToWriteList);
if (!string.IsNullOrEmpty(brokeredMessageProperty.Label))
{
AddItemToList(Label,
BrokeredMessagePropertySchemaNamespace,
brokeredMessageProperty.Label,
PropertiesToWrite,
propertiesToWriteList);
}

if (!string.IsNullOrEmpty(brokeredMessageProperty.MessageId))
{
AddItemToList(MessageId,
BrokeredMessagePropertySchemaNamespace,
brokeredMessageProperty.MessageId,
PropertiesToWrite,

218
propertiesToWriteList);
}
if (!string.IsNullOrEmpty(brokeredMessageProperty.ReplyTo))
{
AddItemToList(ReplyTo,
BrokeredMessagePropertySchemaNamespace,
brokeredMessageProperty.ReplyTo,
PropertiesToWrite,
propertiesToWriteList);
}
if (!string.IsNullOrEmpty(brokeredMessageProperty.ReplyToSessionId))
{
AddItemToList(ReplyToSessionId,
BrokeredMessagePropertySchemaNamespace,
brokeredMessageProperty.ReplyToSessionId,
PropertiesToWrite,
propertiesToWriteList);
}
AddItemToList(ScheduledEnqueueTimeUtc,
BrokeredMessagePropertySchemaNamespace,
brokeredMessageProperty.ScheduledEnqueueTimeUtc,
PropertiesToWrite,
propertiesToWriteList);
AddItemToList(SequenceNumber,
BrokeredMessagePropertySchemaNamespace,
brokeredMessageProperty.SequenceNumber,
PropertiesToWrite,
propertiesToWriteList);
if (!string.IsNullOrEmpty(brokeredMessageProperty.SessionId))
{
AddItemToList(SessionId,
BrokeredMessagePropertySchemaNamespace,
brokeredMessageProperty.SessionId,
PropertiesToWrite,
propertiesToWriteList);
}
AddItemToList(TimeToLive,
BrokeredMessagePropertySchemaNamespace,
brokeredMessageProperty.TimeToLive.TotalSeconds,
PropertiesToWrite,
propertiesToWriteList);
if (!string.IsNullOrEmpty(brokeredMessageProperty.To))

219
{
AddItemToList(To,
BrokeredMessagePropertySchemaNamespace,
brokeredMessageProperty.To,
PropertiesToWrite,
propertiesToWriteList);
}

// Promote or write properties indicated in the
// configuration of the message inspector.
if (brokeredMessageProperty.Properties != null &&
brokeredMessageProperty.Properties.Count > 0)
{
foreach (var property in brokeredMessageProperty.Properties)
{
if (propertiesToPromoteDictionary.ContainsKey(property.Key))
{
AddItemToList(property.Key,
propertiesToPromoteDictionary[property.Key],
property.Value,
PropertiesToPromote,
propertiesToPromoteList);
continue;
}
if (propertiesToWriteDictionary.ContainsKey(property.Key))
{
AddItemToList(property.Key,
propertiesToWriteDictionary[property.Key],
property.Value,
PropertiesToWrite,
propertiesToWriteList);
}
}
}
}

// Notify the WCF Adapter the properties to promote
if (propertiesToPromoteList.Count > 0)
{
request.Properties.Add(PropertiesToPromoteKey, propertiesToPromoteList);
}


220
// Notify the WCF Adapter the properties to write
if (propertiesToWriteList.Count > 0)
{
request.Properties.Add(PropertiesToWriteKey, propertiesToWriteList);
}
}
catch (Exception ex)
{
Trace.WriteLineIf(isTrackingEnabled, string.Format(ExceptionFormat,
ex.Message));
EventLog.WriteEntry(Source, ex.Message, EventLogEntryType.Error);
throw;
}
return Guid.NewGuid();
}

/// <summary>
/// Called after the operation has returned but before the reply message is sent.
/// </summary>
/// <param name="reply">The reply message. This value is null if the operation is one
way.</param>
/// <param name="correlationState">The correlation object returned from the
AfterReceiveRequest method.</param>
public void BeforeSendReply(ref Message reply, object correlationState)
{
return;
}
#endregion

#region IClientMessageInspector Members
//**********************************
// IClientMessageInspector Members
//**********************************

/// <summary>
/// Enables inspection or modification of a message after a reply message is
/// received but prior to passing it back to the client application.
/// </summary>
/// <param name="reply">The message to be transformed into types and handed
/// back to the client application.</param>
/// <param name="correlationState">Correlation state data.</param>
public void AfterReceiveReply(ref Message reply, object correlationState)

221
{
return;
}

/// <summary>
/// Enables inspection or modification of a message before a request message is sent
to a service.
/// </summary>
/// <param name="message">The message to be sent to the service.</param>
/// <param name="channel">The client object channel.</param>
/// <returns>The object that is returned as the correlationState
/// argument of the AfterReceiveReply method.</returns>
public object BeforeSendRequest(ref Message message, IClientChannel channel)
{
try
{
if (!isComponentEnabled)
{
return message.Headers.Action;
}

// Create a new BrokeredMessageProperty object
var brokeredMessageProperty = new BrokeredMessageProperty();

// Adds properties to the BrokeredMessageProperty object
foreach (var property in message.Properties)
{
if (string.Compare(property.Key, ContentTypeProperty, true) == 0)
{
brokeredMessageProperty.ContentType =
message.Properties[ContentTypeProperty] as string;
Trace.WriteLineIf(isTrackingEnabled,
string.Format(PropertyAddedToList,
PropertiesToSend,
ContentType,

BrokeredMessagePropertySchemaNamespace,

brokeredMessageProperty.ContentType));
continue;
}
if (string.Compare(property.Key, CorrelationIdProperty, true) == 0)

222
{
brokeredMessageProperty.CorrelationId =
message.Properties[CorrelationIdProperty] as string;
Trace.WriteLineIf(isTrackingEnabled,
string.Format(PropertyAddedToList,
PropertiesToSend,
CorrelationId,

BrokeredMessagePropertySchemaNamespace,

brokeredMessageProperty.CorrelationId));
continue;
}
if (string.Compare(property.Key, LabelProperty, true) == 0)
{
brokeredMessageProperty.Label = message.Properties[LabelProperty] as
string;
Trace.WriteLineIf(isTrackingEnabled,
string.Format(PropertyAddedToList,
PropertiesToSend,
Label,

BrokeredMessagePropertySchemaNamespace,

brokeredMessageProperty.Label));
continue;
}
if (string.Compare(property.Key, ReplyToProperty, true) == 0)
{
brokeredMessageProperty.ReplyTo = message.Properties[ReplyToProperty]
as string;
Trace.WriteLineIf(isTrackingEnabled,
string.Format(PropertyAddedToList,
PropertiesToSend,
ReplyTo,

BrokeredMessagePropertySchemaNamespace,

brokeredMessageProperty.ReplyTo));
continue;
}
if (string.Compare(property.Key, ReplyToSessionIdProperty, true) == 0)

223
{
brokeredMessageProperty.ReplyToSessionId =
message.Properties[ReplyToSessionIdProperty] as string;
Trace.WriteLineIf(isTrackingEnabled,
string.Format(PropertyAddedToList,
PropertiesToSend,
ReplyToSessionId,

BrokeredMessagePropertySchemaNamespace,

brokeredMessageProperty.ReplyToSessionId));
continue;
}
if (string.Compare(property.Key, SessionIdProperty, true) == 0)
{
brokeredMessageProperty.SessionId =
message.Properties[SessionIdProperty] as string;
Trace.WriteLineIf(isTrackingEnabled,
string.Format(PropertyAddedToList,
PropertiesToSend,
SessionId,

BrokeredMessagePropertySchemaNamespace,

brokeredMessageProperty.SessionId));
continue;
}
if (string.Compare(property.Key, ToProperty, true) == 0)
{
brokeredMessageProperty.To = message.Properties[ToProperty] as
string;
Trace.WriteLineIf(isTrackingEnabled,
string.Format(PropertyAddedToList,
PropertiesToSend,
To,

BrokeredMessagePropertySchemaNamespace,

brokeredMessageProperty.To));
continue;
}


224
// Include properties indicated in the
// configuration of the message inspector.
string[] parts = property.Key.Split('#');
if (parts.Length == 2)
{
if (propertiesToSendDictionary.ContainsKey(parts[1]) &&
propertiesToSendDictionary[parts[1]] == parts[0])
{
brokeredMessageProperty.Properties[parts[1]] = property.Value;
Trace.WriteLineIf(isTrackingEnabled,
string.Format(PropertyAddedToList,

PropertiesToSend,
parts[0],
parts[1],

property.Value));
}
}
}

// Add the BrokeredMessageProperty to the WCF message properties
message.Properties[BrokeredMessageProperty.Name] = brokeredMessageProperty;
}
catch (Exception ex)
{
Trace.WriteLineIf(isTrackingEnabled, string.Format(ExceptionFormat,
ex.Message));
EventLog.WriteEntry(Source, ex.Message, EventLogEntryType.Error);
throw;
}
return Guid.NewGuid();
}
#endregion

#region Private Methods
//******************
// Private Methods
//******************

/// <summary>
/// Initialize the dictionary passed as second parameter

225
/// with the information read from the first argument.
/// In particular, the Name and Namespace properties of the
/// PropertiesElement object contain respectively
/// the comma-separated list of property names and namespaces
/// of a set of properties. The name of each property will
/// become the key in the dictionary, whereas the namespace
/// will become the corresponding value.
/// </summary>
/// <param name="dictionaryName">The name of the dictionary.</param>
/// <param name="properties">A PropertiesElement object.</param>
/// <param name="dictionary">The dictionary to populate.</param>
private void InitializeDictionary(string dictionaryName,
PropertiesElement properties,
Dictionary<string, string> dictionary)
{
if (properties == null)
{
Trace.WriteLineIf(isTrackingEnabled, PropertiesElementCannotBeNull);
throw new ApplicationException(PropertiesElementCannotBeNull);
}
if (!string.IsNullOrEmpty(properties.Name) &&
!string.IsNullOrEmpty(properties.Namespace))
{
var nameArray = properties.Name.Split(new[] { ',', ';', '|' });
var namespaceArray = properties.Namespace.Split(new[] { ',', ';', '|' });
if (namespaceArray.Length == nameArray.Length)
{
if (nameArray.Length > 0)
{
for (int i = 0; i < nameArray.Length; i++)
{
dictionary[nameArray[i]] = namespaceArray[i];
}
}
}
else
{
Trace.WriteLineIf(isTrackingEnabled,
string.Format(NameAndNamespaceNumberAreDifferent,
dictionaryName));
throw new
ApplicationException(string.Format(NameAndNamespaceNumberAreDifferent,

226
dictionaryName));
}
}
}

/// <summary>
/// Adds a new property to the specified list.
/// </summary>
/// <param name="name">The name of a property.</param>
/// <param name="ns">The namespace of a property.</param>
/// <param name="value">The value of a property.</param>
/// <param name="listName">The name of the list.</param>
/// <param name="list">The list to add the property to.</param>
private void AddItemToList(string name,
string ns,
object value,
string listName,
List<KeyValuePair<XmlQualifiedName, object>> list)
{
list.Add(new KeyValuePair<XmlQualifiedName, object>(new XmlQualifiedName(name,
ns), value));
Trace.WriteLineIf(isTrackingEnabled, string.Format(PropertyAddedToList,
listName,
name,
ns,
value));
}
#endregion
}
Component Description
The constructor of the message inspector reads the configuration data and initializes three
dictionaries that contain, respectively, the properties to promote, to write, and to send out.
The AfterReceiveRequest method is executed when a WCF-Custom receive location,
configured to use the message inspector, receives a message from the Service Bus. This
method performs the following actions:
a. Uses the current OperationContext to read the BrokeredMessageProperty from the
Properties collection of the WCF message,
b. Includes the BrokeredMessageProperty properties in the list of the properties to write to
the BizTalk message context
c. Reads the user-defined properties from the Properties collection of the
BrokeredMessageProperty and promote or write only those that have been specified in
The B
the m
follow
1. C
2. R
P
pr
3. A
P
4. F
Comp
You c
monit

the configu
Properties
BeforeSendRe
message inspe
wing actions:
Creates a new
Reads the con
roperties coll
roperties of th
Adds the custo
roperties coll
inally, adds th
ponent Track
can enable the
tor its runtime
uration of the
sToWrite pro
equest metho
ector, is about
w BrokeredMe
ntext propertie
ection of the
he BrokeredM
om context pr
ection of the
he BrokeredM
king
e component
e behavior.
WCF receive
operties of the
od is executed
t to send a m
essagePrope
es, defined in
WCF messag
MessagePrope
roperties, in o
BrokeredMes
MessageProp
tracking and
e location usin
e endpoint be
d when a WC
essage to the
rty object.
the Brokered
ge and assign
erty.
our sample de
ssageProperty
erty to the W
use DebugV
ng the Prope
havior.
CF-Custom se
e Service Bus
dMessagePr
ns their values
efined in the P
y.
CF message
View, as show
rtiesToProm
end port, conf
s. This metho
ropertySchem
s to the corre
PropertySche
properties.
wn in the pictu

22
mote and
igured to use
d performs th
ma, from the
esponding
ema, to the
re below, to

27

he
If you
BizTa
Serve
and, i
by the
W
S
X
an





stop both orc
alk Server, the
er Administra
n particular, t
e following co
WCF-Custom
erviceBusMe
Xml Disassem
nd 4).
chestrations a
e BizTalk Serv
ation Conso
to review the
omponents:
adapter (see
essageInspe
mbler compon
and use the c
ver is suspen
le to examine
properties pro
properties at
ector (see pro
nent within th
client applicat
nded. This sus
e the context
omoted or wr
points 2 and
operties at po
e XMLReceiv
ion to send a
spension allow
properties of
ritten to the m
5).
oint 3)
ve pipeline (s
request mes
ws you to use
the inbound d
message conte
see properties

22
sage to
e the BizTalk
document
ext, in order,
s at points 1

28
k
Inst
In ord
a BizT
1. In
2. R
The fi
autom
the ex
follow

An ea
GAC

gacuti
gacuti




talling Co
der to use the
Talk applicatio
nstall the corre
Register the co
irst step is pre
matically insta
xecution of th
wing picture:
asy and handy
is to use the f
il /lr Micro
il /lr Micro
omponent
ListenUriEn
on, we need t
esponding as
omponents in
etty straightfo
all your assem
e gacutil tool
y way to verif
following com
soft.Windows
soft.Windows
ts in GAC
ndpointBehav
to perform the
ssemblies in t
n the proper m
orward and ca
mbly to the GA
in the post-b
fy that the two
mmands:
Azure.CAT.Sa
Azure.CAT.Sa
C
vior and Serv
e following op
he global ass
machine.config
an be accomp
AC whenever
uild event com
o assemblies
amples.Servic
amples.Servic
viceBusMess
perations:
sembly cache
g file on the B
plished using
you build the
mmand-line o
have been su
ceBus.ListenU
ceBus.Message
sageInspect
e (GAC).
BizTalk Serve
the gacutil too
e class library
of the project,
uccessfully in
UriEndpointBe
eInspector

22
or classes in
er computer.
ol. You can
by including
as shown in

stalled in the
ehavior
29

230
Registering Components in the machine.config file
To accomplish this task you can proceed as follow:
Use a text editor such as Notepad to open the machine.config file, which can be found in
the following folders:
%windir%\Microsoft.NET\Framework\v4.0.30319\Config
%windir%\Microsoft.NET\Framework64\v4.0.30319\Config
Add the following items to the
<configuration><system.serviceModel>\<extensions><bindingElementExtensions> section.
These items contain the fully-qualified names (FQDN) of the behavior extension elements of
the two components:
<add name="serviceBusMessageInspector"
type="Microsoft.WindowsAzure.CAT.Samples.ServiceBus.MessageInspector.
ServiceBusMessageInspectorBehaviorExtensionElement,
Microsoft.WindowsAzure.CAT.Samples.ServiceBus.MessageInspector,
Version=1.0.0.0, Culture=neutral, PublicKeyToken=197ec3eb961f773c"/>
<add name="setListenUri"
type="Microsoft.WindowsAzure.CAT.Samples.ServiceBus.
ListenUriEndpointBehavior.ListenUriBehaviorExtensionElement,
Microsoft.WindowsAzure.CAT.Samples.ServiceBus.ListenUriEndpointBehavior,
Version=1.0.0.0, Culture=neutral, PublicKeyToken=197ec3eb961f773c"/>
<add name="setSessionChannel"
type="Microsoft.WindowsAzure.CAT.Samples.ServiceBus.
SessionChannelEndpointBehavior.SessionChannelBehaviorExtensionElement,
Microsoft.WindowsAzure.CAT.Samples.ServiceBus.
SessionChannelEndpointBehavior,
Version=1.0.0.0, Culture=neutral, PublicKeyToken=197ec3eb961f773c"/>





Receive Locations
We have now all the elements we need to create and configure the artifacts that compose our
BizTalk application. In this section we will describe the steps necessary to create the WCF-
Custom receive location used by the ServiceBusSample application to read messages from the
requestqueue, then well see how to define a WCF-Custom receive location to retrieve
messages from the ItalyMilan subscription defined on the requesttopic.
Lets start by looking at the steps necessary to create the WCF receive location used to read
request messages from the requestqueue.
O
th
C

R
m
S
S
ch
lik
S
W
S
Open the BizT
he ServiceBu
Create a new o
Right-click the
menu.
pecify a nam
ince in our sa
hoose the XM
ke the Messa
taticSendPo
WCF-Custom
pecify the UR
Talk Server A
usSample ap
one-way rece
newly create
e for the rece
ample we wan
MLReceive pi
geType that i
ortOrchestrat
adapter and
RL of the requ
Administratio
plication.
eive port as sh
ed receive por
eive location a
nt to process
peline to prom
is used in the
tion and Dyn
click the Con
uestqueue in
on Console a
hown in the fi
rt and select N
and choose a
incoming req
mote the cont
e subscription
amicSendPo
nfigure button
n the Address
and expand th
gure below:
New Receive
pipeline base
quests using a
text propertie
expression o
ortOrchestra
n.
s (URI) field in
he Receive P
e Location fro
ed on your re
an orchestrati
s of the inbou
of the
ation. Then ch
n the Genera

23
Ports node of
om the contex
equirements.
ion, we
und documen
hoose the
al tab.
31

xt
nt

232


Select the Binding tab, and choose the NetMessagingBinding from the Binding Type drop-
down list. Then set the binding properties in the Configuration property grid.

233


Select the Behavior tab, then right click the EndpointBehavior node and select the
transportClientEndpointBehavior component. Expand the corresponding node and enter
the shared secret credentials for your service namespace that you previously retrieved from
the Windows Azure Management Portal.

R
co
na
ne
S
Right click the
omponent. Th
amespace lis
eed to promo
taticSendPo
EndpointBe
hen click the P
ts of the user
ote the Metho
ortOrchestrat
ehavior node
PropertiesTo
r-defined prop
od property, w
tion and Dyn
and select th
oPromote no
perties that yo
which is used
amicSendPo
e serviceBus
ode and speci
ou want to pro
in the subscr
ortOrchestra
sMessageIns
fy the name a
omote. In our
ription express
ation.

23
spector
and
sample we
sion of
34


C
de
w


Click the Prop
efined proper
we define the f
MachineN
Country, Ci
equal to C
response t
Italy and M
reason, bo
the contex
message.
ertiesToWrit
rties that you
following prop
Name, UserNa
ity: since the
ountry='Italy
to the respon
Milan respect
oth StaticSen
xt properties, i
te node, and
want to prom
perties:
ame: these tw
e client applica
y' and City='M
nsetopic, the
tively,.in orde
ndPortOrches
including Cou
specify the na
mote to write to
wo properties
ation uses a s
Milan', when
client must s
r for the clien
stration and
untry and Cit
ame and nam
o the messag
s are optional
subscription w
the BizTalk a
set the Count
t to receive th
DynamicSen
ty, from the re
mespace lists
ge context. In
in our demo.
whose filter ex
application pu
try and City p
he message F
ndPortOrche
equest to the

23
of the user-
our sample
xpression is
blishes the
properties to
For this
estration cop
response
35

y

B
yo
W
qu
se
W
m
en
ecause the P
ou are definin
When the WCF
ueue, you mu
elect the setS
WCF-Custom
messages from
nd of this sec
PropertiesToS
ng a WCF-Cu
F-Custom rec
ust perform an
SessionChan
adapter uses
m a sessionfu
ction
Send propert
stom receive
ceive location
n additional s
nnel compone
s a session-aw
ul queue or su
ty applies only
location.
is configured
step: right clic
ent. At runtim
ware channel
ubscription. Fo
y to send por
d to receive m
k the Endpoi
e, this compo
(IInputSessio
or more inform
rts, it can be ig
messages from
ntBehavior n
onent makes
onChannel) to
mation, see th

23
gnored when
m a sessionfu
node, and
sure that the
o receive
he note at the
36

ul
e

C
C
Lets
mess
1. O
th
2. R
R
3. S
S
ch
lik
S
W
Click the OK to
Click the OK b
now go throu
ages from the
Open the BizT
he ServiceBu
Right-click the
Receive Loca
pecify a nam
ince in our sa
hoose the XM
ke the Messa
taticSendPo
WCF-Custom
o close the W
button to finish
gh the steps
e ItalyMilan s
Talk Server A
usSample ap
ServiceBusS
ation from the
e for the rece
ample we wan
MLReceive pi
geType that i
ortOrchestrat
adapter and
WCF-Custom a
h creating the
to configure t
subscription o
Administratio
plication.
Sample.OneW
e context men
eive location,
nt to process
peline to prom
is used in the
tion and Dyn
click the Con
adapter config
e WCF-Custom
the WCF rece
of the request
on Console a
Way.Receive
nu.
and choose a
incoming req
mote the cont
e subscription
amicSendPo
nfigure button
guration dialo
m receive loc
eive location u
ttopic.
and expand th
ePortreceive
a pipeline bas
quests using a
text propertie
expression o
ortOrchestra
n.
og.
cation.
used to read
he Receive P
port and sele
sed on your re
an orchestrati
s of the inbou
of the
ation. Then ch

23
request
Ports node of
ect New
equirements.
ion, we
und documen
hoose the
37

nt

4. Specify the UR RL of the requuesttopic in tthe Address (URI) field in the General

23
tab.
38


239


5. Select the Binding tab then choose NetMessagingBinding from the Binding Type drop-
down list. Then set the binding properties in the Configuration property grid.

240


6. Select the Behavior tab then right click the EndpointBehavior node and select the
transportClientEndpointBehavior component. Expand the corresponding node and enter
the shared secret credentials for your service namespace that you previously retrieved from
the Windows Azure Management Portal.

7. R
th
su
in
Right click the
he setListenU
ubscription to
nformation, se
EndpointBe
Uri node in th
o the ListenU
ee the note at
ehavior node
e Behavior p
ri property in
t the end of th
and select th
panel and ass
the Configu
his section.
e setListenU
sign the URL
ration proper
Uri componen
of the ItalyM
rty grid. For m

24
nt. Then click
ilan
more
41


8. R
co
na
th
S
Right click the
omponent. Th
amespace lis
he Method pr
taticSendPo
EndpointBe
hen click the P
ts of the user
roperty, which
ortOrchestrat
ehavior node
PropertiesTo
r-defined prop
h is used in th
tion and Dyn
and select th
oPromote no
perties to prom
he subscriptio
amicSendPo
e serviceBus
ode and speci
mote. In our s
n expression
ortOrchestra
sMessageIns
fy the name a
sample we ne
of the
ation.

24
spector
and
eed to promot
42

te

9. C
de
fo


Click the Prop
efined proper
ollowing prope
MachineN
Country, C
equal to C
response t
must set th
client to re
and Dynam
and City, f
ertiesToWrit
rties promote
erties:
Name, UserNa
City: since th
ountry='Italy
to the respon
he Country a
eceive the me
micSendPort
from the requ
te node and s
to write to the
ame: these tw
e client applic
y' and City='M
nsetopic, in o
and City prope
ssage. For th
tOrchestratio
est to the res
specify the na
e message co
wo properties
cation uses a
Milan', when
order for the c
erties to Italy
his reason, bo
on copy the c
sponse messa
ame and nam
ontext. In our
s are optional
subscription
the BizTalk a
client to receiv
y and Milan, r
oth the StaticS
context prope
age.
espace lists o
sample we d
in our demo.
whose filter e
application pu
ve the messa
respectively, i
SendPortOrc
erties, includin

24
of the user-
define the
expression is
blishes the
age, the client
n order for th
chestration
ng Country
43

e

10. B
W
11. W
su
an
th
m
en
ecause Prop
WCF-Custom
When the WCF
ubscription, y
nd select the
he WCF-Cust
messages from
nd of this sec
pertiesToSen
receive locati
F-Custom rec
you have to pe
setSessionC
om adapter u
m a sessionfu
ction.
nd applies onl
ion.
ceive location
erform an add
Channel com
uses a sessio
ul queue or su
y to send por
is configured
ditional step.
mponent. At ru
n-aware chan
ubscription. Fo
rts, it can be i
d to receive m
Right click the
untime, this co
nnel (IInputSe
or more inform
gnored when
messages from
e EndpointB
omponent ma
essionChanne
mation, see th

24
n you define a
m a sessionfu
Behavior node
akes sure that
el) to receive
he note at the
44

a
ul
e
t
e

12. C
13. C
If
to
re
The ad
"Syste
Intern
caused
sessio
8:43:4
Server
Except
at
at
Micros
el.
N
Click the OK b
Click the OK b
you make the
opic to which
eceive locatio
dapter "WCF-
em.ServiceMo
nal Server E
d by a prema
on shutdown.
47 AM
r stack trac
tion rethrow
Microsoft.S
soft.Service
Note
button to close
button to finish
e mistake of s
the subscript
on, you will se
Custom" rais
del.Communic
rror: The se
ture
TrackingId:1
e:
n at [0]:
erviceBus.Co
Bus.Messagin
e the WCF-Cu
h creating the
specifying the
ion belongs in
ee an error like
ed an error
ationExcepti
rver did not
025c2be-5f49
mmon.AsyncRe
ng.Sbmp.Duple
ustom adapte
e WCF-Custom
e URL of a su
n the Addres
e the followin
message. Det
ion:
t provide a m
9-4639-be01-4
esult.End[TAs
exRequestBind
er configuratio
m receive loc
ubscription ins
ss (URI) field o
g in the Appl
tails
meaningful re
4b2e5749c7ba,
syncResult](I
dingElement.D
on dialog.
cation.
stead of the U
of a WCF-Cu
lication Log.
eply; this mi
Timestamp:9
IAsyncResult
DuplexRequest

24
URL of the
stom
ight be
9/30/2011
result)
tSessionChann
45

n

246
DuplexCorrelationAsyncResult.End(IAsyncResult result)
at Microsoft.ServiceBus.Messaging.Channels.
ReconnectBindingElement.ReconnectChannelFactory`1.RequestSessionChannel.
RequestAsyncResult.b__13(RequestAsyncResult thisPtr, IAsyncResult r)
at Microsoft.ServiceBus.Messaging.IteratorAsyncResult`1.StepCallback(IAsyncResult
result)

Exception rethrown at [1]:
at Microsoft.ServiceBus.Messaging.Channels.
ServiceBusInputChannelBase`1.OnException(Exception exception)
at Microsoft.ServiceBus.Messaging.Channels.
ServiceBusInputChannelBase`1.OnEndTryReceive(IAsyncResult result,
BrokeredMessage&
brokeredMessage)
at Microsoft.ServiceBus.Messaging.Channels.
ServiceBusInputChannelBase`1.System.ServiceModel.Channels.
IInputChannel.EndTryReceive(IAsyncResult result, Message& wcfMessage)
at System.ServiceModel.Dispatcher.
InputChannelBinder.EndTryReceive(IAsyncResult result, RequestContext&
requestContext)
at System.ServiceModel.Dispatcher.
ErrorHandlingReceiver.EndTryReceive(IAsyncResult result, RequestContext&
requestContext)".
If you configure a WCF-Custom receive location to retrieve messages from a sessionful
queue or subscription and you forget to add setSessionChannel to the list of endpoint
behaviors used by the artifact, then, when you enable the receive location, you will see
an error like the following in the Application Log:
The adapter "WCF-Custom" raised an error message. Details
"System.InvalidOperationException: It is not
possible for an entity that requires sessions to create a non-sessionful message
receiver..
TrackingId:9163a41f-d792-406d-acbe-eb93ab2defb8_1_1,TimeStamp:10/3/2011 12:39:02 PM --->
System.ServiceModel.FaultException`1[System.ServiceModel.ExceptionDetail]: It is not
possible for an entity
that requires sessions to create a non-sessionful message receiver..
TrackingId:9163a41f-d792-406d-acbe-eb93ab2defb8_1_1,TimeStamp:10/3/2011 12:39:02 PM
Server stack trace:
at Microsoft.ServiceBus.Messaging.Sbmp.DuplexRequestBindingElement.
DuplexRequestSessionChannel.ThrowIfFaultMessage(Message wcfMessage)
at Microsoft.ServiceBus.Messaging.Sbmp.DuplexRequestBindingElement.
Note

247
DuplexRequestSessionChannel.HandleMessageReceived(IAsyncResult result)
Exception rethrown at [0]:
at Microsoft.ServiceBus.Common.AsyncResult.End[TAsyncResult](IAsyncResult result)
at Microsoft.ServiceBus.Messaging.Sbmp.DuplexRequestBindingElement.
DuplexRequestSessionChannel.DuplexCorrelationAsyncResult.End(IAsyncResult result)
at
Microsoft.ServiceBus.Messaging.Channels.ReconnectBindingElement.ReconnectChannelFactory`1
.
RequestSessionChannel.RequestAsyncResult.<GetAsyncSteps>b__13(RequestAsyncResult
thisPtr, IAsyncResult r)
at Microsoft.ServiceBus.Messaging.IteratorAsyncResult`1.StepCallback(IAsyncResult
result)
Exception rethrown at [1]:
at Microsoft.ServiceBus.Common.AsyncResult.End[TAsyncResult](IAsyncResult result)
at Microsoft.ServiceBus.Messaging.Channels.ReconnectBindingElement.
ReconnectChannelFactory`1.RequestSessionChannel.EndRequest(IAsyncResult result)
at Microsoft.ServiceBus.Messaging.Sbmp.SbmpMessageReceiver.
EndReceiveCommand(IAsyncResult result, IEnumerable`1& messages)
--- End of inner exception stack trace ---
Server stack trace:
at Microsoft.ServiceBus.Messaging.Sbmp.SbmpMessageReceiver.
EndReceiveCommand(IAsyncResult result, IEnumerable`1& messages)
at Microsoft.ServiceBus.Messaging.IteratorAsyncResult`1.StepCallback(IAsyncResult
result)
Exception rethrown at [0]:
at Microsoft.ServiceBus.Common.AsyncResult.End[TAsyncResult](IAsyncResult result)
at Microsoft.ServiceBus.Messaging.Sbmp.SbmpMessageReceiver.
OnEndTryReceive(IAsyncResult result, IEnumerable`1& messages)
at Microsoft.ServiceBus.Messaging.MessageReceiver.EndTryReceive(IAsyncResult result,
IEnumerable`1& messages)
at Microsoft.ServiceBus.Messaging.Channels.ServiceBusInputChannelBase`1.
OnEndTryReceive(IAsyncResult result, BrokeredMessage& brokeredMessage)
at
Microsoft.ServiceBus.Messaging.Channels.ServiceBusInputChannelBase`1.System.ServiceModel.
Channels.
IInputChannel.EndTryReceive(IAsyncResult result, Message& wcfMessage)
at System.ServiceModel.Dispatcher.InputChannelBinder.
EndTryReceive(IAsyncResult result, RequestContext& requestContext)
at System.ServiceModel.Dispatcher.ErrorHandlingReceiver.
EndTryReceive(IAsyncResult result, RequestContext& requestContext)".
The following picture shows the two receive locations we just created along with two other
request-response receive locations used that use the BasicHttpBinding and the
NetTc
locatio

Sen
In this
by the
mess
Static
entity
runtim
conte
Lets s
Static
O
S
C
th
th
ad
cpRelayBindin
ons is out of t
nd Ports
s section we w
e StaticSend
ages to the re
cSendPortOr
, while the Dy
me, based on
xt property, w
start by lookin
cSendPortOr
Open the BizT
erviceBusSa
Create a new S
he send port a
he PassThruT
dapter and cl
ng. As we sai
the scope of t
will describe t
PortOrchest
equestqueue
rchestration
ynamicSendP
the address s
where to send
ng at the step
rchestration
Talk Server A
ample applica
Static One-w
and choose a
Transmit pip
ick the Confi
id in the Scen
this article.
the steps nec
tration and D
e and reques
uses two stat
PortOrchest
specified by t
d the response
ps necessary
to send respo
Administratio
ation
way Send Por
a pipeline base
eline for perfo
gure button.
narios section
cessary to cre
DynamicSend
sttopic. In par
tic one-way s
ration uses a
the client app
e message.
to create the
onse messag
on Console a
rt as shown in
ed on your re
ormance reas
n, the descript
ate the WCF-
dPortOrchest
rticular, the
send ports, on
a dynamic sen
lication and c
WCF send p
ges to the res
and expand th
n the figure b
equirements. I
sons. Then ch
tion of these t
-Custom send
tration to sen
ne for each m
nd port to dec
contained in th
port used by th
ponsequeue
he Send Port
elow. Specify
In this demo,
hoose the WC

24
two receive

d ports used
nd response
essaging
cide at
he ReplyTo
he
e.
ts node of the
y a name for
we will use
CF-Custom
48
e

Specify the UR RL of the respponsequeue in the Addreess (URI) fieldd in the Gene

24
eral tab.
49


250


Select the Binding tab, and then choose NetMessagingBinding from the Binding Type drop-
down list. Then set the binding properties in the Configuration property grid.

251


Select the Behavior tab then right click the EndpointBehavior node and select the
transportClientEndpointBehavior component. Expand the corresponding node and enter
the shared secret credentials for your service namespace that you previously retrieved from
the Windows Azure Platform Management Portal.

R
co
lis
S
S
de
ap
se
w
co
Right click the
omponent. Th
sts of the use
erviceBusSa
ervice Bus m
emonstrates
pplication usi
erviceBusMe
would be able
ontain the Ap
EndpointBe
hen click the P
r-defined pro
ample applica
message. In pa
how a BizTal
ng a Service
essageInspe
to retrieve the
pplication, Co
ehavior node
PropertiesTo
perties to inc
ation includes
articular, the u
k application
Bus message
ector is not st
e message fr
ountry and C
and select th
oSend node a
lude in the ou
s the Applica
use of the Ap
can send con
e. Please not
rictly required
rom the respo
City properties
e serviceBus
and specify th
utgoing messa
ation, Countr
pplication con
ntext informat
te that in this c
d for routing p
onsequeue e
s.
sMessageIns
he name and
age. In our sa
ry and City in
ntext property
tion out-of-ba
case the use
purposes, as t
even if the me

25
spector
namespace
ample, the
the outgoing
y
nd to another
of the
the client
essage did no
52

g
r
ot

Select the Mes ssages tab thhen uncheck UUse Transacction in the Trransactions

25
section.
53


254


Press the OK button to close the WCF-Custom adapter configuration dialog.
Click the OK button to complete the creation of the WCF-Custom send port.
Now, lets go through the steps to create the WCF send port used by the
StaticSendPortOrchestration to send response messages to the responsetopic.
Open the BizTalk Server Administration Console and expand the Send Ports node of the
ServiceBusSample application.
Create a new static one-way send port as shown in the figure below. Specify a name for the
send port and choose a pipeline based on your requirements. In this demo, we will use the
P
ad

S
assThruTran
dapter and cl
pecify the UR
nsmit pipeline
ick the Confi
RL of the resp
e for performa
gure button.
ponsetopic in
ance reasons
n the Addres
s. Then choos
ss (URI) field
se the WCF-C
in the Genera

25
Custom
al tab.
55


S
do
elect the Bin
own list. Then
ding tab then
n set the bind
n choose NetM
ding properties
MessagingBin
s in the Conf
nding from th
figuration pro
e Binding Ty
operty grid.

25
ype drop-
56


257


Select the Behavior tab then right click the EndpointBehavior node and select the
transportClientEndpointBehavior component. Expand the corresponding node and enter
the shared secret credentials for your service namespace that you previously retrieved from
the Windows Azure Platform Management Portal.

258


Right click the EndpointBehavior node and select the serviceBusMessageInspector
component. Then click the PropertiesToSend node and specify the name and namespace
lists of the user-defined properties to include in the outgoing message. In our sample, the
ServiceBusSample application includes the values of the Application, Country, and City
properties in the outgoing Service Bus message. In particular, the use of the Application
context property demonstrates how a BizTalk application can send context information out-of-
band to another application through a Service Bus message. Please note that in this case the
use of the serviceBusMessageInspector is not strictly required for routing purposes as the
client would be able to retrieve the message from the responsequeue even if the message
did not contain the Application, Country, and City properties.

Select the Mes ssages tab thhen uncheck UUse Transacction in the Trransactions

25
section.
59


260


Click the OK button to close the WCF-Custom adapter configuration dialog.
Click the OK button to finish creating the WCF-Custom send port.
As the last step, lets create the dynamic send port used by the
DynamicSendPortOrchestration to send messages to the responsequeue or responsetopic,
based on the URL specified by the client application in the ReplyTo property of the Service Bus
message.
Open the BizTalk Server Administration Console and expand the Send Ports node of the
ServiceBusSample application.
C
th
th

C
The fo
Admi





Create a new d
he send port a
he PassThruT
Click the OK b
ollowing pictu
inistration Co
dynamic one-
and choose a
Transmit pip
button to comp
ure shows the
onsole.
-way send po
a pipeline base
eline for perfo
plete the crea
e three send p
ort as shown i
ed on your re
ormance reas
ation of the W
ports we creat
n the figure b
equirements. I
sons.
WCF-Custom s
ted in the Biz
below. Specify
In this demo,
send port.
zTalk Server

26
y a name for
we will use

61


262
Orchestrations
In this section we will examine the structure of the StaticSendPortOrchestration and
DynamicSendPortOrchestration and the code of the helper classes used by them.
This section covers the structure of the orchestrations and the code of the components they use.
In particular, with the DynamicSendPortOrchestration, well examine how to implement an
asynchronous, request-response pattern where the client application dynamically indicates to the
BizTalk application the address of the queue or topic where to send the response message.
StaticSendPortOrchestration
The following figure shows the structure of the StaticSendPortOrchestration.

The B
orche
that re
descr
for the
BusinessLog
estration invok
eturns the pay
ription of the R
e understand
gic expression
kes the Proce
yload of the r
RequestMan
ing of the ope
n shape at po
essRequestR
response mes
ager helper c
erating mode
oint 1 contains
ReturnStream
ssage as a str
component as
of the solutio
s the code in
m method on
ream. I will om
s the analysis
on.
the box below
a RequestM
mit for brevity
s of its code is

26

w. The
anager objec
y the
s unnecessary
63
ct
y

264
logHelper.WriteLine("[StaticSendPortOrchestration] Request message received from [{0}].",
requestMessage(WCF.To));
stream = requestManager.ProcessRequestReturnStream(requestMessage);
logHelper.WriteLine("[StaticSendPortOrchestration] Request message successfully
processed.");
The MessageAssignment shape at point 2 creates the response message.
responseMessage = null;
responseManager.SetResponse(responseMessage, stream);
responseMessage(*) = requestMessage(*);
guid = System.Guid.NewGuid();
responseMessage(Microsoft.WindowsAzure.CAT.Schemas.Application) = "ServiceBusSample";
responseMessage(Microsoft.WindowsAzure.CAT.Schemas.MessageId) = guid.ToString();
responseMessage(Microsoft.WindowsAzure.CAT.Schemas.CorrelationId) =
responseManager.GetMessageId(requestMessage);
responseMessage(Microsoft.WindowsAzure.CAT.Schemas.SessionId) =
responseManager.GetReplyToSessionId(requestMessage);
As you can notice, in this shape the orchestration performs the following actions:
Invokes the SetResponse method on a ResponseManager object to assign the stream
containing the payload to the response message.
Copies the context properties from the request to the response message.
Assign a value to the Application context property. In particular, the use of this context
property demonstrates how a BizTalk application can send context information out-of-band to
another application through a Service Bus message. In fact, as we have seen in the previous
section, the static send ports are configured to translate this context property into a user-
defined property of the outgoing BrokeredMessage.
Creates a MessageId for the response message.
Copies the MessageId from the request message to the CorrelationId of the response
message. This copy allows the client to correlate the response message with the initial
request.
Copies the ReplyToSessionId from the request message to the SessionId of the response
message. Please note that using sessions is necessary when multiple consumers use the
same sessionful queue or subscription to receive response messages. In this case, a
publisher application can use the ReplyToSessionId property of a BrokeredMessage to
communicate to the consumer the value to assign to the SessionId of response messages.
The queue rule of the code snippet at point 3 invokes the ReturnResponseViaQueue method of
the RequestManager object, which returns true if the value of the ReplyTo context property
contains the word queue or false otherwise. In other words, the orchestration checks the value
of the ReplyTo context property that contains the address specified by the client in the ReplyTo
property of the BrokeredMessage and chooses whether to send the response message to the
responsequeue or responsetopic based on the client demand.
Dyna
The fo

The B
orche
that re
descr
for the
amicSendP
ollowing figur
BusinessLog
estration invok
eturns the pay
ription of the R
e understand
PortOrches
e shows the s
gic expression
kes the Proce
yload of the r
RequestMan
ing of the ope
stration
structure of th
n shape at po
essRequestR
response mes
ager helper c
erating mode
he DynamicS
oint 1 contain
ReturnStream
ssage as a str
component as
of the solutio
SendPortOrc
ns the code in
m method on
ream. I will om
s the analysis
on.
hestration.
the box belo
a RequestM
mit for brevity
s of its code is

26

ow. The
anager objec
y the
s unnecessary
65
ct
y

266
logHelper.WriteLine("[DynamicSendPortOrchestration] Request message received from
[{0}].",
requestMessage(WCF.To));
stream = requestManager.ProcessRequestReturnStream(requestMessage);
logHelper.WriteLine("[DynamicSendPortOrchestration] Request message successfully
processed.");
The MessageAssignment shape at point 2 creates the response message.
responseMessage = null;
responseManager.SetResponse(responseMessage, stream);
responseMessage(*) = requestMessage(*);
guid = System.Guid.NewGuid();
responseMessage(Microsoft.WindowsAzure.CAT.Schemas.Application) = "ServiceBusSample";
responseMessage(Microsoft.WindowsAzure.CAT.Schemas.MessageId) = guid.ToString();
responseMessage(Microsoft.WindowsAzure.CAT.Schemas.CorrelationId) =
responseManager.GetMessageId(requestMessage);
responseMessage(Microsoft.WindowsAzure.CAT.Schemas.SessionId) =
responseManager.GetReplyToSessionId(requestMessage);
responseMessage(WCF.InboundHeaders) = "";
responseMessage(WCF.Action)= responseManager.GetAction(requestMessage);
responseMessage(WCF.BindingType)="netMessagingBinding";
responseMessage(WCF.EnableTransaction) = false;
responseMessage(WCF.BindingConfiguration) =
responseManager.GetBindingConfiguration(requestMessage);
responseMessage(WCF.EndpointBehaviorConfiguration) =
responseManager.GetEndpointBehaviorConfiguration(requestMessage);
OneWayResponsePort(Microsoft.XLANGs.BaseTypes.Address) =
responseManager.GetReplyTo(requestMessage);
OneWayResponsePort(Microsoft.XLANGs.BaseTypes.TransportType)="WCF-Custom";
As you can see in this shape the orchestration performs the following actions:
Invokes the SetResponse method on a ResponseManager object to assign the stream
containing the payload to the response message.
Copies the context properties from the request to the response message.
Assign a value to the Application context property. In particular, the use of this context
property demonstrates how a BizTalk application can send context information out-of-band to
another application through a Service Bus message. In fact, as we have seen in the previous
section, the static send ports are configured to translate this context property into a user-
defined property of the outgoing BrokeredMessage.
Creates a MessageId for the response message.
Copies the MessageId from the request message to the CorrelationId of the response
message. This allows the client to correlate the response message with the initial request.
Finally, the orchestration.

267
Copies the ReplyToSessionId from the request message to the SessionId of the response
message. Please note that sessions are necessary when multiple consumers use the same
sessionful queue or subscription to receive response messages. In this case, a publisher
application can use the ReplyToSessionId property of a BrokeredMessage to communicate to
the consumer the value to assign to the SessionId of response messages.
Invokes the GetAction method on the ResponseManager object to get the value for the
WCF.Action.
Specifies the NetMessagingBinding as WCF.BindingType for the dynamic send port.
Assigns false to the WCF.EnableTransaction property.
Invokes the GetBindingConfiguration method on the ResponseManager object to get the
value for the WCF.BindingConfiguration.
Invokes the GetEndpointBehaviorConfiguration method on the ResponseManager object
to get the value for the WCF.EndpointBehaviorConfiguration.
Invokes the GetReplyTo method on the ResponseManager object to get the address of the
dynamic send port.
Sets the TransportType of the dynamic send port to WCF-Custom.
For your convenience, I included the code of the ResponseManager class in the example below.
ResponseManager Class
public class ResponseManager
{
#region Private Fields
//***************************
// Private Fields
//***************************
private readonly LogHelper logHelper = new LogHelper();
#endregion

#region Public Methods
//***************************
// Public Methods
//***************************

/// <summary>
/// Sets the content of the XLANGMessage passed as first parameter.
/// </summary>
/// <param name="message">An XLANGMessage object.</param>
/// <param name="stream">The stream containing the payload to assign to the
message.</param>
public void SetResponse(XLANGMessage message, Stream stream)
{
try

268
{
if (stream != null &&
message != null &&
message.Count > 0)
{
if (stream.CanSeek)
{
stream.Seek(0, SeekOrigin.Begin);
}
message[0].LoadFrom(stream);
}
}
catch (Exception ex)
{
logHelper.WriteLine(string.Format("[ResponseManager] {0}", ex.Message));
}
finally
{
if (message != null)
{
message.Dispose();
}
}
}
/// <summary>
/// Gets the action for the response request.
/// </summary>
/// <param name="request">The request request.</param>
/// <returns>The action to assign to the Action header of the response
request.</returns>
public string GetAction(XLANGMessage request)
{
const string action = "ReceiveResponse";

try
{
// In a real application, the action should be retrieved from a
// configuration repository based on the request request content and context
// information. In addition, this data should be cached to improve
performance.
return action;
}

269
catch (Exception ex)
{
logHelper.WriteLine(string.Format("[ResponseManager] {0}", ex.Message));
}
finally
{
logHelper.WriteLine(string.Format("[ResponseManager] Action=[{0}]", action ??
"NULL"));
if (request != null)
{
request.Dispose();
}
}
// Return default Action.
return action;
}

/// <summary>
/// Gets the binding configuration for the response request.
/// </summary>
/// <param name="request">The request request.</param>
/// <returns>The binding configuration of the dynamic send port used to send
/// out the response request.</returns>
public string GetBindingConfiguration(XLANGMessage request)
{
const string bindingConfiguration = @"<binding name=""netMessagingBinding"" " +
@"sendTimeout=""00:03:00"" " +
@"receiveTimeout=""00:03:00"" " +
@"openTimeout=""00:03:00"" " +
@"closeTimeout=""00:03:00"" " +
@"sessionIdleTimeout=""00:01:00"" " +
@"prefetchCount=""-1""> " +
@"<transportSettings batchFlushInterval=""00:00:01"" />" +
@"</binding>";
try
{
// In a real application, the binding configuration should be retrieved from
a
// configuration repository based on the request request content and context
// information. In addition, this data should be cached to improve
performance.
return bindingConfiguration;

270
}
catch (Exception ex)
{
logHelper.WriteLine(string.Format("[ResponseManager] {0}", ex.Message));
}
finally
{
logHelper.WriteLine(string.Format("[ResponseManager]
BindingConfiguration=[{0}]",
bindingConfiguration ?? "NULL"));
if (request != null)
{
request.Dispose();
}
}
// Return default binding configuration.
return bindingConfiguration;
}

/// <summary>
/// Gets endpoint behavior configuration of the dynamic
/// send port used to send out the response request.
/// </summary>
/// <param name="request">The request request.</param>
/// <returns>The endpoint behavior configuration of the dynamic
/// send port used to send out the response request.</returns>
public string GetEndpointBehaviorConfiguration(XLANGMessage request)
{
const string endpointBehaviorConfiguration = @"<behavior
name=""EndpointBehavior"">" +
@"<transportClientEndpointBehavior>" +
@"<tokenProvider>" +
@"<sharedSecret issuerName=""owner""
issuerSecret=""2jdfAfOD7Jgd4WmnI7mJStz35hH7NDC7AN20MjbSpxs="" />" +
@"</tokenProvider>" +
@"</transportClientEndpointBehavior>" +
@"<serviceBusMessageInspector>" +
@"<propertiesToSend name=""Application,Country,City"" " +

@"namespace=""http://windowsazure.cat.microsoft.com/samples/servicebus/propertyschema," +
@"http://windowsazure.cat.microsoft.com/samples/servicebus/propertyschema," +
@"http://windowsazure.cat.microsoft.com/samples/servicebus/propertyschema"" />" +

271
@"</serviceBusMessageInspector>" +
@"</behavior>";
try
{
// In a real application, the endpoint behavior configuration should be
retrieved from a
// configuration repository based on the request request content and context
// information. In addition, this data should be cached to improve
performance.
return endpointBehaviorConfiguration;
}
catch (Exception ex)
{
logHelper.WriteLine(string.Format("[ResponseManager] {0}", ex.Message));
}
finally
{
logHelper.WriteLine(string.Format("[ResponseManager]
EndpointBehaviorConfiguration=[{0}]",
endpointBehaviorConfiguration ?? "NULL"));
if (request != null)
{
request.Dispose();
}
}
// Return default endpoint behavior configuration.
return string.Empty;
}

/// <summary>
/// Gets the value of the MessageId context property.
/// </summary>
/// <param name="request">The request request.</param>
/// <returns>The value of the MessageId context property.</returns>
public string GetMessageId(XLANGMessage request)
{
try
{
if (request != null)
{
var messageId = request.GetPropertyValue(typeof (MessageId)) as string;
if (!string.IsNullOrEmpty(messageId))

272
{
logHelper.WriteLine(string.Format("[ResponseManager]
MessageId=[{0}]", messageId));
return messageId;
}
}
}
catch (Exception ex)
{
logHelper.WriteLine(string.Format("[ResponseManager] Exception: {0}",
ex.Message));
}
finally
{
if (request != null)
{
request.Dispose();
}
}
return string.Empty;
}

/// <summary>
/// Gets the value of the ReplyTo context property.
/// </summary>
/// <param name="request">The request request.</param>
/// <returns>The value of the ReplyTo context property.</returns>
public string GetReplyTo(XLANGMessage request)
{
try
{
if (request != null)
{
var replyTo = request.GetPropertyValue(typeof(ReplyTo)) as string;
if (!string.IsNullOrEmpty(replyTo))
{
logHelper.WriteLine(string.Format("[ResponseManager] ReplyTo=[{0}]",
replyTo));
return replyTo;
}
}
}

273
catch (Exception ex)
{
logHelper.WriteLine(string.Format("[ResponseManager] Exception: {0}",
ex.Message));
}
finally
{
if (request != null)
{
request.Dispose();
}
}
return string.Empty;
}

/// <summary>
/// Gets the value of the ReplyToSessionId context property.
/// </summary>
/// <param name="request">The request request.</param>
/// <returns>The value of the ReplyToSessionId context property.</returns>
public string GetReplyToSessionId(XLANGMessage request)
{
try
{
if (request != null)
{
var replyToSessionId = request.GetPropertyValue(typeof(ReplyToSessionId))
as string;
if (!string.IsNullOrEmpty(replyToSessionId))
{
logHelper.WriteLine(string.Format("[ResponseManager]
ReplyToSessionId=[{0}]",
replyToSessionId));
return replyToSessionId;
}
return !string.IsNullOrEmpty(replyToSessionId) ? replyToSessionId :
string.Empty;
}
}
catch (Exception ex)
{

ex.Mes









}
#e
}

The fo
Cons
reque
orche

Test
Assum
as fol
T
M
T
M
T
bu
T
bu
T
Q
T
T
logHe
ssage));
}
finally
{
if (r
{
r
}
}
return st
endregion
ollowing pictu
sole along wit
ests coming fr
estration is ou
ting the S
ming that you
lows to test it
o send a requ
Method prope
o send a requ
Method prope
o send a requ
utton in the R
o send a requ
utton in the R
o ask BizTalk
Queue radio b
o ask BizTalk
opic radio bu
lper.WriteLi
equest != nu
equest.Dispo
ring.Empty;
ure shows the
th another orc
rom the reque
t of the scope
Solution
have already
t.
uest message
rty to StaticS
uest message
rty to Dynam
uest message
Request Meth
uest message
Request Meth
k Server to se
button in the R
k Server to se
utton in the Re
ne(string.Fo
ull)
se();
e two orchestr
chestration ca
est-response
e of this article
y deployed an
e to the Static
SendPort.
e to the Dyna
micSendPort.
e to BizTalk S
hods group
e to BizTalk S
hods group.
end the respo
Response Me
end the respo
esponse Met
ormat("[Respo
rations in the
alled RelaySe
receive locat
e.
nd properly co
cSendPortOr
amicSendPor
Server via the
Server using t
nse message
ethods group
nse message
thods group.
onseManager]
BizTalk Serv
erviceOrches
ions. The des
onfigured the
rchestration
rtOrchestrati
requestque
the requestto
e to the respo
p
e to the respo
Exception: {
ver Administ
stration used
scription of th
solution, you
, set the value
ion, set the v
ue, select the
opic, select th
onsequeue, s
onsetopic, se

27
0}",
tration
d to serve
is

u can proceed
e of the
value of the
e Queue radio
he Topic radi
select the
elect the
74
d
o
o
The fi

T
T
re
T
T
de
Looki
T
T
igure below s
he client send
he BizTalk ap
equesttopic u
he DynamicS
he client appl
efined on the
ng at the log,
he request m
he response
hows the mo
ds the reques
pplication rea
using the WC
SendPortOrc
lication receiv
responseto
you can note
message was s
message was
st interesting
st message to
ds the reques
CF-Custom re
chestration s
ves the respo
pic.
e the following
sent to the re
s received fro
combination:
o requesttop
st from the Ita
eceive location
sends the res
onse message
g:
equesttopic.
om the respo
:
ic.
alyMilan subs
n.
ponse messa
e through the
nsetopic.
scription for th
age to the res
ItalyMilan su

27

he
sponsetopic.
ubscription
75

T
m
T
in
th
W
S
flo
T
W
m
T
Imp
You c
BizTa
and s
Servic
notific
The fo
this sc
reque
receiv

To se
C
U
C
U
he Correlatio
message.
he Label pro
nto the Label
he Label cont
WCF send por
ervice Bus m
ow.
he orchestrat
WCF send por
message.
he Area prop
plementin
can use the S
alk receives a
ends a notific
ce Bus topic.
cations sent b
ollowing figur
cenario one in
ests to BizTalk
ve notification
et up this scen
Create the ind
UK.
Copy the clien
UK.
onId of the re
perty of the re
context prope
text property f
rt copied the v
message. The
tion added the
rt translated th
perty was add
g a Mess
erviceBusSa
request throu
cation to multi
In this contex
by BizTalk Se
e shows the a
nstance of the
k Server, whil
ns of a change
nario, you nee
ividual subsc
t applications
esponse mess
equest messa
erty in the WC
from the requ
value of the L
user defined
e Application
his property in
ded by the rule
sage Fan-
ample solutio
ugh a Service
iple front-offic
xt, each subs
rver.
architecture a
e client applic
le the other in
e of state (for
ed to proceed
riptions for th
s configuratio
sage is equal
age has been
CF receive loc
uest message
Label context
properties Co
n context pro
nto a user-de
e action defin
-Out Sce
on to simulate
e Bus queue o
ce application
criber uses a
a demo that s
cation simulat
nstances simu
r example, if t
d as follows:
e requesttop
on file to three
to the Messa
n translated by
cation. Then
e to the respo
property to th
ountry and C
perty to the re
efined property
ned on the Ita
nario
a message f
or topic from
ns located in d
different sub
imulates a me
tes a back-off
ulate front-off
the price of a
pics called, re
e different fold
ageId of the r
y the messag
the orchestra
nse message
he Label prop
City followed t
esponse mes
y of the Servi
lyMilan subs
fan-out scena
a back-office
different locat
bscription to re
essage fan-o
fice applicatio
fice applicatio
product has
espectively, U
ders called US

27
request
ge inspector
ation copied
e. Finally, the
perty of the
the same dat
ssage, and the
ice Bus
scription.
ario where
application
tion via a
eceive
ut scenario. I
on sending
ons that
changed).

US, China and
S, China and
76
ta
e
n
d

277
Open the configuration files of the three instances and change the listenUri of the
subscription service endpoint as follows:
US: sb://paolosalvatori.servicebus.windows.net/responsetopic/Subscriptions/US
China:
sb://paolosalvatori.servicebus.windows.net/responsetopic/Subscriptions/China
UK: sb://paolosalvatori.servicebus.windows.net/responsetopic/Subscriptions/UK
Use a forth instance of the client application to send a message to the BizTalk application,
and make sure to select Topic in the Response Methods.




Conclusion
In this article we have seen how to bridge the gap between BizTalk Server and the Service Bus
and how to achieve full interoperability between the two simply by using the native features
provided out-of-the-box by these technologies and by adding a couple of extensions and a pinch
of imagination. Is this the only way to integrate BizTalk Server with the Service Bus? The answer
is clearly no. An alternative solution is to build a custom Service Bus adapter by using the WCF
LOB Adapter SDK and the .NET Messaging API. This adapter should integrate the functionalities
that I described in this article and implemented in my code.
Another alternative, and an easier way to enable a BizTalk application to exchange messages
with Service Bus queues and topics, is to create a reusable helper class. This component should
use the .NET Messaging API and the BizTalk Runtime libraries to accomplish the following tasks:
Transform an XLANGMessage into a BrokeredMessage and vice versa. In particular, map
BizTalk message context properties into BrokeredMessage user-defined properties and vice
versa.
Receive messages from a Service Bus queue or subscription.
Send messages to a Service Bus queue or topic.
I look forward hearing your feedback. In the meantime, here you can download the companion
code for this article.



278
Managing and Testing Topics, Queues and Relay
Services with the Service Bus Explorer Tool
Authors: Paolo Salvatori
Introduced in the September 2011, queues and topics represent the foundation of a new cloud-
based messaging and integration infrastructure that provides reliable message queuing and
durable publish/subscribe messaging capabilities to both cloud and on-premises applications
based on Microsoft and non-Microsoft technologies. .NET applications can use the new
functionality offered by queues and topics by using the new messaging API
(Microsoft.ServiceBus.Messaging) released with the Windows Azure AppFabric SDK V1.5 or via
WCF by using the new NetMessagingBinding. Likewise, any Microsoft or non-Microsoft
applications can use a Service Bus REST API to manage and access messaging entities over
HTTPS.
Queues and topics were first introduced by the Community Technology Preview (CTP) of
Windows Azure AppFabric that was released in May 2011. At that time, the Windows Azure
Management Portal didnt provide any user interface to administer, create and delete messaging
entities and the only way to accomplish this task was using the .NET or REST API. For this
reason, In June 2011 I decided to build a tool called Service Bus Explorer that would allow
developers and system administrators to connect to a Service Bus namespace and administer its
messaging entities. During the last few months I continued to develop this tool and add new
features with the intended goal to facilitate the development and administration of new Service
Bus-enabled applications. In the meantime, the Windows Azure Management Portal introduced
the ability for a user to create queues, topics, and subscriptions and define their properties, but
not to define or display rules for an existing subscription. Besides, the Service Bus Explorer
enables to accomplish functionalities, such as importing, exporting and testing entities, that are
not currently provided by the Windows Azure Management Portal. For this reason, the Service
Bus Explorer tool represents the perfect companion for the official Azure portal and it can also be
used to explore the features (session-based correlation, configurable detection of duplicate
messages, deferring messages, etc.) provided out-of-the-box by the Service Bus brokered
messaging.
This post is the first of a two-part article where I will explain the functioning and implementation
details of my tool whose source code is available on MSDN Code Gallery for use or modification
by any developer. In particular, in the first part I will explain how to use my tool to manage and
test Queues and Topics whereas in the second post I will go through the code and explain the
approach I followed to realize the application.
For more information on the AppFabric Service Bus, you can use the following resources:
Service Bus topic on the MSDN Site.
Now Available: The Service Bus September 2011 Release article on the Windows Azure
Blog.
Queues, Topics, and Subscriptions article on the MSDN site.
AppFabric Service Bus topic on the MSDN site.
Understanding Windows Azure AppFabric Queues (and Topics) video on the AppFabric
Team Blog.
B
on
S
S
Que
Queu
applic
reliab

Queu
size o
produ
unlim
follow
T
W
A
Queu
S
ea
R
D
m
D
D
m
w
m
a
In the
variou
capab
uilding loosel
n the channe
ervice Bus To
ecuring Servi
eues
es provide m
cations runnin
le fashion acr
es are hosted
of a queue du
uction. The ma
ited-size sequ
wing APIs:
he new mess
WCF via the n
An extension o
e entities pro
ession-based
asy way.
Reliable delive
Detection of in
multiple times
Dead-letter fac
Deferring mess
messages are
while the proce
messages nee
traffic peak.
e .NET API the
us properties
bilities such a
ly-coupled ap
l9 site.
opics And Qu
ice Bus with A
essaging cap
ng on premise
ross network
d in Windows
ring the CTP
aximum mess
uences of rela
saging API av
ew ServiceBu
of the Service
vide the follow
d correlation,
ery patterns su
bound messa
without adver
cility for mess
sages for late
received out
ess waits for a
ed to be proce
e message en
such as Mes
s automatic d
pps with Wind
ueues video o
ACS video o
pabilities that
es or in the clo
and trust bou
Azure by a r
is 100MB, wh
sage size is 2
ated message
vailable via the
usMessaging
e Bus REST A
wing capabilit
meaning that
uch as Peek/
age duplicate
rse conseque
sages that can
er processing
of the expect
a particular m
essed based o
ntity is model
sageId, Sess
duplicate dete
ows Azure Se
on the channe
n the channe
enable a larg
oud to exchan
undaries.
eplicated, dur
hich is expec
256KB, but th
es. Queues fu
e new Micros
Binding
API.
ties:
t you can buil
/Lock.
s, allowing cl
ences.
nnot be proce
. This function
ted sequence
message to pe
on a set of pr
ed by the Bro
ionID and Co
ection or sess
ervice Bus To
el9 site.
l9 site.
ge and hetero
nge message
rable store in
ted to rise to
e use of sess
unctionality ar
soft.ServiceBu
d multiplexed
ients to send
essed or expir
nality is partic
e and need to
ermit further p
roperties that
okeredMessa
orrelationId th
sion-enabled c
opics and Qu
geneous clas
es in a flexible
frastructure. T
1GB (or more
sions allows c
re accessible
us.Messaging
d request/repl
the same me
re before bein
cularly handy
o be safely pu
progress or w
define their p
ge class that
at can be use
communicatio

27
eues video
ss of
e, secure and

The maximum
e) for
creating
e through the
g assembly
y paths in a
essage
ng received.
when
t on the side
when
priority during
exposes
ed to exploit
ons. A
79

m

280
QueueDescription object can be used to specify the metadata that models the behavior of the
queue being created:
The DefaultMessageTimeToLive property specifies the default message time to live value.
The DuplicateDetectionHistoryTimeWindow property defines the duration of the duplicate
detection history.
The EnableDeadLetteringOnMessageExpiration property allows to enable\disable the dead-
lettering on message expiration.
The LockDuration property defines the duration of the lock used by a consumer when using
the PeekLock receive mode. In the ReceiveAndDelete receive mode, a message is deleted
from the queue as soon as it is read by a consumer. Conversely, in the PeekLock receive
mode, a message is hidden from other receivers until the timeout defined by the
LockDuration property expires. By that time the receiver should have deleted the message
invoking the Complete method on the BrokeredMessage object.
The MaxQueueSizeInBytes property defines the maximum queue size in bytes.
The RequiresDuplicateDetection property enables\disables duplicate message detection.
Since metadata cannot be changed once a messaging entity is created, modifying the
duplicate detection behavior requires deleting and recreating the queue. The same principle
applies to any other metadata.
The RequiresSession property enables\disables sessions.
The use of Queues permits to flatten a highly-variable traffic into a predictable stream of work and
distribute the load across a set of worker processes which size can vary dynamically to
accommodate the incoming message volume. In a Competing Consumers scenario, when a
publisher writes a message to a queue, multiple consumers compete with each other to receive
the message, but only one of them will receive and process the message in question.
In a service-oriented or a service bus architecture composed of multiple, heterogeneous systems,
interactions between autonomous systems are asynchronous and loosely-coupled. In this
context, the use of messaging entities like Queues and Topics (see the next section) allows to
increase the agility, scalability and flexibility of the overall architecture and helps decreasing the
loose coupling of individual systems.
For more information on Queues, see the following articles:
Introducing the Windows Azure AppFabric Service Bus May 2011 CTP article on the
AppFabric Team Blog.
Understanding Windows Azure AppFabric Queues (and Topics) video on the AppFabric
Team Blog.
An Introduction to Service Bus Queues article on the AppFabric Team Blog.
Windows Azure AppFabric Service Bus Queues and Topics on Neil Mackenzie's Blog.
Windows Azure AppFabric Learning Series available on CodePlex.
Topics
Topics extend the messaging features provided by Queues with the addition of Publish-Subscribe
capabilities.

A Top
(this n
mess
entitie
the su
SqlFil
O
S
Conve
filter e
clause
S
S
Each
can p
Queu
have
mess
broad
sets o
The fo
Azure
U
T
A
W
W
Que
In ord
first o
name
Mana
the fig
pic entity cons
number is sub
age copies to
es. Each Rule
ubscription an
lterExpression
OrderTotal > 5
hipDestinatio
ersely, the Sq
expression is
e of an UPDA
ET AuditReq
ET Priority =
matching rule
otentially gen
es, Topics su
a single cons
ages on a firs
dcast messag
of competing w
ollowing secti
e AppFabric S
Understanding
eam Blog.
An Introduction
Windows Azur
Windows Azur
eues, Top
der to use the
peration to pe
espace to inclu
agement Porta
gure below.
sists of a sequ
bject to vary in
o a poll of wor
e specifies a f
nd a filter actio
n class allows
5000 OR Clien
onCountry = U
qlFilterAction
true and the
ATE SQL-com
uired = 1
'High', Sever
e generates a
nerate more c
upport a Com
sumer that rec
st-come-first-s
es to many c
worker proce
ion provides b
Service Bus. F
g Windows Az
n to Service B
re AppFabric
re AppFabric
pics and S
Brokered an
erform is to p
ude the Servi
al respectively
uential messa
n the future) c
rker processe
filter expressio
on that can m
s you to defin
ntPriority > 2
USA AND Sh
class can be
message sele
mmand.
rity = 1
a separate co
copies of the s
peting Consu
ceives all mes
served basis.
onsumer app
sses.
basic informa
For more info
zure AppFabr
Bus Topics a
Service Bus Q
Learning Ser
Subscrip
d Relay mess
rovision a new
ice Bus. You
y by clicking t
age store like
concurrent an
es. Each Subs
on that is use
modify messag
ne a SQL92-lik

hipDestination
used to mod
ected, using a
opy of the pub
same messag
umers scenar
ssages or a s
Topics are th
plications or d
tion on the Q
rmation, see
ric Queues (a
rticle on the A
Queues and T
ries available
tions
saging functio
w Service Bu
can accompli
the New butto
e a Queue, bu
nd durable su
scription can
ed to filter mes
ge properties
ke condition o
nState = WA
ify, add or rem
a syntax simi
blished messa
ge, one for ea
io: in this con
set of competi
he ideal mess
istribute the w
Queue and To
the following
nd Topics) v
AppFabric Te
Topics on Ne
e on CodePlex
onality provide
s namespace
ish this task f
on or the Mod
ut it supports u
bscriptions w
define one or
ssages that p
. In particular
on message p

move propert
lar to that use
age, so any su
ach matching
text, a subsc
ing consumer
saging solutio
work load acro
pic entities in
articles:
video on the A
am Blog.
eil Mackenzie
x.
ed by the Ser
e or modify an
from the Wind
dify button hig

28

up to 2000
which relay
r multiple Rule
pass through
r, the
properties:
ties, as the
ed by the SET
ubscription
rule. As
ription can
rs that fetch
on to
oss multiple
Windows
AppFabric
e's Blog.
rvice Bus, the
n existing
dows Azure
ghlighted in
81
e
T
e

Once
have
T
ex
qu
ap
E
T
m
H
on
fr
na
Finally
Azure
highlig
entity
displa
press
completed th
many options
his first option
xposed by the
ueues, topics
pproach that
xplorer tool.
he second ap
messaging ent
HTTP requests
n non-Micros
om the Servic
amespace.
y, you can ad
e Managemen
ghted in red a
. Then you ca
ay its propertie
the Delete b
his step, you c
s to provision
n is to exploit
e Namespace
s, subscription
I followed to b
pproach is to
tities. By usin
s, without the
oft technolog
ce Bus, but a
dminister mes
nt Portal. In p
at point 1 in th
an use the na
es in the verti
button in the M
can start crea
and manage
the CreateQ
eManager cla
ns, and rules,
build the prov
rely on the Se
g REST, you
e need for a cl
ies (e.g. Java
lso to create
ssaging entitie
articular, the
he figure belo
avigation tree-
ical bar highli
Manage Entit
ating and usin
messaging e
ueue, Create
ass that can b
in your servic
visioning func
ervice Bus RE
can write ap
lient SDK. In
a, Ruby, etc.)
or delete que
es from the us
buttons in the
ow allow creat
-view shown a
ghted at poin
ties command
ng queues, top
entities:
eTopic and Cr
be used to ma
ce namespac
ctionality expo
EST API to cr
plications in a
particular, thi
not only to se
eues, topics a
ser interface s
e Manage En
ting a new qu
at point 2 to s
nt 3. To remov
d bar.
pics and subs
reateSubscrip
anage entities
ce. This is exa
osed by the S
reate and del
any language
s allows appl
end and rece
nd subscriptio
supplied by th
ntities comma
ueue, topic an
select an exis
ve the selecte

28

scriptions. Yo
ption methods
s, such as
actly the
ervice Bus
ete
that supports
ications base
ive messages
on in a given
he Windows
and bar
nd subscriptio
ting entity and
ed entity, you
82
ou
s
s
ed
s
n
d

Using
mess
of ope
quite
create
rules
.NET
AddR
Subsc
GetRu
Porta
P
M
yo
na
cr
E
bi
se




In
E
E
g Windows Az
aging entities
erations that a
limited. For e
e queues, top
for an existin
Messaging A
Rule(String, Fi
criptionClient
ules method o
l actually doe
roperly visua
Management P
ou can organ
ame as an ab
rm/prod/queu
xport the mes
inding file (a-l
elect and exp
Individual e
Entities by
Entities co
All the enti
ndividual entit
ntities by type
ntities contain
zure Managem
s in a given Se
a developer o
example, the W
pics, and subs
g subscription
API. In particu
lter) or the Ad
class, while t
of the Names
es not provide
lize entities in
Portaldisplays
ize messagin
bsolute path c
ues/orderqueu
ssaging entiti
la BizTalk Se
port
entities.
y type (queues
ontained in a c
ities in a give
ties.
e (queues or
ned in a certa
ment Portal is
ervice Bus na
or a system ad
Windows Azu
scriptions and
n. At the mom
ular, to add a n
ddRule(RuleD
to enumerate
spaceManage
e the ability to
n a hierarchic
s queues, top
ng entities in a
composed of
ue.
es contained
rver). Instead
s or topics).
certain path (e
n namespace
topics).
ain path (e.g.
s a handy and
amespace. Ho
dministrator c
ure Managem
d define their
ment, you can
new rule to an
Description) m
the rules of a
er class. Besid
perform the f
al manner. Ac
pics and subsc
a hierarchical
multiple path
in a given Se
d, the Service
e.g. crm/prod
e.
crm/prod/que
d convenient m
owever, at lea
can perform u
ent Portalactu
properties, bu
n accomplish t
n existing sub
methods expo
an existing su
des, the Wind
following ope
ctually, the W
criptions with
structure sim
segments, fo
ervice Bus na
e Bus Explore
d/queues).
eues).
manner to ha
ast at the mom
using its user
ually allows a
ut not to creat
this task only
bscription you
osed by the
ubscription, yo
dows Azure M
erations:
Windows Azure
a flat treeview
mply by specif
or example
amespace to a
r tool provide

28

andle the
ment, the set
interface is
a user to
te or display
by using the
u can use the
ou can use th
Management
e
w. However,
fying their
an XML
es the ability to
83
he
o

284
All the entities in a given namespace.
Import queues, topics and subscriptions from an XML file to an existing namespace. The
Service Bus Explorer supports the ability to export entities to an XML file and reimport them
on the same or another Service Bus namespace. This feature comes in a handy to perform a
backup and restore of a namespace or simply to transfer queues and topics from a testing to
a production namespace.
Thats why in June, I created a tool called Service Bus Explorer that allows a user to create,
delete and test queues, topics, subscriptions, and rules. My tool was able to manage entities in
the AppFabric Labs Beta environment. However, the new version of the Service Bus API
introduced some breaking changes, as you can read here, so I built a new version of the Service
Bus Explorer tool that introduces a significant amount of new features.
Solution
The following picture illustrates the high-level architecture of the Service Bus Explorer tool. The
application has been written in C# using Visual Studio 2010 and requires the installation of the
.NET Framework 4.0 and Windows Azure AppFabric SDK V1.5. The tool can be copied and used
on any workstation that satisfies the prerequisites mentioned above to manage and test the
Brokered and Relay messaging services defined in a given Service Bus namespace.

In the
Con
When
mana
opens
Servic
crede
e remainder of
nnecting t
n you launch t
age its entities
s up a modal
ce Bus name
entials.
f the article I'l
to an exis
the applicatio
s by choosing
dialog shown
space that yo
ll explain how
sting Ser
n, you can co
g the Connect
n in the picture
ou want to con
w to configure
rvice Bus
onnect to an e
t command un
e below that a
nnect to and t
and use the
s Namesp
existing Servi
nder the File m
allows the use
the correspon
Service Bus E
pace
ce Bus name
menu. This o
er to enter the
nding authent

28

Explorer tool.
espace to
peration
e name of the
tication
85
.
e

286


The authentication and authorization to use a service exposed by the Service Bus are controlled
by the Access Control Service. In particular, the Service Bus currently supports three different
types of credential schemes:
SharedSecret, a slightly more complex but easy-to-use form of username/password
authentication.
Saml, which can be used to interact with SAML 2.0 authentication systems.
SimpleWebToken, which uses the OAuth Web Resource Authorization Protocol (WRAP) and
Simple Web Tokens (SWT).
The Service Bus Explorer tool currently supports only SharedSecret credentials. However, you
can easily extend its code to support other credential schemes. You can retrieve the issuer-secret
key for a given namespace from the Windows Azure Management Portal by clicking the View
button highlighted in red in the picture below.

This o
button

B
For yo
conne
snipp
n
N
opens up a m
n highlighted
y convention
our convenien
ection string f
et below, eac
amespace: th
Note
modal dialog w
in red in the p
, the name of
nce, the tool g
for the most fr
ch connection
his property d
where you can
picture below
f the Default
gives the pos
requently acc
n string is com
defines the Se
n retrieve the
.
Issuer is alwa
sibility to defi
cessed names
mposed of 4 e
ervice Bus na
key by clickin
ays the owne
ne, in the con
spaces. As yo
lements:
amespace.
ng the Copy t
er.
nfiguration file
ou can see in

28

to Clipboard

e, a
the xml
87


288
servicePath: this property is optional and specifies the service path within the namespace.
issuerName: this field contains the Issuer Name. As mentioned above, by convention, the
default value for this field is owner.
issuerSecret: this element contains the Issuer Secret that you can copy from the Windows
Azure Platform Management Portal.

<?xml version="1.0"?>
<configuration>
<configSections>
<section name="serviceBusNamespaces"
type="System.Configuration.DictionarySectionHandler,
System, Version=4.0.0.0, Culture=neutral,
PublicKeyToken=b77a5c561934e089"/>
</configSections>
<serviceBusNamespaces>
<add key="Namespace1" value="namespace=namespace1;
servicePath=;issuerName=owner;issuerSecret=..."/>
<add key="Namespace2" value="namespace=namespace2;servicePath=;
issuerName=owner;issuerSecret=..."/>
</serviceBusNamespaces>
<appSettings>
<add key="debug" value="true"/>
<add key="scheme" value="sb"/>
<add key="message" value="Hi mate, how are you?" />
<add key="file" value="C:\Sample.xml" />
<add key="label" value="Service Bus Explorer" />
<add key="retryCount" value="10" />
<add key="retryTimeout" value="100" />
<add key="messageDeferProvider"
value="Microsoft.AppFabric.CAT.WindowsAzure.Samples.ServiceBusExplorer.InMemoryMessageDef
erProvider,ServiceBusExplorer"/>
</appSettings>
<system.net>
<connectionManagement>
<add address="*" maxconnection="50"/>
</connectionManagement>
</system.net>
<startup>
<supportedRuntime version="v4.0" sku=".NETFramework,Version=v4.0"/>
</startup>
</configuration>


289

If you define one or more namespaces in the configuration file, this gives you the ability to select
one of them from the corresponding drop-down list and the fields within the modal dialog will be
automatically populated with the corresponding connection string data.


In the appSettings section of the configuration file you can define the following default settings:
debug: some components use a custom trace listener to write message in the log view. By
setting this parameter to false, you can disable this behavior.
scheme: this parameter can be used to define the transport protocol that the tool uses to
communicate with the Service Bus. At the moment, the only supported scheme is sb.
message: you can use this parameter to specify the default text for the test message.
file: this parameter can be used to define the location of the default message when testing
queues, topics and relay services.
label: this setting defines the default value for the Label property of a brokered message.
retryCount: this parameter can be used the default retry count in case of exception.
retryTimeout: this parameter allows to control the time in milliseconds between two
consecutive attemps.
messageDeferProvider: when testing deferred messages, the application uses a default
provider to store in-process the sequence number of deferred messages. You can create an
alternative provider and replace the default one. In this case, you have to use this parameter
to declare its fully-qualified name.
Gra
The G
1. T
In
en
2. T
br
ac
3. T
se
of
en
S
4. T
ha
re
re
th

Once
follow
Men
The M
F
phical Us
GUI of the Ser
he Menu bar
n particular, th
ntity.
he Service B
rowse and se
ccess a conte
he Main pane
elected by the
f relay service
ndpoints defin
erviceRegistr
he Log pane
andy especia
elay services
eceiver tasks.
he View menu
you connecte
wing operation
nu Bar
Menu bar con
ile:
ser Interf
rvice Bus Exp
contains the
he commands
Bus Namespa
elect entity no
ext-menu whi
el displays the
e user, it prov
es, the Servic
ned in the Se
rySettings end
l displays info
ally when you
to trace send
. The Log Pan
u.
ed to an exist
ns.
ntains the follo
face
plorer is comp
main menu t
s displayed in
ace panel con
odes. Each no
ch command
e information
vides the abili
ce Bus Explor
ervice Bus reg
dpoint behav
ormation and
use the Serv
d and receive
nel can be hid
ting Service B
owing menus:
posed of follow
hat allows the
n the Actions
ntains a Tree
ode is right-cli
s depend on
for the select
ty to create, d
rer tool only s
gistry by a giv
ior.
error messag
vice Bus Explo
operations pe
dden by dese
Bus namespa
:
wing parts:
e user to sele
menu vary b
View control
ck enabled a
the selected
ted entity. Ba
delete or test
supports the p
en relay serv
ges. The log f
orer to test ex
erformed by i
electing the Lo
ace, you can u
ect commands
based on the s
that allows to
nd offers the
entity.
sed on the co
the current e
possibility to te
vice using the
functionality c
xisting queues
ndividual sen
og Window o
use the tool to

29
s and options
selected
o navigate,
possibility to
ommand
ntity. In case
est the servic
comes in
s, topics and
nder and
option under

o perform the
90
s.
ce

291


Connect: opens a modal dialog that allows to connect to a given Service Bus
namespace.
Exit: closes the application.
Edit:


Clear Log: clears the log view.
Save Log As: saves the content of the log to a text file.
Actions: contains the same commands available in the context menu of the currently
selected node in the treeview.
View:


Set Default Layout: reset the layout to the default configuration
Log Window: enable or disable the display of the log window.
Options: shows a modal dialog that allows to define the layout parameters:

292


Help: contains the About Service Bus Explorer command.
Namespace Commands
By right clicking the namespace root node in the Service Bus Namespace panel, you can open
the following context menu. The commands provided by the context menu are also available
under the Actions menu in the Menu bar.


The context menu provides access to the following commands:

293
Delete Entities: as shown in the picture below, prompts the user to confirm whether to delete
all the entities in the current Service Bus namespace.


Refresh Entities: refreshes all the entities (queues, topics, subscriptions, rules, relay
services) in the current Service Bus namespace.
Export Entities: exports all the entities in the current Service Bus namespace to an XML file.
Import Entities: import entities from an XML file. The file in question can contain the
definition of one or multiple queues and topics.
Expand Subtree: expands the subtree below the namespace root node.
Collapse Subtree: collapse the subtree below the namespace root node.
Queues Commands
By right clicking the Queues node in the Service Bus Namespace panel, you can open the
following context menu. The commands provided by the context menu are also available under
the Actions menu in the Menu bar.

294


The context menu provides access to the following commands:
Create Queue: displays a custom control in the Main Panel that allows to specify the properties
of a new queue. For more information on how to create a new queue, see later in this article.
Delete Queues: prompts the user to confirm whether to delete all the queues in the current
Service Bus namespace.


Refresh Queues: refreshes all the queues in the current Service Bus namespace.
Export Queues: exports all the queues in the current Service Bus namespace to an XML file.
Expand Subtree: expands the subtree below the Queues node.
Collapse Subtree: collapse the subtree below the Queues node.

295
Queue Path Segment Commands
By right clicking a Path Segment node in the Queues subtree, as shown in the picture below,
you can open a context menu. The commands provided by the context menu are also available
under the Actions menu in the Menu bar.


The context menu provides access to the following commands:
Create Queue: displays a custom control in the Main Panel that allows to specify the
properties of a new queue. In particular, it initializes the value of the queue Name field with
the absolute path of the selected node (e.g. crm/prod in the sample above). For more
information on how to create a new queue, see later in this article.
Delete Queues: as shown in the picture below, prompts the user to confirm whether to delete
all the queues in the current path.


E
E
C
Que
If you
displa

By rig
The c
Menu
xport Queue
xpand Subtr
Collapse Sub
eue Comm
click an exist
ayed in the Ma
ght clicking a Q
commands pro
u bar.
es: exports all
ree: expands
tree: collapse
mands
ting Queue n
ain panel, as
Queue node
ovided by the
l the queues
the subtree b
e the subtree
node in the Se
shown in the
in the Queue
e context men
in the current
below the sele
below the se
ervice Bus N
e picture below
es subtree, yo
nu are also av
t path to an X
ected path se
elected path s
amespace p
w.
ou can open th
vailable under
XML file.
egment node.
egment node
anel, its infor
he following c
r the Actions

29
e.
mation will be

context menu
menu in the
96
e
.

297


The context menu provides access to the following commands:
Delete Queue: as shown in the picture below , prompts the user to confirm whether to delete
the selected queue in the current Service Bus namespace.


Test Queue: displays a custom control in the Main Panel that allows to test the selected
queue by sending and receiving messages. The tool allows the user to select a wide range of
testing options. In addition, the Service Bus Explorer enables to monitor the progress and
performance of an ongoing test run by using the information traced in the log view. Finally,
the tool allows to display real-time performance data, such as latency and throughput for

298
sender and receiver tasks, simply by enabling statistics and chart features. For more
information on how to test a queue, see later in this article
Refresh Queue: refreshes the current information for the selected queue. This feature can
be used to refresh the value of the MessageCount and SizeInBytes properties for the current
queue.
Export Queue: exports the current queue definition to an XML file.
Copy Queue Url: as shown in the picture below, this command copies the url of the current
queue to the clipboard.


Copy Deadletter Queue Url: as shown in the picture below, this command copies the url of
the deadletter queue for the current queue to the clipboard.


Expand Subtree: expands the subtree below the Queues node.
Collapse Subtree: collapse the subtree below the Queues node.
Topics Commands
By right clicking the Topics node in the Service Bus Namespace panel, you can open the
following context menu. The commands provided by the context menu are also available under
the Actions menu in the Menu bar.

299


The context menu provides access to the following commands:
Create Topic: displays a custom control in the Main Panel that allows to specify the
properties of a new topic. For more information on how to create a new topic, see later in this
article.
Delete Topics: as shown in the picture below, prompts the user to confirm whether to delete
all the topics in the current Service Bus namespace.


Refresh Topics: refreshes all the topics in the current Service Bus namespace.
Export Topics: exports all the topics in the current Service Bus namespace to an XML file.
Expand Subtree: expands the subtree below the Topics node.
Collapse Subtree: collapse the subtree below the Topics node.

300
Topic Path Segment Commands
By right clicking a Path Segment node in the Topics subtree, as shown in the picture below, you
can open a context menu. The commands provided by the context menu are also available under
the Actions menu in the main Menu bar.


The context menu provides access to the following commands:
Create Topic: displays a custom control in the Main Panel that allows to specify the
properties of a new topic. In particular, it initializes the value of the topic Name field with the
absolute path of the selected node (e.g. erp/test in the sample above). For more information
on how to create a new topic, see later in this article.
Delete Topics: as shown in the picture below, prompts the user to confirm whether to delete
all the topics in the current path.


Expo
Expa
Colla
Top
If you
displa

By rig
The c
Menu
rt Topics: ex
nd Subtree:
pse Subtree
ic Comm
click an exist
ayed in the Ma
ght clicking a T
commands pro
u bar.
xports all the t
expands the
: collapse the
mands
ting Topic no
ain panel, as
Topic node in
ovided by the
topics in the c
subtree below
e subtree belo
de in the Serv
shown in the
n the Topics s
e context men
current path to
w the selected
ow the selecte
vice Bus Nam
picture below
subtree, you c
nu are also av
o an XML file
d path segme
ed path segm
mespace pane
w.
can open the
vailable under
.
ent node.
ment node.
el, its informa
following con
r the Actions

30
ation will be

ntext menu.
menu in the
01

302


The context menu provides access to the following commands:
Delete Topic: as shown in the picture below, prompts the user to confirm whether to delete
the selected topic in the current Service Bus namespace.


Test Topic: displays a custom control in the Main Panel that allows to test the selected topic
by sending and receiving messages. The tool allows the user to select a wide range of testing
options. In addition, the Service Bus Explorer enables to monitor the progress and

303
performance of an ongoing test run by using the information traced in the log view. Finally,
the tool allows to display real-time performance data, such as latency and throughput for
sender and receiver tasks, simply by enabling statistics and chart features. For more
information on how to test a queue, see later in this article.
Refresh Topic: refreshes the current information for the selected topic. This feature can be
used to refresh the value of the SizeInBytes property for the current topic.
Export Topic: exports the current topic definition to an XML file along with its subscriptions.
Create Subscription: displays a custom control in the Main Panel that allows to specify the
properties of a new subscription for the selected topic. For more information on how to create
a new subscription, see later in this article.
Delete Subscriptions: as shown in the picture below, prompts the user to confirm whether to
delete all the subscriptions for the selected topic.


Copy Topic Url: as shown in the picture below, this command copies the url of the current
topic to the clipboard.
Expand Subtree: expands the subtree below the Topics node.
Collapse Subtree: collapse the subtree below the Topics node.
Subscriptions Commands
By right clicking a Subscriptions node as shown in the picture below, you can open the following
context menu. The commands provided by the context menu are also available under the
Actions menu in the Menu bar.

304


The context menu provides access to the following commands:
Create Subscription: displays a custom control in the Main Panel that allows to specify the
properties of a new subscription for the current topic. For more information on how to create a
new subscription, see later in this article.
Delete Subscriptions: as shown in the picture below, prompts the user to confirm whether to
delete all the subscriptions for the current topic.

305


Refresh Subscriptions: refreshes the current information for the subscriptions of the current
topic.
Expand Subtree: expands the subtree below the Subscriptions node.
Collapse Subtree: collapse the subtree below the Subscriptions node.
Subscription Commands
By right clicking a Subscription node as shown in the picture below, you can open the following
context menu. The commands provided by the context menu are also available under the Actions
menu in the Menu bar.


The context menu provides access to the following commands:

306
Delete Subscription: as shown in the picture below, prompts the user to confirm whether to
delete the selected subscription.


Test Subscription: displays a custom control in the Main Panel that allows to test the
selected subscription by receiving messages from it. The tool allows the user to select a wide
range of testing options. In addition, the Service Bus Explorer enables to monitor the
progress and performance of an ongoing test run by using the information traced in the log
view. Finally, the tool allows to display real-time performance data, such as latency and
throughput for receiver tasks, simply by enabling statistics and chart features. For more
information on how to test a queue, see later in this article.
Refresh Subscription: refreshes the current information for the selected subscription. This
feature can be used to refresh the value of the MessageCount property for the current
subscription.
Add Rule: displays a custom control in the Main Panel that allows to specify the properties of
a new rule for the currently selected subscription. For more information on how to create a
new topic, see later in this article.
Copy Queue Url: as shown in the picture below, this command copies the url of the current
queue to the clipboard.


Copy Deadletter Queue Url: as shown in the picture below, this command copies the url of the
deadletter queue for the current queue to the clipboard.

E
C
Rule
By rig
menu
in the


The c
xpand Subtr
Collapse Sub
es Comm
ght clicking a R
u. The comma
Menu bar.
context menu
ree: expands
tree: collapse
mands
Rules node a
ands provided
provides acc
the subtree b
e the subtree
as shown in th
d by the conte
cess to the fol
below the Sub
below the Su
he picture belo
ext menu are
lowing comm
bscription nod
ubscription no
ow, you can o
also available
mands:
de.
ode
open the follo
e under the A


30

owing context
Actions menu
07
u

308
Add Rule: displays a custom control in the Main Panel that allows to specify the properties of
a new rule for the currently selected subscription. For more information on how to create a
new topic, see later in this article.
Delete Rules: as shown in the picture below, prompts the user to confirm whether to delete
all the rules for the current subscription.


Refresh Rules: refreshes the current information for the rules of the current topic.

Rule Commands
By right clicking a Rule node as shown in the picture below, you can open the following context
menu. The commands provided by the context menu are also available under the Actions menu
in the Menu bar.

309


The context menu provides access to the following commands:
Remove Rule: as shown in the picture below, prompts the user to confirm whether to remove the
selected rule.




310
Relay Services Commands
By right clicking the Relay Services node in the Service Bus Namespace panel, you can open the
following context menu. The commands provided by the context menu are also available under
the Actions menu in the Menu bar.


The context menu provides access to the following commands:
Refresh Relay Services: refreshes all the relay services in the current Service Bus namespace.
In order to appear in the treeview, a relay service needs to use the ServiceRegistrySettings
endpoint behavior to define its public endpoints in the Service Bus registry.
Expand Subtree: expands the subtree below the Relay Services node.
Collapse Subtree: collapse the subtree below the Relay Services node.

Relay Service Path Segment Commands
By right clicking a Path Segment node in the Relay Service subtree, as shown in the picture
below, you can open the following context menu. The commands provided by the context menu
are also available under the Actions menu in the Menu bar.

311


The context menu provides access to the following commands:
Expand Subtree: expands the subtree below the selected path segment node.
Collapse Subtree: collapse the subtree below the selected path segment node.
Relay Service Commands
By right clicking a Relay Service node as shown in the picture below, you can open the following
context menu. The commands provided by the context menu are also available under the Actions
menu in the Menu bar.

312


The context menu provides access to the following commands:
Expand Subtree: expands the subtree below the selected path segment node.
Collapse Subtree: collapse the subtree below the selected path segment node.
Relay Service Endpoint Commands
If you click an Endpoint node of a relay service in the Service Bus Namespace panel, the
following custom control appears in the Main panel. This control allows to configure a client
endpoint and send messages to the selected endpoint. For more information on how to test a
relay service, see later in this article

By rig
The c
Menu

The c
ght clicking an
commands pro
u bar.
context menu
n Endpoint no
ovided by the
provides acc
ode of a relay
e context men
cess to the fol
service, you
nu are also av
lowing comm
can open the
vailable under
mands:
e following con
r the Actions

31

ntext menu.
menu in the

13
C
cu


View
To vie
corres

The c
queue
Crea
To cre
segm
1. R
2. C
This o
user t
Copy Relay S
urrent service
w a Queu
ew the proper
sponding nod
current version
e once create
ate a Que
eate a new qu
ent node und
Right-click the
Choose Create
operation disp
to specify the
ervice Url: a
e endpoint to
ue
rties of an exi
de in the treev
n of the Servi
ed. The same
eue
ueue in the cu
der the Queue
node and ch
e Queue com
plays a custom
following pro
s shown in th
the clipboard
sting queue i
view.
ice Bus does
principle app
urrent names
es subtree an
oose Create
mmand under
m control (see
operties for th
he picture belo
.
n the Main Pa
not allow to c
plies to the oth
pace, you ca
nd then perfor
Queue from
the Actions m
e the picture
e new queue
ow, this comm
anel, you can
change the pr
her entities (to
n select the Q
rm one of the
the context m
menu.
below) in the
:
mand copies t

n just select th
roperties of a
opics, subscr
Queues node
following act
menu.
Main Panel t

31
the url of the
he

n existing
riptions, rules
e or a path
tions:
that allows the
14
).
e

P
ab
D
T
du
D
fo
L
pr
re
as
is
B
th
M
m
M
m
E
va
E
fo
le
R
R
de
R
pr
The M
queue
ath: this field
bsolute path l
Duplicate Det
imeSpan valu
uration of the
Default Messa
or the Default
ock Duration
roperty that d
eceive mode.
s soon as it is
s hidden from
y that time th
he BrokeredM
Max Delivery
message is au
Max Size In M
megabytes, wh
nable Batch
alue that indic
nable Dead
or the EnableD
ettering functio
Requires Dup
RequiresDuplic
etection featu
Requires Ses
roperty that e
MessageCoun
e. In addition,
defines the p
like erp/prod/
tection Histo
ue for the Dup
duplicate de
ge Time To L
MessageTime
n: this section
defines the du
In the Receiv
s read by a co
other receive
e receiver sh
Message objec
Count: the M
utomatically de
Megabytes: th
hich is the siz
ed Operation
cates whethe
Lettering on
DeadLettering
onality on me
plicate Detec
cateDetection
ure.
sion: this che
enables or dis
nt and SizeInB
since metad
path for the q
queues/order
ory Time Win
plicateDetecti
tection history
Live: this sect
eToLive prop
n gives the po
uration of the
veAndDelete
onsumer. Con
ers until the ti
ould have de
ct or called th
MaxDeliveryCo
eadlettered a
he MaxSizeIn
ze of memory
ns: the Enabl
r server-side
Message Ex
gOnMessageExp
essage expira
tion: this che
n property tha
eckbox specif
sables session
Bytes are rea
ata cannot be
ueue within th
rqueue.
dow: this sec
ionHistoryTim
y.
ion gives the
perty that spec
ossibility to en
lock used by
receive mode
nversely, in th
meout define
eleted the mes
he Abandon m
ount property
after this numb
Megabytes p
allocated for
eBatchedOpe
batched oper
xpiration: this
piration prop
ation.
eckbox specifi
at enables or
fies a Boolean
ns support on
ad-only prope
e changed on
he namespac
ction gives the
meWindow pro
possibility to
cifies the defa
nter a TimeSp
a consumer w
e, a message
he PeekLock
d by the Lock
ssage invokin
method to rele
y defines the m
ber of deliveri
roperty define
the queue.
erations prop
rations are en
s checkbox sp
perty that ena
ies a Boolean
disables the d
n value for the
n the queue b
rties and can
nce a messag
ce. The value
e possibility to
operty that de
enter a Time
ault message
pan value the
when using th
e is deleted fro
receive mode
kDuration pro
ng the Comple
ease the lock.
maximum del
ies.
es the maxim
erty accepts a
nabled.
pecifies a Boo
bles or disabl
n value for the
duplicate mes
e RequiresSe
eing created.
not be set wh
ging entity is c

31

can be an
o enter a
efines the
eSpan value
time to live.
LockDuration
he PeekLock
om the queue
e, a message
perty expires
ete method o

livery count. A
um size in
a Boolean
olean value
les the dead-
e
ssage
ession

hen creating a
created,
15
n
e

.
n
A
a
modif
princi
fields,
Dele
To de
perfor
R
C
C
Test
To tes
and th
R
C
This o
M
S
R
G
Mess

The M
M
T
bu
fying the dupl
ple applies to
, when you cl
ete a Que
elete an existi
rm one of the
Right-click the
Choose Delete
Click the Delet
t a Queue
st a queue in
hen perform o
Right-click the
Choose Test Q
operation sho
Message Tab
ender Tab
Receiver Tab
Graph Tab
sage Tab
Message tab a
Message Text
he tool provid
utton and sele
icate detectio
o the other en
ick the Create
eue
ng queue, yo
following act
node and ch
e Queue com
te button in th
e
the current n
one of the foll
node and ch
Queue comm
ows a custom
allows the use
t: specifies th
des the ability
ecting a file.
on behavior re
tities. If you d
e button, the
u can select t
tions:
oose Delete
mmand under
he Main Pane
amespace, y
owing actions
oose Test Qu
mand under th
control in the
er to specify t
e payload of
y to read the m
equires deletin
dont explicitly
default value
the correspon
Queue from t
the Actions m
el.
ou can select
s:
ueue from the
e Actions me
e Main panel t
the following
a BrokeredM
message from
ng and recrea
y specify any
will be used.
nding node in
the context m
menu.
t the correspo
e context men
enu.
that is compo
information:
essage as a
m the file syste
ating a queue
value for one

the treeview
menu.
onding node i
nu.
osed of four ta
TEXT or XML
em by pressin

31
e. The same
e or more
and then
n the treeview
abs:

L message.
ng the Open
16
w
M
ad
al
fe
C
D
pr
Speci
at a s
A
ea
to
sy
oc
si
S
co
Send

The S
E
M
ob
is
S
ob
ex
th
N
Message Prop
dded to the P
llows to defin
eature of a Br
Carry the paylo
Define applica
rocess the cu
ify filter and a
ubscription le
As indicated in
ach property
o the entire he
ystem proper
ccupied by pr
ize of 256K. If
erializationEx
ondition..
der Tab
Sender tab all
nabled: this c
MessageId: th
bject. When a
s used to indiv
essionId: thi
bject. A Sess
xample, the s
hat these mes
Note
perties: you c
Properties dict
e application
okeredMessa
oad of a mess
tion specific p
urrent messag
action express
evel.
n Windows Az
is 32K. Cumu
eader of the B
ties (such as
roperties coun
f an applicatio
xception exce
ows the user
checkbox allo
his field specif
a queue is co
viduate duplic
s field specifi
ionId is a log
sender can se
ssages are re
can use the d
tionary of a B
specific mess
age entity as
sage. In this c
properties tha
ge.
sions that can
zure AppFabr
ulative size of
BrokeredMess
SequenceNu
nts towards th
on exceeds a
eption is gene
to specify the
ows enabling
fies a value fo
nfigured to us
cates or alrea
es a value fo
ical and lightw
et SessionId =
elated to each
datagrid to def
BrokeredMess
sage properti
user-defined
context, the b
at can be used
n be used to d
ric Service Bu
f all properties
sage, which h
umber,Label,
he overall size
any of the limi
erated, so you
e following inf
or disabling s
or the Messag
se the duplica
dy received m
r the Session
weight way to
= "SessionId"
other.
fine the key\v
sage object. T
es. This is pr
properties ca
body of the me
d by a worker
define routing
us Quotas, the
s cannot exce
has both user
MessageId, a
e of the mess
ts mentioned
u should expe
formation:
sender tasks
geId property
ate detection
messages.
Id property of
o correlate me
of several m
value pairs tha
The Properties
robably the m
an be used for
essage could
r process to d
and data enr
e maximum s
eed 64K. This
r properties as
and so on). T
sage and its m
above, a
ect to handle t
when testing
of a Brokere
mechanism, t
f a BrokeredM
essages to ea
essages as a

31
at will be
s collection
ost important
r the following
d be empty.
decide how to
richment rules
size for
s applies
s well as
he space
maximum
this error

a queue.
dMessage
this property
Message
ach other. For
an indicator
17
t
g:
o
s
r

318
CorrelationId: the CorrelationId property can be used to implement a request-reply message
exchange pattern where the client application uses the MessageId property of an outgoing
request message and the CorrelationId property of an incoming response message to
correlate the two messages.
Label: specifies a value for the Label property of a BrokeredMessage object.
ReplyTo: the ReplyTo property can be used to define the address of the queue to reply to. In
an asynchronous request-reply scenario where a client application A sends a request
message to a server application B via a Service Bus queue or topic and waits for a response
message, by convention the client application A can use the ReplyTo property of the request
message to indicate to the server application B the address of the queue or topic where to
send the response.
ReplyToSessionId: by definition, the ReplyToSessionId property of a BrokeredMessage
object can be used by a client application to indicate to a service the session identifier to
reply to. When a client application A sends a flow of request messages to a server application
B via a session-enabled queue or topic and waits for the correlated response messages on a
separate session-enabled queue or a subscription, the client application A can assign the id
of the receive session to the ReplyToSessionId property of outgoing messages to indicate to
the application B the value to assign to the SessionId property of response messages.
Task Count: this field can be used to define the number of publisher tasks that will be used
to send messages. Every task runs on a separate .NET thread pool thread.
Message Count: this integer field can be used to specify the number of messages to write to
the queue.
Time to Live: specifies the value in seconds for the TimeToLive property of a
BrokeredMessage object. This is the duration after which the message expires, starting from
when the message is sent to the Service Bus. Messages older than their TimeToLive value
will expire and no longer be retained in the message store. Subscribers will be unable to
receive expired messages.
Use Transaction: when this option is checked, publisher tasks will use a TransactionScope
object to send a batch of messages within a transaction context. For more information, see
the code snippet in the table below.
Commit Transaction: when this option is checked, publisher tasks will invoke the Complete
method to commit the transaction and persist messages to the queue. Please note that when
a publisher sends one or multiple messages within a TransactionScope and at the end does
not call the Complete method, messages are not written to the queue. The Use Transaction
and Commit Transaction options can be used to test the behavior of send-side transactions.
For more information, see the code snippet in the table below.

using (var scope = new TransactionScope())
{
serviceBusHelper.SendMessages(...);
...
if (checkBoxSenderCommitTransaction.Checked)
{
scope.Complete();
...

319
}
...
}


Enable Logging: when this option is checked, sender tasks will track the operations
performed in the log view.
Enable Verbose: when this option is checked, sender tasks will log the payload and
properties of transmitted messages.
Enable Statistics: when this option is checked, sender tasks will update statistics in the
Graph tab.
Enable Graph: when this option is checked, sender tasks will update the latency and
throughput charts in the Graph tab.
One Session/Task: when this option is checked, each sender task will use a different value
for the SessionId property. This feature allows to test session-enabled queues with multiple
sender and consumer tasks.
Update MessageId: when this option is checked, the tool will generate a different MessageId
for each outbound message. You can disable this option when testing the duplicate message
detection feature of a queue.
Add Message Number: when this option is checked, sender tasks will add the actual
message number as a user-defined property to the Properties collection of the outgoing
messages.
Send WCF Messages: when this option is checked, outgoing messages are wrapped in a
SOAP envelope. This feature can be used to send messages to one or multiple consumers
that use WCF and NetMessagingBinding to receive messages from a queue. The following
table shows the code used by sender tasks to create a WCF message from a
BrokeredMessage object used as template. As you can see, the WCF message needs to be
encoded using a message encoder that depends on the protocol scheme used by a
consumer to receive and decode the message.

/// <summary>
/// Create a BrokeredMessage for a WCF receiver.
/// </summary>
/// <param name="messageTemplate">The message template.</param>
/// <param name="taskId">The task Id.</param>
/// <param name="updateMessageId">Indicates whether to use a unique id for each
message.</param>
/// <param name="oneSessionPerTask">Indicates whether to use a different session for
each sender task.</param>
/// <param name="messageText">The message text.</param>
/// <param name="to">The name of the target topic or queue.</param>
/// <returns>The cloned BrokeredMessage object.</returns>
public BrokeredMessage CreateMessageForWcfReceiver(BrokeredMessage messageTemplate,

320
int taskId,
bool updateMessageId,
bool oneSessionPerTask,
string messageText,
string to)
{

if (!IsXml(messageText))
{
throw new ApplicationException(MessageIsNotXML);
}

MessageEncodingBindingElement element;
if (scheme == DefaultScheme)
{
element = new BinaryMessageEncodingBindingElement();
}
else
{
element = new TextMessageEncodingBindingElement();
}
using (var stringReader = new StringReader(messageText))
{
using (var xmlReader = XmlReader.Create(stringReader))
{
using (var dictionaryReader =
XmlDictionaryReader.CreateDictionaryReader(xmlReader))
{
var message = Message.CreateMessage(MessageVersion.Default, "*",
dictionaryReader);
message.Headers.To = new Uri(namespaceUri, to);
var encoderFactory = element.CreateMessageEncoderFactory();
var encoder = encoderFactory.Encoder;
var outputStream = new MemoryStream();
encoder.WriteMessage(message, outputStream);
outputStream.Seek(0, SeekOrigin.Begin);
var outboundMessage = new BrokeredMessage(outputStream, true)
{
ContentType = encoder.ContentType
};
if (!string.IsNullOrEmpty(messageTemplate.Label))
{

321
outboundMessage.Label = messageTemplate.Label;
}
outboundMessage.MessageId = updateMessageId ?
Guid.NewGuid().ToString() : messageTemplate.MessageId;
outboundMessage.SessionId = oneSessionPerTask ? taskId.ToString() :
messageTemplate.SessionId;
if (!string.IsNullOrEmpty(messageTemplate.CorrelationId))
{
outboundMessage.CorrelationId = messageTemplate.CorrelationId;
}
if (!string.IsNullOrEmpty(messageTemplate.ReplyTo))
{
outboundMessage.ReplyTo = messageTemplate.ReplyTo;
}
if (!string.IsNullOrEmpty(messageTemplate.ReplyToSessionId))
{
outboundMessage.ReplyToSessionId =
messageTemplate.ReplyToSessionId;
}
outboundMessage.TimeToLive = messageTemplate.TimeToLive;
foreach (var property in messageTemplate.Properties)
{
outboundMessage.Properties.Add(property.Key, property.Value);
}
return outboundMessage;
}
}
}
}


Rece

The R
E
W
to
T
to
R
ou
T
M
S
se
A
F
D
de
S
co
ob

...
var sq


sqlFil
filter
...
if (fi
eiver Tab
Receiver tab a
nabled: this c
When using th
o uncheck this
ask Count: th
o receive mes
Receive Time
ut occurs. Wh
his value corr
MessageRece
ession Time
ession before
AcceptMessag
ilter: the filter
Deadletter Que
eadletter que
qlExpression
onsumer app
bject matches
qlFilter = n


lter.Validat
r = sqlFilte
ilter.Match(
allows the use
checkbox allo
e Service Bu
s option.
his field can b
ssages from th
eout: specifies
hen this happ
responds to th
iver and Que
eout: specifies
e a time out oc
geSession me
r field can be
eue options to
ue. This field
property of a
lication can u
s a given con
ew SqlFilter


e();
r.Preprocess
inboundMessa
er to specify t
ows enabling
s Explorer too
be used to de
he queue.
s the time spa
pens, consum
he TimeSpan
eueClient clas
s the time spa
ccurs. This va
ethod expose
used in conju
o determine,
accepts the S
anSqlFilter ob
use a SqlFilter
dition.
r(!string.IsN
? txtFilter
: DefaultFi
();
ge))
the following i
or disabling r
ol to send me
efine the numb
an in seconds
er tasks stop
n parameter o
sses.
an in seconds
alue correspo
d by the Que
unction with th
respectively,
SQL92 stand
bject. Note: as
r to check at r
NullOrEmpty(t
rExpression.T
ilterExpressi
information:
receiver tasks
essages to a r
ber of consum
s that consum
receiving me
of the Receive
s that consum
onds to the Ti
ueClient clas
he Defer Mes
the message
ard syntax ac
s shown in the
runtime wheth
txtFilterExpr
Text
ion);
s when testing
real applicatio
mer tasks that
mer tasks wait
essages from
e method exp
mer tasks wait
meSpan para
s.
ssage and Mo
es to defer or
ccepted by th
e following co
her a Brokere
ression.Text)

32

g a queue.
on, make sure
t will be used
t before a time
the queue.
osed by the
t for a new
ameter of the
ove to
move to
e
ode snippet, a
edMessage

22
e

e
a

323
{
inboundMessage.Defer();
...
}


Prefetch Count: specifies a value for the PrefetchCount property.of the MessageReceiver
objects used by receiver tasks to consume messages from the current queue. This number
indicates the number of messages that the message receiver can simultaneously request.
Receive Mode: indicates the receive mode (PeekLock or ReceiveAndDelete) used by
consumer tasks.
Use Transaction: when this option is checked, consumer tasks will use a TransactionScope
object to receive a batch of messages within a transaction context. For more information, see
the code snippet in the table below.
Commit Transaction: when this option is checked, the consumer tasks will invoke the
Complete method to commit the transaction and delete messages from the queue. Please
note that when a consumer task receives one or multiple messages within a
TransactionScope and at the end doesnt call the Complete method, messages are not delete
from the queue. The Use Transaction and Commit Transaction options can be used to test
the behavior of receive-side transactions. For more information, see the code snippet in the
table below.

using (var scope = new TransactionScope())
{
serviceBusHelper.ReceiveMessages(...);
...
if (checkBoxReceiverCommitTransaction.Checked)
{
scope.Complete();
...
}
...
}


Enable Logging: when this option is checked, receiver tasks will track the operations
performed in the log view.
Enable Verbose: when this option is checked, receiver tasks will log the payload and
properties of transmitted messages.
Enable Statistics: when this option is checked, receiver tasks will update statistics in the
Graph tab.
Enable Graph: when this option is checked, receiver tasks will update the latency and
throughput charts in the Graph tab.

324
Move to Deadletter queue: when this option is checked, consumer tasks will move
messages matching the SQL expression contained in the Filter field to the deadletter queue.
Read from Deadletter queue: when this option is checked, consumer tasks will read
messages from the deadletter queue. Consider disabling sender tasks when using this
option.
Defer Message: when this option is selected, consumer tasks will defer messages matching
the SQL expression defined in the Filter field. Deferred messages will be processed at the
end after the messages that do not match the filter expression. In order to save the
SequenceNumber of deferred messages, consumer tasks use the following helper class. You
can replace this component with another class that implements the custom
IMessageDeferProvider interface. In this case, you need to indicate the fully-qualified-name
of the new provider class in the messageDeferProvider setting in the configuration file.

public interface IMessageDeferProvider
{
int Count { get; }
void Enqueue(long sequenceNumber);
bool Dequeue(out long sequenceNumber);
}

public class InMemoryMessageDeferProvider : IMessageDeferProvider
{
#region Private Fields
private readonly Queue<long> queue = new Queue<long>();
#endregion

#region Public Properties
public int Count
{
get
{
lock (this)
{
return queue.Count;
}
}
}
#endregion

#region Public Methods
public bool Dequeue(out long sequenceNumber)
{
sequenceNumber = -1;

325
lock (this)
{
if (queue.Count > 0)
{
sequenceNumber = queue.Dequeue();
return true;
}
}
return false;
}

public void Enqueue(long sequenceNumber)
{
lock (this)
{
queue.Enqueue(sequenceNumber);
}
}
#endregion
}

Read WCF Message: when this option is checked, incoming messages are extracted from a
SOAP envelope. This feature can be used to receive messages from an application that used
WCF and NetMessagingBinding to send messages to a queue. The following table shows the
code used by consumer tasks to extract a BrokeredMessage object from a WCF message.
As you can see, the WCF message needs to be decoded using a message encoder that
depends on the protocol scheme used by a WCF publisher to encode and send the message.

MessageEncodingBindingElement element;
if (scheme == DefaultScheme)
{
element = new BinaryMessageEncodingBindingElement();
}
else
{
element = new TextMessageEncodingBindingElement();
}
var encoderFactory = element.CreateMessageEncoderFactory();
encoder = encoderFactory.Encoder;
...
var stream = inboundMessage.GetBody<Stream>();
if (stream != null)

326
{
var stringBuilder = new StringBuilder();
var message = encoder.ReadMessage(stream, MaxBufferSize);
using (var reader = message.GetReaderAtBodyContents())
{
// The XmlWriter is used just to indent the XML message
var settings = new XmlWriterSettings { Indent = true };
using (var writer = XmlWriter.Create(stringBuilder, settings))
{
writer.WriteNode(reader, true);
}
}
messageText = stringBuilder.ToString();
}


Complete Receive: when this option is checked and the selected receive mode is equal to
PeekLock, consumer tasks will call the Complete method on received messages to complete
the receive operation and delete the message from the queue. Conversely, when this option
is unchecked, consumer tasks will call the Abandon method on received messages to and
relinquish the lock. This option can be used to test and examine the behavior of the Complete
and Abandon methods of the BrokeredMessage class.

Graph Tab
When the Enable Statistics and Enable Graph options are enabled, this tab displays performance
statistics for sender and receiver tasks and four distinct chart series:
Sender Latency
Receiver Latency
Sender Throughput
Receiver Throughput

View
To vie
corres

The c
topic o
Crea
To cre
segm
1. R
2. C
This o
user t
w a Topic
ew the proper
sponding nod
current version
once created
ate a Top
eate a new to
ent node und
Right-click the
Choose Create
operation disp
to specify the
c
rties of an exi
de in the treev
n of the Servi
. The same p
pic
opic in the cur
der the Topics
node and ch
e Topic comm
plays a custom
following pro
sting topic in
view.
ice Bus does
principle appli
rrent namespa
s subtree and
oose Create
mand under t
m control (see
operties for th
the Main Pan
not allow to c
es to the othe
ace, you can
then perform
Topic from th
the Actions m
e the picture
e new topic:
nel, you can ju
change the pr
er entities (qu
select the To
m one of the fo
he context me
menu.
below) in the
ust select the
roperties of a
ueues, subscr
opics node or
ollowing actio
enu.
Main Panel t

32

e

n existing
riptions, rules
a path
ons:
that allows the
27
).
e

P
ab
D
fo
D
T
du
M
m
E
va
R
R
de
The S
dont
defau
Dele
To de
perfor
1. R
2. C
3. C
Test
To tes
and th
1. R
ath: this field
bsolute path l
Default Messa
or the Default
Duplicate Det
imeSpan valu
uration of the
Max Size In M
megabytes, wh
nable Batch
alue that indic
Requires Dupl
RequiresDuplic
etection featu
SizeInBytes is
explicitly spe
ult value will b
ete a Top
elete an existi
rm one of the
Right-click the
Choose Delete
Click the Delet
t a Topic
st a topic in th
hen perform o
Right-click the
defines the p
like erp/prod/
ge Time To L
MessageTime
tection Histo
ue for the Dup
duplicate de
egabytes: the
hich is the siz
ed Operation
cates whethe
icate Detectio
cateDetection
ure.
s a read-only
cify any value
e used.
pic
ng topic, you
following act
node and ch
e Topic comm
te button in th
he current nam
one of the foll
node and ch
path for the to
topics/orderto
Live: this sect
eToLive prop
ory Time Win
plicateDetecti
tection history
e MaxSizeInM
ze of memory
ns: the Enabl
r server-side
on: this check
n property tha
property and
e for one or m
can select th
tions:
oose Delete T
mand under th
he Main Pane
mespace, you
owing actions
oose Test To
opic within the
opic.
ion gives the
perty that spec
dow: this sec
ionHistoryTim
y.
Megabytes pro
allocated for
eBatchedOpe
batched oper
kbox specifies
at enables or
as such cann
more fields, wh
he correspond
Topic from th
he Actions me
el.
u can select t
s:
opic from the c
e namespace
possibility to
cifies the defa
ction gives the
meWindow pro
operty defines
the topic.
erations prop
rations are en
s a Boolean v
disables the d
not be set wh
hen you click
ding node in t
e context me
enu.
the correspon
context menu
. The value c
enter a Time
ault message
e possibility to
operty that de
s the maximu
erty accepts a
nabled.
value for the
duplicate mes
en creating a
the Create b
he treeview a
nu.
nding node in
u.

32

can be an
eSpan value
time to live.
o enter a
efines the
um size in
a Boolean
ssage
a topic. If you
utton, the
and then
the treeview
28
2. C
This o
M
S
R
G
The m
be us
tab as

View
To vie
Subsc

Choose Test T
operation sho
Message Tab
ender Tab
Receiver Tab
Graph Tab
module that al
ed to test a q
s shown in the
w a Subs
ew the proper
criptions node
Topic comman
ows a custom
llows to test a
queue. The on
e following pi
cription
rties of a subs
e under the to
nd under the
control in the
a topic is virtu
nly addition is
cture.
scription asso
opic node and
Actions menu
e Main panel t
ually identical
the ability to
ociated to a gi
d select the co
u.
that is compo
to that we de
choose a Su
iven topic, yo
orresponding
osed of four ta
escribed abov
bscription in t
ou can expand
subscription

32
abs:
ve and that ca
the Receiver

d the
node.

29
an
The c
subsc
Crea
To ad
then p
R
E
S
C
This o
user t

N
M
m
D
fo
L
pr
re
E
va
E
fo
le
R
R
de
current version
cription once c
ate a Sub
dd a new subs
perform one o
Right-click the
xpand the top
ubscription fr
Choose Create
operation disp
to specify the
Name: this fiel
Max Delivery
message is au
Default Messa
or the Default
ock Duration
roperty that d
eceive mode.
nable Batch
alue that indic
nable Dead
or the EnableD
ettering functio
Requires Dup
RequiresDuplic
etection featu
n of the Servi
created. The
bscription
scription to an
of the followin
topic node an
pic node, righ
rom the conte
e Subscription
plays a custom
following pro
d defines the
Count: the M
utomatically de
age Time To
MessageTime
n: this section
defines the du
ed Operation
cates whethe
Lettering on
DeadLettering
onality on me
plicate Detec
cateDetection
ure.
ice Bus does
same princip
n
n existing topi
ng actions:
nd choose Cr
ht-click the Su
ext menu.
n command u
m control (see
operties for th
e name for the
MaxDeliveryCo
eadlettered a
Live: this sec
eToLive prop
n gives the po
uration of the
ns: the Enabl
r server-side
Message Ex
gOnMessageExp
essage expira
tion: this che
n property tha
not allow to c
ple applies to t
ic, you can se
reate Subscri
ubscriptions n
under the Acti
e the picture
e new subscr
e subscription
ount property
after this numb
ction gives th
perty that spec
ossibility to en
lock used by
eBatchedOpe
batched oper
xpiration: this
piration prop
ation.
eckbox specifi
at enables or
change the pr
the other enti
elect the topic
ption from the
ode under the
ions menu.
below) in the
ription:
n being create
y defines the m
ber of deliveri
e possibility t
cifies the defa
nter a TimeSp
a consumer w
erations prop
rations are en
s checkbox sp
perty that ena
ies a Boolean
disables the d
roperties of a
ties (queues,
c node in the
e context men
e latter and c
Main panel th
ed.
maximum del
ies.
to enter a Tim
ault message
pan value the
when using th
erty accepts a
nabled.
pecifies a Boo
bles or disabl
n value for the
duplicate mes

33
n existing
topics, rules
treeview and
nu.
hoose Create
hat allows the

livery count. A
meSpan value
time to live.
LockDuration
he PeekLock
a Boolean
olean value
les the dead-
e
ssage
30
).
e
e
A
e
n
R
pr
F
A
Dele
To rem
then p
1. R
2. C
3. C
Test
To tes
one o
1. R
2. C
This o
R
G
Pleas
Recei

View
To vie
node
pictur
Requires Ses
roperty that e
ilter: this field
Action: this fie
ete a Sub
move an exis
perform one o
Right-click the
Choose Delete
Click the Delet
t a Subsc
st a subscript
of the following
Right-click the
Choose Test S
operation sho
Receiver Tab
Graph Tab
se refer to the
iver and Grap
w a Rule
ew the proper
under the sub
re below.
sion: this che
enables or dis
d defines the
eld defines the
bscription
sting subscript
of the followin
node and ch
e Subscriptio
te button in th
cription
tion, you can
g actions:
node and ch
Subscription
ows a custom
Test a Queu
ph tabs.
rties of a rule
bscription nod
eckbox specif
sables session
SQL filter exp
e SQL filter a
n
tion, you can
ng actions:
oose Delete
on command
he Main pane
select the cor
oose Test Su
n command un
control in the
ue section for
associated to
de and select
fies a Boolean
ns support on
pression for th
ction for the d
select the co
Subscription
under the Ac
el.
rresponding n
ubscription f
nder the Actio
e Main panel t
a description
o a given sub
t the correspo
n value for the
n the subscrip
he default rul
default rule. T
orresponding
n from the co
ctions menu.
node in the tre
from the conte
ons menu.
that is compo
of the functio
scription, you
onding rule no
e RequiresSe
ption being cre
e.
This field is op
node in the tr
ntext menu.
eeview and th
ext menu.
osed of two ta
onality and op
u can expand
ode, as illustra

33
ession
eated.
ptional.
reeview and
hen perform
abs:
ptions of the

the Rules
ated in the
31

Add
To ad
treevi
1. R
2. E
R
3. C
This o
user t
N
Is
de
ru
F
A
d a Rule
dd a new rule
ew and then
Right-click the
xpand the su
Rule from the
Choose Add R
operation disp
to specify the
Name: this fiel
s Default Rul
efault rule for
ule, this field i
ilter: this field
Action: this fie
to an existing
perform one
subscription
bscription no
context menu
Rule comman
plays a custom
following pro
d defines the
e: this checkb
r the current s
s disabled.
d defines the
eld defines the
g subscription
of the followin
node and cho
de, right-click
u.
nd under the A
m control (see
operties for th
e name for the
box can be us
subscription. I
SQL filter exp
e SQL filter a
n, you can sel
ng actions:
oose Add Ru
k the Rules no
Actions menu
e the picture
e new rule:
e new rule.
sed to indicat
If the subscrip
pression for th
ction for the n
lect the subsc
ule from the c
ode under the
u.
below) in the
te that the rule
ption in quest
he new rule.
new rule.
cription node
context menu.
e latter and ch
Main panel th
e being creat
tion already o

33

in the

hoose Add
hat allows the
ed is the
wns a default
32
e
t

Rem
To rem
treevi
1. R
2. C
3. C
Test
To tes
the tre
move a Ru
move an exis
ew and then
Right-click the
Choose Remo
Click the Remo
t a Relay
st one of endp
eeview as hig
ule
sting rule from
perform one
node and ch
ove Rule comm
ove button in
Service
points expose
ghlighted in re
m a subscriptio
of the followin
oose Remov
mand under t
the Main Pa
Endpoint
ed by a relay
ed in the follow
on, you can s
ng actions:
ve Rule from t
the Edit menu
nel.
t
service, you
wing picture.
select the corr
the context m
u.
can select the
responding no
menu.
e correspond

33

ode in the
ing node in
33

334


This operation shows a custom control in the Main panel that is composed of three tabs:
Message Tab
Sender Tab
Graph Tab
Mess

The M
M
re
fil
M
th
Send

The S
T
tr
th
M
M
sage Tab
Message tab a
Message Text
ead the mess
le.
Message Prop
he picture abo
der Tab
Sender tab all
ask Count: th
ansmit messa
hread.
Message Cou
Method Actio
allows the use
t: specifies th
age from the
perties: you c
ove.
ows the user
his field can b
ages to the re
unt: this intege
n: specified th
er to specify t
e payload of
file system b
can use the d
to specify the
be used to de
elay service. E
er field can be
he value for t
the following
a WCF mess
y pressing the
datagrid to def
e following inf
efine the numb
Every task ru
e used to spe
he WS-Addre
information:
sage. The too
e Open butto
fine the custo
formation:
ber of sender
ns on a sepa
ecify the numb
essing Action
l provides the
n and selectin
om headers, a
r tasks that wi
rate .NET thr
ber of messag
header.

33

e ability to
ng an XML
as shown in

ill be used to
ead pool
ges to send.
35
E
pe
E
of
E
G
E
th
B
co
th
Grap
When
statist
S
S

Con
The S
and a
In the
availa
impro
custo

nable Loggin
erformed in th
nable Verbo
f request and
nable Statist
Graph tab.
nable Graph
hroughput cha
Binding: you c
ommunicate w
he security se
ph Tab
n the Enable S
tics for sende
ender Latenc
ender Throug
nclusions
Service Bus E
a useful comp
e future I will c
able by the Se
ove the tool an
mize my code
ng: when this
he log view.
se: when this
response me
tics: when th
h: when this o
arts in the Gra
can use this s
with the unde
ection to matc
Statistics and
er tasks and tw
cy
ghput
s
Explorer is the
anion for all d
continue to de
ervice Bus tea
nd I strongly e
e to accommo
s option is che
s option is che
essages (if an
is option is ch
option is check
aph tab.
section to sele
erlying relay s
ch settings spe
Enable Grap
wo distinct ch
e perfect tool t
developers an
evelop and ex
am. In the me
encourage yo
odate your ne

ecked, sende
ecked, sende
ny).
hecked, send
ked, sender ta
ect and config
ervice. In par
ecified in the
ph options are
hart series:
to explore the
nd system ad
xtend the tool
eantime, I invi
ou to downloa
eeds.
er tasks will tra
er tasks will lo
er tasks will u
asks will upda
gure the bindi
rticular, make
configuration
e enabled, thi
e features pro
ministrators t
with new fea
ite you to sha
ad from the M
ack the opera
g the payload
update statist
ate the latenc
ing used by th
sure to prope
n file of the rel
s tab displays
ovided by the
hat work with
atures as they
are your ideas
SDN Code G

33
ations
d and headers
ics in the
cy and
he tool to
erly configure
lay service.
s performanc

Service Bus
h it every day.
y will be made
s on how to
Gallery and
36
s
e
e

e

337
Testing


338
Using Visual Studio Load Tests in Windows Azure
Roles
Load testing is the process of measuring the performance of an application under a simulated
usage scenario. For example, you want to load test a website by requesting it to return pages to
1000 users in a five-minute period. By running the load tests in Windows Azure roles, you can
reduce your reliance on hardware through the lifecycle of an application.
Authors: Paolo Salvatori, Sidney Higa.
Jaime Alva Bravo also cowrote Running Load Tests in Mixed Environments.
Acknowledgements: Stphane Crozatier, Benjamin Guinebertire in Load testing from Windows
Azure | Tests de charge depuis Windows Azure. Also thanks to Janice Choi for technical review.
Visual Studio Ultimate enables you to create load tests that are easily managed by a person or a
team of testers. The basic method is to create one "controller" and one or more "agents." You
have one controller, but as many agents as you require. Each agent generates a part of the load.
The controller directs the lifetime of the agent or agents, and records the results of the test.
However running a load test typically requires at least two computers: the first runs the controller,
and the second runs the agent or agents.
With Windows Azure, you can create worker roles to take the place of multiple computers.
Virtualization of computing resources eliminates the need for dedicated hardware for load testing.
This series of topics explains the basic mechanics of setting up Windows Azure to run a load test
controller and agents in two worker roles.
In This Section
Visual Studio Load Test in Windows Azure Overview
Describes the benefits of using Visual Studio Load Test and outlines the required steps.

Windows Azure Load Test Prerequisites and Setup
Lists the requirements for the solution.

Provisioning Windows Azure For a Load Test
Detailed instructions on how to set up the load test application before publishing.

Publishing the Load Test To Windows Azure
Describes the steps for publishing a Load Test to Azure.

Running Load Tests In Mixed Environments
A mixed environment is one in which the components of a load test (test controller,
agents, results repository, and tested system) reside in different environments, such as
on-premises and in Windows Azure. This document explains how you can proceed to

339
configure such a scenario.


Visual Studio Load Test in Windows Azure Overview
The major components of the load test are described here. For a general explanation of the
architecture of the application, see Load Testing with Agents running on Windows Azure part 1.
For an overview of load tests in Visual Studio, see Understanding Load Tests.
This document is about a mostly "pure" solution. Except for a copy of Visual Studio, the
components of the load test are run as Windows Azure worker roles. There is an
alternative to this "mostly pure" scenario: if you run or host any of the components in a
mixed environment. For example, you may want to run the test controller on-premises
with agents that run on a worker role. Or you want to use SQL Azure to store your load
test data. In other words, you can distribute the components of the load test across
Windows Azure roles and SQL Azure and on-premises. For more documentation on
selecting and setting up these alternative setups, see this topic: Running Load Tests In
Mixed Environments.
Benefits
Before continuing with the overview, here are some of the advantages of using Visual Studio load
tests in Windows Azure.
Entry Cost
The cost of doing load tests decreases greatly after the initial investment. Your first
costs are for Visual Studio Ultimate and for the license to use the Load Test feature.
After that, you create the load test harnessa Visual Studio project. You must have a
Windows Azure subscription and capacity to deploy and run web and worker roles on
an as-needed basis. These costs are balanced against the costs of owing hardware and
all of the collateral costs. (Those costs include software, electrical power, a data center,
and people to maintain the whole system.)

Maintenance Cost
Using the procedures in here, you can create an easily maintained harness for load
testing. After creation, updating the project will be minimal in the foreseeable future.

Elasticity
The load test can be easily modified to accommodate different scenarios and
conditions. For example, you can configure the Azure hosted server with a larger
number of worker roles. Each role becomes a test agent that can be used to increase
the load.
Note

340

Repeatability
After the initial investment however, you can repurpose the load test against various
test subjects with minimal time and costs. Simply reconfigure the test project, and
redeploy it on Windows Azure. Run the test for the necessary time and undeploy it.
Repeat as needed.

Real-life fidelity
As you host your services in a large data center, the concurrency of the system is
greatly increased (more about concurrency later.) Hosting your final application on
Windows Azure presents perfect fidelity to the final deployment.

Concurrency
Concurrency is defined as a property of system where several tasks are executed
simultaneously, and are possibly interacting. A factor that limits concurrency is the number of
available IP addresses. The more IP addresses the system leverages, the greater the concurrent
processing. Typically, the number of available addresses depends on the size of your IP provider.
If your service level agreement is substantial, it is typically allocated a large number of IP
addresses. But such agreements are not common. However, when you use Windows Azure as
your platform, you have the benefit of using a Microsoft data center and its resources. That
includes a large pool of IP addresses. Hosted services in Windows Azure are assigned virtual IP
addresses. In this discussion, the outward facing (Internet) load balancer (not the hosted
services) use the IP addresses. And having a large number is an advantage of the Microsoft data
center. Also note that not all systems require this level of concurrency. This is an advantage only
when testing a high-concurrency system.
This increased capacity for concurrency is another great benefit of running the load tests on
Window Azure. This level of concurrency is also hardest to reproduce outside of a large data
center.
Overview
This graphic illustrates the components of a load test. A copy of Visual Studio Ultimate is used to
configure the load test as well as start and control it. Visual Studio also provides graphic views of
the test results. The load test pieces include the test controller, the test agents, and the system
under test. The test controller and the agents are running in a Windows Azure data center as
worker roles. Except for the system under test, all of the load test components must belong to the
same Windows Azure Connect group. That enables the pieces to communicate freely with each
other, and with the Visual Studio instance. To create counters sets that track performance
counters on the machines that host the system under test, join those machines to the Windows
Azure Connect group.

The lo
this do
Prere
Azure
The M
test. S
Azure
Crea
wor
ena
can
you
Dec

Storag
You
the
softw
the
Cre

Conne
To e
The
cond
cont
copy
oad test uses
ocument. For
quisites and S
e Manageme
Management
Specifically, y
e Hosted Servi
ate a hosted s
ker roles. The f
bles remote ex
use the instan
need. For mor
creasing the N
ge account
u need a Windo
worker roles. T
ware for SQL E
agent code. Fo
eate a Storage
ect group
enable commu
e Connect featu
duct the agent
troller. The Con
y of Visual Stu
a few major
r the complete
Setup
ent Portal
Portal is used
you use the fo
ice
service to run t
first role hosts
xecution of the
nce feature of A
re information,
Number of Ro
ows Azure Stor
The storage blo
Express. The s
or more informa
e Account for
nication betwe
ure functions as
activity. It also
nnect group als
dio. For more i
technologies
e list of down
d to create an
ollowing Azure
the load test ap
a single instan
load test. The
Azure to create
see How to S
ole Instances
rage account to
obs contain thre
second contains
ation about cre
r a Windows A
en the controlle
s a virtual priva
o allows the age
so allows you t
information, se
. Their functio
loads require
nd manage th
e features:
pplication. The
nce of the test c
second role ho
as many insta
Scale Applicat
o hold the comp
ee .zip files. On
s the controller
eating a storage
Azure Subscr
er and agents,
ate network tha
ents to return t
to run the contr
ee Overview o
ons in the load
d, see Windo
e primary stru
application co
controller. The
osts a single ag
ances of the ag
tions by Incre
pressed code t
ne file contains
r code. The thir
e account, see
ription
create a Conn
at allows the co
est results to th
roller from an o
of Windows Az
d test are des
ows Azure Loa
uctures used
nsists of two
controller
gent. You
ent role as
easing or
that runs on
s the set-up
rd contains
How to
nect group.
ontroller to
he
on-premises
zure

34

scribed later i
ad Test
for the load
41
n

342
Connect.

In addition, upload two X.509 certificates to the Azure portal. The certificates are required to
authenticate and encrypt the data streams between your on-premises computer and the Azure
portal. The first certificate is used to enable publishing an application from Visual Studio directly
into Azure. That certificate is called a management certificate in Azure parlance. The second
certificate is used to enable a remote desktop connection between an on-premises computer and
an Azure worker role. That certificate is called a service certificate in Azure terms.
Visual Studio Ultimate
Visual Studio is the development environment, and it is the only version that provides the load
test functionality. For a comparison of Visual Studio versions, see Visual Studio 2010
Comparison. Using Visual Studio Ultimate, you can create test scripts to load test your
application.
Visual Studio also has the facility to manage the controller remotely. In a non-Azure scenario, the
remote controller would be running on a second computer. In the Azure scenario, the controller
runs in a worker role in the cloud.
SQL Server 2012 Express Edition
Collecting and storing the data from the agents requires a database. Fortunately, SQL Server
2012 Express is a free edition, and the load test program uses it by default. All that is required is
that the database engine is also deployed onto the worker role that is also running the controller
software. Setup is accomplished using code in the RoleEntryPoint.OnStart method.
The SQL data files cannot reside in the virtual hard drive of the worker role because the data is
lost when the role is recycled. The recommendation is to mount a shared drive for all roles to
write to. Details are found in Using SMB to Share a Windows Azure Drive among multiple Roles.
Another alternative is to use SQL Azure. Download the loadtest2010.dacpac file to provision SQL
Azure with the database. For more information, see Running Load Tests In Mixed Environments.
Assemble the pieces you need
To begin, download the software required to provision the Azure portal and to configure the
application. You can find details about the downloads in this topic: Windows Azure Load Test
Prerequisites and Setup.
Provision Windows Azure and configure the application
Once you have the downloaded all extra software, you can provision the Windows Azure portal
and configure the application as outlined later. Details for completing each step in detail are found
in the Provisioning Windows Azure For a Load Test.
1. Provision the Windows Azure portal with a new hosted service, a storage account, and
Connect virtual network.
2. After creating the storage and hosted-service accounts, use the values from the Windows
Azure portal to configure the application.
3. Create three .zip files that contain software that automatically installs on the worker roles.

343
The application requires both the controller and agent software to run on worker roles. The
software for creating the controller and agents are precreated, and available for you to
download. In order to get the software onto the worker roles involves one technique: First
create two .zip files that contain the test controller and agent software. Use a tool of your
choice (there are free versions) to create the .zip files. Once the .zip files are created, upload
the files to your Windows Azure storage account. When the application runs, the .zip files are
retrieved from the storage, unpacked and loaded onto the worker roles.
There is a third .zip file you must create: one that contains the setup package for SQL Server
2008 SQL Express. As with the controller and agent packages, the SQL Server setup
automatically runs and deploys an instance of SQL Express onto the worker role.
4. Upload the .zip files to the Azure storage account. For this task, use the Storage Service
Client, a free project that you can download.
5. Create a Windows Azure Connect group in the Windows Azure portal.
6. Install a Connect endpoint onto the computer that is used to connect to the controller role.
7. Configure the application with the Connect activation token. The token allows the worker
roles to join the Connect group.
Publish the application
Once both the portal and the application are correctly provisioned and configured, you can
publish the application to Windows Azure. Detailed steps are found in Publishing the Load Test
To Windows Azure.
1. Retrieve the Windows Azure subscription ID. The ID is used to enable Visual Studio to
upload the application to your account.
2. Create and upload a management certificate. The procedure is as follows.
a. Use the Publish Windows Azure dialog box and create a new management certificate.
The certificate enables Windows Azure to authorize when you upload the application.
b. Paste the subscription ID into the dialog box. The ID enables Visual Studio to identify
your account.
3. Export the management certificate. To upload the certificate, export it from the computer's
local certificate cache.
4. Upload the management certificate. Use the Management Portal to upload the certificate.
5. Publish the application.
Use Remote Desktop
Once the application is published, you use Remote Desktop to connect to it. Visual Studio
supplies the tools and features to enable this functionality. The steps for enabling a remote
desktop connection are found in Publishing the Load Test To Windows Azure
Next Steps
To begin, download the tools and components of the solution: Windows Azure Load Test
Prerequisites and Setup

344
Windows Azure Load Test Prerequisites and Setup
This is one of a series of topics that cover load testing using Visual Studio Ultimate and Windows
Azure. This topic catalogs the prerequisites and samples to download.
Prerequisites
The following are required:
Visual Studio Ultimate
The load testing feature uses Visual Studio to control a remote instance of a controller.
The controller is running as a Windows Azure worker role. This capability is only
available with Visual Studio Ultimate.

Windows Azure SDK
Please search on the Internet for the latest version of the SDK.

Windows Azure Account
A Windows Azure account is required to complete the examples. The features of
Windows Azure that will be used include:
Windows Azure Connect
Windows Azure Service Bus
Windows Azure Storage
MSDN Subscription


Downloads
The following can be downloaded from the Internet:
1. Load testing from Windows Azure. Save the file to your computer and unzip it. The sample
contains the code running the load test.
2. SQL Server 2008 RS Express and Management ToolsThe load test must save the data to a
SQL Express database running on the controller.
3. Load Test Key from MSDN. You must sign in to obtain a key, which the load test feature
requires to run. The MSDN subscription is included with Visual Studio Ultimate.
4. Visual Studio Test Agents 2010 - ISO. The test agents and the controller code are included in
this download.
5. Storage Service Smart Client. Use this tool to upload files to Azure storage.
6. Service Bus Explorer. This tool creates Service Bus relays and message queues or topics.
In addition, you will need a tool for unpacking and creating .zip files, several of which are
available for free on the Internet.

345
Create the Azure Hosted Service
The first step is to create a hosted service in the Azure portal. But do not deploy it yet.

1. Sign into the Windows Azure portal.
2. In the left pane, click Hosted Services, Storage Accounts & CDN.
3. If you manage more than one subscription, select the subscription to use.
4. In the Windows Azure Platform ribbon, click New Hosted Service.
5. In the Create a New Hosted Service dialog, fill in appropriate values for:
a. Name of the service
b. URL for the service. The URL must be unique.
c. Region or affinity group.
d. Deployment name. For this exercise, type Version 1
6. Uncheck the box for Start after successful deployment.
7. Under Deployment options, select Do not deploy.
For more information about creating hosted services, see How to Create a Hosted Service
Create the Azure storage account
Create an Azure storage account. This will be used to create blobs that contain the zip files for
the agents and controller.
Follow the instructions found in this topic: How to Create a Storage Account for a Windows Azure
Subscription.
Set up Windows Azure Connect
The Azure Connect feature allows you to create a virtual private network (VPN) of computers.
You can add any on-premise computer to the VPN by installing an endpoint onto the computer.
You can also add Azure worker roles to the VPN through configuration. Windows Azure Connect
is used to establish an IPSec, IPv6 VPN between one or multiple roles and one or multiple local
machines. In this scenario, this enables the connection between the controller role and the local
Visual Studio machine. Also, if the tested application needs to access local resources (such as a
database, file system, LOB application, etc.), be sure to enlist the agent role in the same
Windows Azure Connect group as the local machine hosting the resource.
Follow the instructions found in this topic: Tutorial: Setting up Windows Azure Connect.
Next Steps
Download the application (Load testing from Windows Azure) and begin with Provisioning
Windows Azure For a Load Test.

346
Provisioning Windows Azure For a Load Test
When using Visual Studio Ultimate to do a load test in Azure, you must provision the Azure with
the following components:
1. Hosted Service
2. Storage Account
3. Connect Group
In addition, you must create three .zip files that are uploaded to the Azure portal. The .zip files
contain the controller and agent code, as well as the setup files for SQL Server 2008 R2 Express.
The sections below walk through the steps of provisioning the Azure Management portal.
For an overview about running visual Studio load test in Azure, see Using Visual Studio Load
Tests in Windows Azure Roles
Creating a Hosted Service
For this application, you must create a hosted service, however you should not deploy it.
1. Log in to the Windows Azure Management portal.
2. In the left pane, click Hosted Services, Storage Accounts & CDN.
3. In the left pane, click the Hosted Services node.
4. In the ribbon, click New Hosted Service.
5. Select the subscription where the service will be created.
6. Type the name of the service. For this tutorial, type "Azure LoadTest"
7. Type a URL prefix for the service. The URL must be unique; if is not, this message will
appear:
8. Choose a region or affinity group for the service.
9. Under Deployment options select the Do not deploy option.
Creating a Storage Account
A storage account for containing public blobs must be created.
1. In the left pane, click Storage Accounts.
2. Right-click the subscription name that hosts the service, then click New Storage
Account.
3. In the dialog box, type a unique URL. Note that the URL must be in all lower-case letters,
and can contain only letters and numbers.
4. Choose a region or affinity group for the storage. It should be the same region or affinity
group as the hosted service.
5. Click OK.
6. After the account is created, refresh the browser to see the Properties of the account.
To create a hosted service
To create the storage account

347
7. From the Properties pane, copy the Name value, and save it for later use. The value is
also referred to as the storage account name.
8. Under Primary access key click the View button.
9. In the dialog box, click the "copy" icon (to the right of the access key value). Save the
value for later use. The value is also called the account key.
Configuring the Application with the Storage Account Name and Key
With a storage account created, you can configure the application.
1. Run Visual Studio as an Administrator.
2. Open the AzureLoadTest solution. (See Windows Azure Load Test Prerequisites and
Setup.)
3. In Solution Explorer, expand the AzureLoadTest project, then expand the Roles folder.
4. Right-click the Agent role and click Properties.
5. In the Agent [Role] page, click the Settings tab.
6. Select the Microsoft.WindowsAzure.Plugins.Diagnostics.ConnectionString setting.
7. In the Type column, set the drop-down value to Connection String.
8. At the right of the Value column, click the ellipsis () button.
9. In the Storage Account Connection String dialog box select the Enter storage
account credentials option.
10. Paste in the account name into the Account name box.
11. Paste the primary access key value into the Account key box.
12. Click the OK button.
13. In Solution Explorer, in the Roles folder, right-click the Controller role and click
Properties.
14. In the Controller [Role] page, click Settings.
15. Repeat steps 6 to 12.
Configuring the Application with the Load Test Key
The Visual Studio load test feature requires that you supply a valid key in order to work. The key
can be obtained from your MSDN subscription, which is free with a copy of Visual Studio
Ultimate. Once you have the key, configure the application with it. For more information about the
load test key key, see:
Managing Your Virtual User Licenses for Load Testing with a Test Controller
Microsoft Visual Studio 2010 Load Test Feature Pack
Visual Studio 2010 Feature Packs Frequently Asked Questions
1. Sign in to your MSDN subscription. Go to: MSDN Subscriptions and click the Sign In link.
To configure the Load Test application
To obtain the load test key

348
2. Once signed in, go to the Secure Downloads page.
3. Find the Visual Studio 2010 Load Test Feature Pack entry. In the Keys column, click the
View link.
4. Find the value in the Product Keys column.
1. In Solution Explorer, open the Properties page of the Controller role.
2. In the Properties page, click the Settings tab.
3. Paste the product key into the TestController_License setting.
Creating Controller, Agent, and SQL Express .Zip Files
The load test relies on two sets of code that run on the two worker roles. The controller and agent
code are downloaded from the Microsoft Download Center and must be repackaged as .zip files.
The .zip files must then be uploaded to the Azure storage account. When the application is
deployed, the OnStart method of the RoleEntryPoint class is called. Within the method,
installation code is then invoked, and the controller and agents are run as services on the worker
roles.
1. Download the X16-49583VS2010Agents1.iso file from the Download Center: Visual
Studio Agents 2010 ISO.
The file contains both the controller and agent software.
2. Find and download .zip utility on the Internet by searching for "free .zip utilities" or similar.
3. Run the utility and open the X16-49583VS2010Agents1.iso file.
4. Extract the testagent and testcontroller folders.
5. Using the utility, create two new .zip files with the contents of the extracted folders. The
new .zip files are named as follows:
testagent.zip
testcontroller.zip
Important
The contents of the folders must be at the top level of the .zip file. By default,
simply repackaging the folders will put the contents in a folder in the .zip file,
which is one level too deep. The contents must be accessible at the first level for
the installation code to extract the code.
A third .zip file must be created. The SQL Express engine must also be installed on the
controller worker role. In order to do this, you must first retrieve the (free) SQL Server
Express installation package, the repackage it as a .zip file.
1. Download the the SQL Server 2008 R2 Express database and management setup:SQL
Server 2008 RS Express and Management Tools
To configure the application with the load test key
To create the controller agent .zip files
To create the SQL Server Express installation file

349
2. Run the .zip utility and open the SQLEXPRWT_x64_ENU.exe. (Despite the .exe
extension, the file is also an .iso archive file.
3. Extract the contents.
4. Use the .zip utility and create a new file named SQLEXPRWT_x64_ENU.zip.
Important
As with the controller and agent .zip files, the contents of the installation package
must be at the top level of the new .zip file.
Uploading the .Zip Files to the Azure Storage Account.
The .zip files must be available to the worker role as it starts up and runs the installation
procedures. To do that, you must create a single public blob container, and upload the files there.
While you can use code to upload the files, this procedure uses a downloadable tool.
The storage container must be public.
1. Download the tool here: Storage Service Smart Client
2. Open the StorageServicesSmartClient solution in Visual Studio.
3. Configure the application with your storage account name and key.
a. In Solution Explorer, open the App.config file.
b. Find the <storageAccounts> element.
c. Under the <storageAccounts> element, select one of the <add> elements. Refer to
the following example. In the key attribute, replace "<MyStorageName>" with your
Azure storage account name. In the value attribute, replace "<MyStorageName>"
with your storage name. Also in the value attribute replace "<MyPrimaryAccessKey>"
with your storage account key.
<add key="<MyStorageName>"
value="DefaultEndpointsProtocol=https;AccountName=<MyStorageNa
me>;AccountKey=<MyPrimaryAccessKey>"/>

4. Run the application.
5. In the lower left corner, under Storage Accounts, click the drop-down box and select
your storage name.
6. Under Container Commands click the New button. Leave the default set to Public.
7. Give the new container a name and click OK.
8. In the left pane, under Blob Services, select the new container.
9. In the section named Blob Commands click the New button.
10. In the Upload Files dialog box, click Select.
11. Navigate to the directory where the controller, agent, and SQL Express setup .zip files
are saved to upload them.
Important
To upload the .zip files

350
12. Record the name of the new container for later reference. It is used when configuring the
application.
Configuring the Application for the .Zip Files
Having created the .zip files and uploaded them to the Azure storage account, you must now
configure the Load Test application with the proper values.
1. In Visual Studio, open theAzureLoadTest solution.
2. In Solution Explorer, expand the AzureLoadTest project, then expand the Roles folder.
3. Right-click the Agent role and click Properties.
4. In the Agent [Role] page, click the Settings tab.
5. Select the TestAgent_Zip setting.
6. In the Value box, replace the values for the storage address as shown. Refer to the
following example. In the URL, replace "<MyStorageName>" with your Azure storage
account name. Then replace "<mycontainername>" with your blob container name.
http://<MyStorageName>.blob.core.windows.net/<mycontainername
>/testagent.zip
7. In Solution Explorer, right-click the Controller role and click Properties.
8. In the Settings tab, select the SqlExpress_Zip setting.
9. In the Value box, replace the values for the storage address as shown. Refer to the
following example. In the URL, replace "<MyStorageName>" with your Azure storage
account name. Then replace "<mycontainername>" with your blob container name.
http://<MyStorageName>.blob.core.windows.net/<mycontainername
>/SQLEXPRWT_x64_ENU.zip
10. Select the TestController_Zip setting. Replace the
11. In the Value box, replace the values for the storage address as shown. Refer to the
following example. In the URL, replace "<MyStorageName>" with your Azure storage
account name. Then replace "<mycontainername>" with your blob container name.
http://<MyStorageName>.blob.core.windows.net/<mycontainername
>/testcontroller.zip
Creating a Connect Group
The Connect feature of Azure allows you to create a Virtual Private Network. Members of the
group can include on-premise computers as well as Azure role instances. For the Load Test
solution, the Connect group allows communication between the test controller and the agents.
For more details about setting up a Connect group, see http://msdn.microsoft.com/en-
us/library/gg508836.aspx
If you have not created a Connect group before, install a local endpoint onto the development
computer (the computer used to develop and manage the controller instance).
To configure the application for the .zip files

351
The local endpoint can be installed only once on any computer, and it works only with
Connect groups created in a specific Azure subscription. If you have already installed a
local endpoint for a particular Azure subscription, you must uninstall it before installing a
new local endpoint for a different subscription. For more information, see How to Install
Local Endpoints with Windows Azure Connect.
1. In the Azure Management portal, click Virtual Network in the left pane.
2. Select the subscription where the application will be hosted.
3. In the ribbon, click Install Local Endpoint.
4. In the Install Windows Azure Connect Endpoint Software dialog box, click Copy Link
to Clipboard and click OK.
5. Paste the link into the Internet Explorer address box and press Enter.
6. Follow the instructions to install the endpoint.
1. In the Azure Management portal, ensure that you have clicked Virtual Network in the left
pane.
2. Select the subscription in which the application will be hosted.
3. Expand the node where the service is hosted.
4. Select the Groups and Roles node.
5. In the ribbon, click Create Group.
6. In the dialog box, type a name for the group.
7. Click the Create button.
8. In the Windows tray, click the Connect Endpoint tray icon and click Refresh Policy.
9. To add the local endpoint to the group:
a. In the Azure Management portal, select the newly-created group.
b. In the ribbon, click Edit Group.
c. In the Edit Endpoint Group dialog box, in the Connect from section, click the Add
button
d. In the Select Local Endpoints dialog box select the computer and click OK.
If the computer does not appear, see Troubleshooting Windows Azure Connect
e. Click the Save button.
At this point, you cannot add the worker role endpoints to the group. Instead, the roles are
added when you configure the worker roles with the activation token, as shown below in
"Configuring the Application for the Connect Group" below. Additionally, the roles only
appear in the Connect group after the application is deployed, and the roles are created.
Important
To install a local endpoint
To create the Connect group

352
Configuring the Application for the Connect Group
1. In the Windows Azure Management portal, click Virtual Network to open the Connect
user interface.
2. Select the subscription that hosts the Load Test application Connect group.
3. In the ribbon, click Get Activation Token.
4. Click the Copy Token to Clipboard button and click OK.
5. In Visual Studio, open the AzureLoadTest solution.
6. In Solution Explorer, expand the AzureLoadTest project, then expand the Roles folder.
7. Right-click the Agent role and click Properties.
8. In the Agent [Role] page, click the Virtual Network tab.
9. Select the Activate Windows Azure Connect setting.
10. Paste the activation token value into the box.
11. In Solution Explorer, right-click the Agent role, and click Properties.
12. Repeat steps 8 to 10.
Next Steps
The steps here have configured both the Azure Management portal and the Load Test application
with the necessary files and values to get to the next stage of development. To continue, go to
Publishing the Load Test To Windows Azure.
Publishing the Load Test To Windows Azure
Publishing an application to Azure has one requirement: creating and supplying a management
certificate to the Azure Management portal. Once the portal is correctly configured, you can
upload a new version of an application. If there is an existing application already running, you
have the option of stopping and overwriting it.
Publishing an Application from Visual Studio
1. Log in to the Azure Management portal.
2. In the left pane, click Hosted Services, Storage Accounts & CDN.
3. In the upper left pane, click Hosted Services.
4. Select the subscription where you will publish the application.
5. In the Properties pane, find the Subscription ID value. Select it and copy it.
Store the value, or keep the portal open so you can retrieve the value again if needed.
6. Run Visual Studio as an Administrator.
7. Open the Load Test application. (See Windows Azure Load Test Prerequisites and
Setup.)
To configure the application for use in the Connect group
To prepare the application

353
8. In Solution Explorer, right-click the Azure project node, then click Publish. You will see
the Publish Windows Azure Application dialog box.
If you have not previously published an application, you will be at the Sign in stage. If
you have already published an application, you can select an existing target profile.
Otherwise, to create a new profile, click the <Previous button until you return to the Sign
in stage.
9. Click the drop-down box, and from the list select <Manage>.
10. In the Windows Azure Project Management Settings dialog box, click the New button.
11. In the Windows Azure Project management Authentication dialog box, under Create
or select an existing certificate for authentication click the drop-down box.
12. At the bottom of the drop-down list, click <Create>.
Alternatively, select an existing certificate to use.
13. In the Create Certificate dialog box type in a friendly name such as
"LoadTestManagementCertificate," Then click the OK button. Copy and store the friendly
name for later reference. (The name is used in Exporting and Uploading the Management
Certificate below.) Also copy and store the thumbprint. In the Azure Management portal,
you can view the thumbprints of all management certificates, which is a quick way to find
a specific certificate.
14. In the Windows Azure Project management Authentication dialog box, paste the
Subscription ID value into the third box.
15. In the box under Name these credentials type in a friendly name, such as
LoadTestCredentials
16. Click the OK button.
17. In the Windows Azure Project Management Setting dialog box click Close.
18. Before you can publish the application, you must export the certificate and upload it to the
Azure Management portal, as described below. Keep the Publish Windows Azure
Application dialog box while doing so.
Exporting and Uploading the Management Certificate
Before you can upload a certificate, you must export it into a file format. Export the certificate
using these steps:
1. Open the Visual Studio Command Prompt. See Visual Studio and Windows SDK
Command Prompts
2. In the prompt, type mmc and press enter to open the Microsoft Management Console.
3. On the File menu, click Add/Remove Snap In.
4. In the list of available snap-ins, click Certificates. Then click the Add> button.
5. In the Certificates snap-in dialog box, ensure My user account is selected and click
Finish. Then click the OK button.
6. Expand the following Certificates-Current User node. Then expand the Personal node,
and finally expand the Certificates node.
To export the certificate

354
7. In the list of certificates, right-click the certificate that was just created, then click All
Tasks, and click Export.
8. In the Certificate Export Wizard, click Next. Ensure that the No, do not export the
private key option is selected and click Next.
9. In the Export File Format dialog use the default settings and click Next>.
10. Click the Browse button and navigate to a folder where you will save the file. Copy the
path for the next step (uploading the certificate).
11. Type in a file name and click Save.
12. Click the Next button.
13. In the final page click the Finish button.
1. In the Azure Management portal, click Hosted Services, Storage Accounts & CDN.
2. In the left pane, click the Management Certificates folder.
3. Select the subscription where the application will be published.
4. In the ribbon, click Add Certificate.
5. In the Add New Management Certificate dialog box, click the Browse button.
6. Navigate to the folder where the certificate is saved, and select the certificate.
7. Close the dialog box.
Creating a Local User To Manage a Load Test Controller
The Load Test feature of Visual Studio allows you to remotely access the load test controller from
Visual Studio. For example, you can change the settings of test agents using this functionality.
However, to do this, the identity running the Visual Studio application must be authenticated to
the load test controller, which will be running on an Azure worker role. To enable the remote
computer to authenticate the Visual Studio user, first create a new local user on the computer that
will be running Visual Studio.
1. Open a command prompt. At the Start menu, type cmd and press Enter.
2. In the command prompt, type mmc and press Enter to run the Microsoft Management
Control.
3. On the File menu click Add/Remove Snap-in.
4. In the Add or Remove Snap-ins dialog box, double-click Local Users and Groups.
5. In the Choose Target Machine dialog box keep the default, Local computer, and click
Finish.
6. In the Add or Remove Snap-ins dialog box, click OK. The Console Root and Local
Users and Groups nodes appear.
7. Expand the Local Users and Groups node.
8. Right-click the Users folder and click New User.
9. In the New User dialog, clear the User must change password at next logon box.
To upload the certificate to the Azure Management Portal
To create a new local user

355
10. In the User Name, type the new user name.
11. In the Password and Confirm Password boxes type a valid password.
12. Click the Create button, and close the New User dialog box.
13. In the management console, right-click the new user and click Properties. You will now
add the user to the Administrators group of the computer.
14. In the user Properties dialog box, click the Member Of tab.
15. Click the Add button.
16. In the Select Groups dialog box, type Administrators and click the Check Names
button. The name of the computer with a slash and Administrators should appear.
17. Click the OK button.
18. In the user properties dialog box, click OK to close it.
19. Close the management console.
Managing a Load Test Controller From Visual Studio
After the application is published on Windows Azure(see "Publishing the Application" below for
instructions), you will be able to connect to the virtual machine running on the worker role.
1. Log onto the computer as the new local user.
2. Run Visual Studio as an Administrator.
3. In Visual Studio, on the Test menu, click Manage Test Controllers. If you have correctly
configured the application, you will see a list of virtual machines running on Windows
Azure. Note that all worker roles and the computer you are running Visual Studio on must
all be part of the same Azure Connect group. If the connection fails, check the Azure
Connect configuration. Ensure that the local computer is in the same Connect group as
the worker role. Also check the ActivationToken value of the controller and agent service
definitions are identitcal:
4. In the left pane of the Windows Azure Management Portal, click Virtual Network.
5. Click the subscription containing the load test service.
Configuring the Application for Remote Desktop
Before publishing, you must also enable a remote desktop connection to the application. Doing so
allows you to remotely connect to any running instance of the application. Remote desktop
authentication requires a certificate; you must also configure the application to use the remote
desktop feature.
Configuration consists of two procedures. The first procedure is to enable a remote desktop
connection to the application, as shown below in "To configure the application for Remote
Desktop connection."
You must also specify which users can access the remote desktop. When you do so, you must
also specify a password to authenticate the user at log-in. If you are developing the Azure
application on a Windows Server machine, you have the added option of creating a new user
account specifically for using the Load test application.
To manage load test controllers from Visual Studio

356
1. In Visual Studio, open the Load Test application.
2. Right-click the AzureLoadTest project and click Configure Remote Desktop.
3. Select the Enable connections for all roles box.
4. Click the drop-down list and click <Create>.
5. Type a friendly name for the certificate, such as "RemoteDesktopServiceCertificate," and
click OK.
At this point, Visual Studio creates the certificate for you. When you publish the application,
the correct certificate is also uploaded.
Viewing the running instance using Remote Desktop
After the application is published on Windows Azure(see "Publishing the Application" below for
instructions), you can connect to the deployed, running instance.
1. Go to the Azure Management portal.
2. In the left pane, click Hosted Services, Storage Accounts & CDN.
3. Under the hosted service node, select the Controller node.
4. In the Remote Access section of the ribbon, click Configure.
The Set Remote Destkop Credentials dialog box appears. The User name box will be
filled with a default value.
5. In the Password box type the value that will be used when you connect to the controller
instance. Retype the same password in the second box.
6. Click the OK button.
7. Select the instance. By default, it is named Controller_IN_0.
8. In the Remote Access section of the ribbon, click Connect.
9. If prompted to "Open" or "Save," choose Open.
10. In the Remote Desktop Connection dialog, click Connect.
11. Type the password for the user.
12. In the Remote Desktop Connection dialog box click Yes.
The connection using Remote Desktop will be made and you will be able to examine the
worker role.
Publishing the Application
If you have not closed the Publish Windows Application dialog box, you can click Publish.
If you have closed the dialog box, but have uploaded the management certificate to the Windows
Azure Management portal, do the following:
In Solution Explorer, right-click the AzureLoadTest project and click Publish. Then click the
Publish button.
To configure the application for Remote Desktop connection
To view the Controller instance using Remote Desktop

357
Running Load Tests In Mixed Environments
This topic covers running Visual Studio Load tests that run in a mixed environment. A "mixed
environment" is one where the components of the load test are hosted in different contexts, such
as on-premises, or as Windows Azure worker roles. For example, a mixed environment would be
one where the test agents run on-premises and the data repository is sited on a SQL Azure
server. This is in contrast to the system described in Visual Studio Load Test in Windows Azure
Overview. In that topic, all of the components, except for the work station are run as Windows
Azure worker roles.
Running the load test in these mixed environments may be necessary due to regulatory reasons,
or for reasons specific to your application. At any rate, this topic covers the choices you have, and
how to set up such varied scenarios.
Download
The files required for these procedures are found here: VSLoadTestInMixedEnvironments
Concurrency
One factor used in weighing different configurations is the increased concurrency from running
many processes in a large data center. Concurrency is defined as a property of system where
several tasks execute simultaneously, and are possibly interacting. A factor that limits
concurrency the number of available IP addresses. The more IP addresses the system leverages,
the greater the concurrent processing. Typically, the number of addresses available depends on
the size of your IP provider. If your service level agreement is substantial, it is typically allocated a
large number of IP addresses. But such agreements are not common. However, when you use
Windows Azure as your platform, you have the benefit of using a Microsoft data center and its
resources. That includes a large pool of IP addresses. Hosted services in Windows Azure are
assigned virtual IP addresses. In this discussion, the IP addresses are used by the outward
facing (Internet) load balancer (not the hosted services). And having a large number is an
advantage of the Microsoft data center. Also note that not all systems require this level of
concurrency.
This increased capacity for concurrency is another great benefit of running the load tests on
Window Azure. This level of concurrency is also hardest to reproduce outside of a large data
center.
Factors
There are six factors that affect concurrency. The first two factors are the contexts where you run
the load test: cloud and on-premises.
On-premises
Cloud
The other four factors are the components of the load test. They are:
Test Controller
Test Agents
Results Repository

358
Tested System
The four components are shown here:


In the graphic, the relative widths of the lines represent the amount of data transferred; the thicker
the line, the greater the amount of data. The heaviest data transfer occurs between the test
controller and the results repository. The lightest load occurs between the tested system and the
controller. (However the actual load depends on how much logging the tested system produces.)
For reference, see the graphic in Visual Studio Load Test in Windows Azure Overview.
For the sake of clarity, the graphic assumes that the workstation that hosts Visual Studio
also hosts the Test Controller. This configuration facilitates the communication between
Visual Studio and the test controller. During a load test, the controller streams a large
amount of data back to Visual Studio for real-time performance monitoring.
Topologies
Given the four components, and the two contexts, various topological choices become evident.
The four components of the load test can exist in one or the other context. For example, the two
simplest topologies are stated here.
1. All components on cloud.
2. All components on-premises.
For the sake of clarity, the two simplest choices are shown in this table.

Note


359
Controller Agents Repository Tested System Notes
On-premises On-premises On-premises On-premises Limited
concurrency no
network traffic
outside premises.
Cloud Cloud Cloud Cloud Large concurrency
(More IP
addresses) and no
network traffic
outside the cloud.

Now the topologies are more complicated. To keep things simple, this table shows one major
division. In this table, the controller runs on-premises.
The traffic from agents to the tested system is not accounted for, as it is assumed to be part of
any test cost. Also note that the network traffic in the following tables has a monetary cost. You
are charged for data transferred out of the data center. Internal traffic is not billed. For pricing
details, see Pricing Details, and search for "data transfers measured in GB".
This next table shows a major splitting point: when the test controller runs on-premises. In that
case, the traffic between the components must cross the boundary, with various degrees of
consequences, depending on the component and its level of "chattiness."
Controller Runs On-Premises
Controller Agents Repository Tested System Notes
On-premises On-premises On-premises Cloud Limited
concurrency,
network traffic from
tested system to
controller.
On-premises On-premises Cloud On-premises Limited
concurrency,
network traffic from
controller to
repository.
On-premises On-premises Cloud Cloud Limited
concurrency,
network traffic from
tested system to
controller and back
to the repository.
On-premises Cloud On-premises On-premises Larger
concurrency,

360
Controller Agents Repository Tested System Notes
network traffic from
agents to controller.
On-premises Cloud On-premises Cloud Larger
concurrency,
network traffic from
agents and tested
system to
controller.
On-premises Cloud Cloud On-premises Larger
concurrency,
network traffic from
agents to controller
and back to the
repository.
On-premises Cloud Cloud Cloud Larger
concurrency,
network traffic from
agent and tested
system to controller
and back to
repository.

This table shows the second major division. In this table, the controller runs in the cloud.
Controller Runs in Cloud
Controller Agents Repository Tested System Notes
Cloud On-premises On-premises On-premises Limited
concurrency,
network traffic from
agent and test
system to controller
and back to
repository.
Cloud On-premises On-premises Cloud Limited
concurrency,
network traffic from
agent to controller
and back to
repository.
Cloud On-premises Cloud On-premises Limited

361
Controller Agents Repository Tested System Notes
concurrency,
network traffic from
agent and test
system to
controller.
Cloud Cloud Cloud Cloud Larger
concurrency,
network traffic from
agents to controller.
Cloud Cloud On-premises On-premises Larger
concurrency,
network traffic from
tested system to
controller and back
to repository.
Cloud Cloud On-premises Cloud Larger
concurrency,
network traffic from
controller to
repository.
Cloud Cloud Cloud On-premises Larger
concurrency,
network traffic from
tested system to
controller.
Cloud Multi-cloud Cloud Cloud Largest
concurrency, at
least from tested
system in DC1 to
controller in DC2
(possibly more,
depending on
setup).

Recommendations
The topologies are shown for the sake of completeness. The choice of a topology may not be
yours. For example, if you require more than 100 GB of SQL Server data storage, you must store
it on-premises; currently, 100 GB is the limit of SQL Azure. However, if you have some leeway,
here are the best topologies. These recommendations are the most efficient and give the highest
concurrency levels in the two major divisions (controller on-premises or controller on-cloud).

362
Best When Controller Runs On-Premises
Controller Agents Repository Tested System Notes
On-premises Cloud On-premises On-premises Larger concurrency,
network traffic from
agents to controller.
On-premises Cloud On-premises Cloud Larger concurrency,
network traffic from
agents and tested
system to controller.

Best When Controller Runs in Cloud
Controller Agents Repository Tested System Notes
Cloud On-premises Cloud On-premises Limited
concurrency,
network traffic from
agent and test
system to
controller.
Cloud Cloud Cloud Cloud Larger
concurrency,
network traffic from
agents to controller.
Cloud Multi-cloud Cloud Cloud Largest
concurrency, at
least from tested
system in DC1 to
controller in DC2
(possibly more,
depends on setup)

Increased Complexity Demands
In the real world, your actual layout of components can be complex. A system under test can
have many sub-components, each running as a worker role or web role, or using on-premises
services. For example, a component initializes the system, or provides validation services. It is
likely then that many of the components must communicate to other parts of the system. This set
of components is in addition to the load-test system itself (consisting of controller, agents, and
repository). The following graphic shows a system that has many parts. It is only meant to
illustrate that a real solution can many pieces, and with many communication requirements. The
details of the system are not the main point.

363


Given this amount of complexity, it is still possible to site various components in Windows Azure,
or on-premises, as needed. The next section outlines the required steps.
Setup
The setup here is a set of alternatives to the basic setup found in this document: Provisioning
Windows Azure For a Load Test. The procedures there specify how to set up the Management
Portal with the correct pieces. For example, it explains how to set up a storage account, and how
to set up a Windows Azure Connect group.
The first alternative here enables you to use SQL Azure as a results repository. The second
alternative instructs how to set up endpoints on each worker role. Endpoints enable
communication between the agents, tested system, the test controller, and (if needed) an
auxiliary factorthe workstation that hosts Visual Studio. In the cleanest mode, the machine is

364
sited on-premises. But if your application requires you to use it during the test execution, it can
also be hosted as a Windows Azure worker role.
Prerequisites
The following are required:
Download the LoadTest2010.bacpac file.
Optional. Install Visual Studio 2010 Ultimate on a worker role.
If you intend to run the test controller on a worker role, install Visual Studio 2010 Ultimate on
the worker role. Add a worker role to the Windows Azure project in Visual Studio. Ensure that
Remote Desktop is enabled. Add the worker role to a Connect group in order to connect it to
your on-premises site. Use Remote Desktop to connect to the worker role, and thence install
Visual Studio 2010 Ultimate.
Using a SQL BACPAC to Provision SQL Azure
Use this method to store results on a SQL Azure database to store the load test data. Upload the
file named LoadTest2010.bacpac to your Azure Storage account. You must also have the
following prerequisites:
1. A SQL Azure account.
2. A SQL server instance created in a subscription. For more information, see How to Create a
SQL Azure Server.
3. A log-in name and password that has permission to modify the server instance.
1. Upload the LoadTest2010.bacpac file to your Azure Storage account.
Use the StorageServicesSmartClient to upload the file to a public container. Once the
.dacpac file is uploaded, you can use the Import function of the Management Portal to
create the database.
2. In the Management Portal, click Database.
3. Expand the subscription that contains the SQL Azure server you use, and select the
server.
4. In the Database section of the ribbon, click Import.
5. Follow the instructions here to import the LoadTest2010.bacpac file: How to: Import a
Data-tier Application (SQL Azure).
Once the repository is created in SQL Azure, configure the Test Controller. Specifically, set the
connection string for the results store. The dialog for setting the value is found only in Visual
Studio Ultimate. However before you proceed, you have a choice to make. Decide whether to run
the Test Controller:
1. From an on-premises computer.
If you this option, from an on-premises computer, follow the next set of instructions.
2. From a worker role on Windows Azure, running an instance of Visual Studio.
If you have decided to run the Test Controller from a worker role on Windows Azure:
a. Create the Windows Azure Connect Group.
To provision SQL Azure with a BACPAC

365
b. Add the on-premises workstation to the group.
c. After deploying the hosted service, use Remote Desktop to connect to the worker role
that will host Visual Studio.
d. Install Visual Studio 2010 Ultimate from a setup (.msi) file on your network
1. Run Visual Studio as an Administrator.
2. In Visual Studio, on the Test menu, click Manage Test Controllers.
3. Under the text Load test results store, click the ellipsis button.
4. In the Connection Properties dialog box, paste in the name of the SQL Azure server.
5. Under the text Log on to the server, select the Use SQL Server Authentication option.
6. Set the User name value to the log-in name for the SQL Azure administrator.
7. Set the Password field value to the password value for the SQL Azure administrator.
8. In the Connect to a database section, select the LoadTest2010 database.
If the connection string and the credentials are correct, examine the Configure Test
Controller dialog box. The dialog is populated with correct values.
Using Endpoints Instead of an Azure Connect Group
There is an alternative to using the Connect group: configuring endpoints on each worker role to
communicate between the instances. There is some advantage here in that the connection is
more reliable. Use the following steps to try this alternative.
1. In Visual Studio, open the Load Test project.
2. Configure each worker role with internal TCP endpoints. For general instructions on
adding endpoints to a role, see How to Define Internal Endpoints for a Role. For each
endpoint, specify a different private port number. The following table shows the endpoint
configurations for each role. Notice that all of these endpoints are private ports using the
TCP prototol:

Role Name Port Name Port Number
Agent SQL 1433
Agent WFS 137-445
Agent Agent Listener Port 6910
Agent AgentDownload 80
Controller SQL 1433
Controller WFS 137-445
Controller Controller 6910
To configure the On-Premises instance of the Test Controller
To configure endpoints on worker role instances

366
Controller AgentDownload 80
Controller Controller Listener Port 6901
Controller Agent 6910
Controller/Workstation Talkback 49152

3. Configure a specific port to allow the controller to send messages to the agents. Use the
following registry key on the worker role that hosts the controller:
a.
HKEY_LOCAL_MACHINE\SOFTWARE\MICROSOFT\VisualStudio\10.0\Enterpri
seTools\QualityTools\ListenPortRange\PortRangeStart
b.
HKEY_LOCAL_MACHINE\SOFTWARE\MICROSOFT\VisualStudio\10.0\Enterpri
seTools\QualityTools\ListenPortRange\PortRangeEnd
4. In the startup of the worker role, run the setupwfw.cmd file.
5. Remote Desktop to each role and do the following:
a. Open the following directory: \Windows\System32\drivers\etc\
Open the hosts file to edit it. The file can be opened using Notepad.exe. Place the
"hosts" file on each system that has the host name and IP address. Editing the file is
a manual process. To find the IP address of a role, open a command window on
each role and type IPCONFIG. An example of the Hosts file:
b.
6. Each role can communicate and you can install the binaries on each system. To speed
up the process, place the binaries in blob store. Also, run additional commands that as
part of the startup task. For more information, see How to Define Startup Tasks for a
Role.)
7. On SQL Server, create a local SQL Account for the controller and workstation to access
the results database. Then configure the repository to use those accounts. Create the
database when you set up the controller and then configure it to use the account from the
IDE.



367
Guidance on Efficiently Testing Azure Solutions
Author: Suren Machiraju
Reviewers: Jaime Alva Bravo and Steve Wilkins
After designing, coding and deploying your Azure AppFabric solution you may find that it does not
work. This article provides guidance on how to efficiently test your Azure AppFabric applications
through the software development life cycle both at the business logic and the comprehensive
end-to-end scenario tests. In this article, we demonstrate how to:
Develop unit tests for the business logic components while eliminating dependencies on
Azure AppFabric components.
Design integration end-to-end tests and eliminate overhead for setup, initialize, cleanup and
teardown of the Azure AppFabric Service resources (for example, Queues) for every test run.
Further eliminate creating duplicate infrastructures for ACS namespaces, Service Bus
Queues, and other resources.
Additionally, this article provides an overview of the various testing technologies and techniques
for testing your Azure AppFabric applications.
The Types of Tests
In this article, we will focus on the following tests:
Unit Tests are narrowly focused tests designed to exercise one specific bit of functionality.
These tests are often referred to as the Code Under Test or CUT. Any dependencies taken
by the CUT are somehow removed.
Integration Tests are broader tests for exercising multiple bits of functionality at the same
time. In many cases, these resemble unit tests that cover multiple feature areas and include
multiple dependencies.
Overall our tests focus on creating and using Test Doubles. We will use the following types of
Test Doubles:
Mocks are simulated objects that imitate the behavior of real objects in controlled ways. They
are replacements of what is dynamically produced by a framework.
Fakes are simulated objects that implement the same interface as the object that they
represent. Fakes return pre-defined responses. Fakes contain a set of method stubs and
serve as replacements for what you build by hand.
Stubs simulate the behavior of software objects.
Our tests, upon executing can verify both state (for example, after you make a call a specific
value is returned) and behavior (the method is called in a certain order or a certain number of
times).
The Azure Dependencies
One of the major goals of unit testing is to eliminate dependencies. For the Azure framework,
these dependencies include the following:
Service Bus Queues
Access Control Service

368
Cache
Windows Azure Tables, Blobs, Queues
SQL Azure
Cloud Drive
Other Web Services
When building tests for Azure Applications we will replace these dependencies to better focus our
tests on exercising the desired logic.
In this article, we will provide examples of Service Bus Queues with full knowledge that the tools
and techniques demonstrated here will also apply to all other dependencies.
The Testing Framework
To implement the testing framework for your Azure AppFabric applications, you will need:
A unit a testing framework to define and run your tests.
A mocking framework to help you isolate dependencies and build narrowly scoped unit tests.
Tools to help with automatic unit test generation for increased code coverage.
Other frameworks that can help you with testable designs by taking advantage of
dependency injection and applying the Inversion of Control Pattern.
Unit Testing with MS Test
Visual Studio includes a command line utility MS Test that you can use to execute unit tests
created in Visual Studio. Visual Studio also includes a suite of project and item templates to
support testing. You typically create a new test project and then add classes (known as test
fixtures) adorned with [TestClass] attributes that contain methods adorned with [TestMethod].
Within MS Test, various windows within Visual Studio enable you to execute unit tests defined in
the project, and review the results after you execute them.
Visual Studio Express and Test Professional editions do not contain MS Test.
In MS Test, unit tests follow an AAA pattern - Arrange, Act, Assert.
Arrange - build up any pre-requisite objects, configurations, and all other necessary
preconditions and inputs needed by the CUT.
Act - perform the actual, narrowly scoped test on the code.
Assert - verify that the expected results have occurred.
The MS Test framework libraries include PrivateObject and PrivateType helper classes that use
reflection to make it easy to invoke non-public instance members or static members, from within
the unit test code.
Premium and ultimate editions of Visual Studio include enhanced unit test tooling that integrate
with MS Test that enable you to analyze the amount of code exercised by your unit tests, and
visualize the result by color coding the source code, in a feature referred to as Code Coverage.
Note

369
Mocking with Moles
A goal of unit testing is to be able to test in isolation. However, often the code under test cannot
be tested in isolation. Sometimes the code is not written for testability. To rewrite the code would
be difficult because it relies on other libraries that are not easily isolated, such as those that
interact with external environments. A mocking framework helps you isolate both types of
dependencies.
A list of mocking frameworks you might consider is available in the links section at the end of this
article. For our purposes, we will focus on how to use Moles, a mocking framework from Microsoft
Research.
Most mocking frameworks will create mock types dynamically by deriving from the type you
indicate for the members you want to change. Fewer frameworks have support for handling
sealed classes, non-virtual methods, static members, delegates or events that require alternative
techniques (such as using the .NET Profiling API). Moles framework provides this advanced
support.
Most of the mocking frameworks do not enable you to rewrite the logic within a constructor. If you
are creating an instance of a type that you have to isolate, and the constructor of this type takes
dependencies you want to isolate, you will have to creatively mock the calls made by the
constructor instead of the constructor itself. Moles can be used to insert logic before constructor
calls, but it stops short of letting you completely rewrite the constructors logic.
Moles provides all these features at no cost to you. In addition, it integrates well with Visual
Studio 2010 and MS Test, and tests generated with it are fully debuggable within Visual Studio.
Automated Test Creation with PEX
PEX (Program EXploration) is a white-box test generation tool available to MSDN subscribers
from Microsoft Research. You can use PEX to automatically create unit tests with high code
coverage by intelligently selecting input parameters for CUT, and recursively for parameterized
unit tests. PEX is capable of building the smallest number of tests that produce the greatest code
coverage.
PEX installs and runs fully integrated within Visual Studio 2010, and installs Moles as part of the
installation process.
Frameworks for Dependency Injection and Inversion of Control (IoC)
Containers
You can use Microsoft Unity for extensible dependency injection and IoC. It supports interception,
constructor injection, property injection, and method call injection.
Microsoft Unity and similar tools help you build testable designs that let you insert your
dependencies at all levels of your application, assuming that your application was designed and
built with dependency injection and one of these frameworks in mind.
These frameworks are great for writing testable code, and ultimately good code, but can be
somewhat heavy in their up-front design requirements. We will not cover DI or IoC containers in
this article.
Test
In this
mess
intere
Unit
Let us
by us

When
contro
[HttpP
public
{
tr
{


RoleEn

ting Azur
s section, we
ages to a Qu
ested in testin
Testing Co
s say you hav
ing an AppFa
n the user clic
oller. The acti
Post]
c ActionResu
ry
string qN
string qN
nvironment.G
string qU
re Solutio
will describe
eue and a wo
g aspects of a
ontroller A
ve a website t
abric Queue. O
cks Create, it
ion is implem
lt Create(Or
ame = RoleEn
amespace =
etConfigurat
sername = Ro
ons
a solution tha
orkers role th
all three.
Actions in a
that creates o
Our example
posts the new
ented as follo
rder order)
nvironment.Ge
ionSettingVa
leEnvironmen
at includes a w
at processes
a Web Role
orders, and th
web page loo
w order to the
ows:
etConfigurati
alue("QueueSe
nt.GetConfigu
web role host
messages fr
e
ese orders ar
oks like this:
e Create actio
ionSettingVal
erviceNamespa
urationSettin
ted web site t
rom the Queu
re queued for
on on the asso
lue("QueueNam
ace");
ngValue("Queu

37
that pushes
e. We are
r processing

ociate
me");
ueUsername");
70
;

371
string qPassword = RoleEnvironment.GetConfigurationSettingValue("QueuePassword");

//Push order to AppFabric Queue
AppFabricQueue queue = new AppFabricQueue(new QueueSettings(qNamespace, qName,
Username, qPassword));

queue.Send(order);

queue.CloseConnections();

return View("OrderCreated");
}
catch (Exception ex)
{
Trace.TraceError(ex.Message);
return View("Error");
}
}
Notice that the code retrieves settings from configuration through RoleEnvironment and sends a
message that contains the order to the queue. The AppFabricQueue class we use here is a
wrapper class that uses the Service Bus .NET APIs (MessageSender, MessageReceiver, etc.) to
interact with Service Bus Queues.
Let us build a unit test for this method using Moles. Notice that the Create action uses the
following methods:
GetConfigurationSettingsValue (RoleEnvironment class)
Send (AppFabricQueue class)
What we want to do is build detours for these methods using Moles to control their behavior and
remove the dependencies on the real environment. Controlling behavior of these methods and
removing their dependencies on the real environment will eliminate the need to run the tests in
the emulator, on Azure, or to call out to the Service Bus Queue. We will exercise the Create
action on the controller to verify that the Order input is the one sent to the queue (checking that it
has the Order Id and Description as input to the action) and that the OrderCreated view is
displayed as a result.
Accomplishing this with moles is easy. Within the Test project, right click the assembly containing
the types you want to mock and select Add Moles Assembly.
In our example, we select Microsoft.Windows.Azure.ServiceRuntime and select Add Moles
Assembly. An XML file named Microsoft.WindowsAzure.ServiceRuntime.moles will be added to
the test project. Then repeat for the AF.Queues assembly.
When you build the Test project, references will be added to the auto-generated moled versions
of these assemblies.
In our example these are Microsoft.Windows.Azure.ServiceRuntime.Moles and
AF.Queues.Moles.

372
Create a new unit test method and decorate it with [HostType("Moles")]. Within the Unit Test, you
use the "moled" types by accessing them through a Moles namespace. All moled types start with
an M.
For example, the moled type for AppFabricQueue is accessed through
AF.Queues.Moles.MAppFabricQueue.
You define your detours (the alternate logic to execute) by assigning a delegate to the method
you want to replace.
For example, to detour RoleEnvironment.GetConfigurationSetting we define a delegate for
MRoleEnvironment.GetConfigurationSettingValueString.
The complete unit test is as follows. Observe the detours we provide for
RoleEnvironment.GetConfigurationSetting, the QueueSettings constructor, and the Send and
CloseConnection methods of the AppFabricQueue class.
[TestMethod]
[HostType("Moles")]
public void Test_Home_CreateOrder()
{
//Arrange

Microsoft.WindowsAzure.ServiceRuntime.Moles.MRoleEnvironment.GetConfigurationSettingValue
String = (key) =>
{
return "mockedSettingValue";
};

bool wasQueueConstructorCalled = false;
AF.Queues.Moles.MAppFabricQueue.ConstructorQueueSettings = (queue, settings) =>
{
wasQueueConstructorCalled = true;
};

Order orderSent = null;
AF.Queues.Moles.MAppFabricQueue.AllInstances.SendObject = (queue, order) =>
{
orderSent = order as Order;
};

bool wasConnectionClosedCalled = false;
AF.Queues.Moles.MAppFabricQueue.AllInstances.CloseConnections = (queue) =>
{
wasConnectionClosedCalled = true;
};


373
Order inputOrder = new Order()
{
OrderId = "Test123",
Description = "A mock order"
};

HomeController controller = new HomeController();

//Act
ViewResult result = controller.Create(inputOrder) as ViewResult;

//Assert
Assert.IsTrue(wasConnectionClosedCalled);
Assert.IsTrue(wasQueueConstructorCalled);
Assert.AreEqual("OrderCreated", result.ViewName);
Assert.IsNotNull(orderSent);
Assert.AreEqual(inputOrder.OrderId, orderSent.OrderId);
Assert.AreEqual(inputOrder.Description, orderSent.Description);

}
Unit Testing Worker Roles that use AppFabric Queues
The web role took care of adding an order to the queue. Let us consider a workers role that
would process the orders by retrieving them from the queue and see how we can test that. The
core item to test in the worker role is defined in the implementation of RoleEntryPoint, typically in
the OnStart and Run methods.
The Run method in our worker role periodically polls the queue for orders and processes them.
public override void Run()
{
AppFabricQueue queue = null;

try
{
string qName = RoleEnvironment.GetConfigurationSettingValue("QueueName");
string qNamespace =
RoleEnvironment.GetConfigurationSettingValue("QueueServiceNamespace");
string qUsername =
RoleEnvironment.GetConfigurationSettingValue("QueueUsername");
string qPassword =
RoleEnvironment.GetConfigurationSettingValue("QueuePassword");


374
queue = new AppFabricQueue(new QueueSettings(qNamespace, qName,
qUsername, qPassword));
queue.CreateQueueIfNotExists();

while (true)
{
Thread.Sleep(2000);

//Retrieve order from AppFabric Queue
TryProcessOrder(queue);
}
}
catch (Exception ex)
{
if(queue !=null)
queue.CloseConnections();

System.Diagnostics.Trace.TraceError(ex.Message);
}

}
We want a test verifying that the routine correctly retrieves a message.
In this case we add a moled assembly for the worker role. For our specific project, this is the
Mvc3Web assembly because we host the RoleEntryPoint there. The following is the complete
unit test for run method of the worker role.
[TestMethod]
[HostType("Moles")]
public void Test_WorkerRole_Run()
{
//Arrange

Microsoft.WindowsAzure.ServiceRuntime.Moles.MRoleEnvironment.GetConfigurationSettingValue
String = (key) =>
{
return "mockedSettingValue";
};

bool wasQueueConstructorCalled = false;
AF.Queues.Moles.MAppFabricQueue.ConstructorQueueSettings = (queue, settings) =>
{
wasQueueConstructorCalled = true;
};

375

bool wasEnsureQueueExistsCalled = false;
int numCallsToEnsureQueueExists = 0;
AF.Queues.Moles.MAppFabricQueue.AllInstances.CreateQueueIfNotExists = (queue) =>
{
wasEnsureQueueExistsCalled = true;
numCallsToEnsureQueueExists++;
};

bool wasTryProcessOrderCalled = false;
int numCallsToTryProcessOrder = 0;
Mvc3Web.Worker.Moles.MWorkerRole.AllInstances.TryProcessOrderAppFabricQueue =
(actualWorkerRole, queue) =>
{
wasTryProcessOrderCalled = true;
numCallsToTryProcessOrder++;

if (numCallsToTryProcessOrder > 3) throw new Exception("Aborting Run");
};

bool wasConnectionClosedCalled = false;
AF.Queues.Moles.MAppFabricQueue.AllInstances.CloseConnections = (queue) =>
{
wasConnectionClosedCalled = true;
};

Mvc3Web.Worker.WorkerRole workerRole = new Worker.WorkerRole();

//Act
workerRole.Run();

//Assert
Assert.IsTrue(wasConnectionClosedCalled);
Assert.IsTrue(wasQueueConstructorCalled);
Assert.IsTrue(wasEnsureQueueExistsCalled);
Assert.IsTrue(wasTryProcessOrderCalled);
Assert.AreEqual(1, numCallsToEnsureQueueExists);
Assert.IsTrue(numCallsToTryProcessOrder > 0);
}
This time we focus on detouring the instance methods of the WorkerRole class. You should set
the delegate of the AllInstances property of the generated mole type. By using the delegate, any
instan
define
In our
instan
detou
throw
time.
You m
classe
comp
fakes
create
appro
Queu
requir
actua
the de
then t
Servic
Ena
To an
both o
from t

By clic
Test R
Howe
that n
before
Code
Soluti
If
N
N
nce you creat
ed delegates.
r example, we
nce methods
urs we have a
w an exception
may be asking
es from the S
letely isolate
that inherit fr
e mock types
oaches, you w
eClient, burd
red overrides,
lly rely on, an
ependency on
that is all you
ce Bus.
abling Co
nalyze what th
of the unit tes
the Test Resu
cking the Sho
Results windo
ever, if we do
no coverage d
e running our
coverage da
ion Items. Op
you have a s
New Item to th
Note
e of the actua

e want to use
CreateQueue
lready seen f
n so that the i
g yourself, wh
Service Bus S
our code from
rom the abstra
for them all.
will ultimately f
ened with cre
, struggling to
nd recreating
n the real Ser
need to moc
de Cover
hese unit tests
sts with MS Te
ults dialog tha
ow Code Cov
ow), we can s
this now, we
data could be
r tests.
ta collection i
pening this file
solution that d
he solution an
al type will de
the original in
eIfNotExists a
for RoleEnviro
nfinite loop m
hy not just use
DK directly in
m calling the r
act classes in
The problem
find yourself r
eating derived
o expose the i
their construc
rvice Bus. On
k and detour
rage
s actually test
est, we can s
at appears.
verage Result
see the code c
will not see a
found. We fir
is enabled sim
e brings up th
does not have
d then select
tour through
nstances Ru
and TryProces
onement.GetC
maintained wit
e the Messag
nstead of injec
real-world Se
n the Microsof
with either ap
reflecting into
d types from t
internal types
ctors or factor
the other han
in order to iso
ted, we can e
ee that they p
s button (the
covered by al
any data. Inste
rst have to en
mply by config
e Test Setting
e Local.testse
Test >Test S
any of the me
n method, bu
ssOrder (in ad
Configuration
thin Run is ex
geSender/Mes
cting a helper
rvice Bus, we
ft.ServiceBus
pproach is co
o classes like
hese abstrac
s that real ver
ry methods in
nd, if you inse
olate yourself
examine code
pass and also
rightmost but
ll unit tests ex
ead, we will g
nable collectio
guring the Loc
gs editor.
ettings file, you
Settings.
ethods for wh
ut provide deto
ddition to the
nSettings). In
xited at a pred
ssageReceive
r type? In orde
e have two ch
s namespace,
omplexity. Wit
TokenProvid
t types that e
rsions of these
n clever ways
ert your own h
f from the rea
e coverage da
o see related
tton on near t
xecuted in the
get a message
on of code cov
cal.testsetting
u can add it u

37
ich you have
ours for the
static method
our code we
determined
er and related
er to
hoices - write
, or let moles
th both
er and
xpose all thei
e classes
just to excise
helper type,
al-world
ata. If we run
run details

the top of the
e run.
e indicating
verage data
gs under the
using Add
76
d
d
ir
e

Click
Next,
Code

the Data and
click the Con
Coverage De
Diagnostics
nfigure button
etail window s
tab, and the c
(the one just
select all your
check the box
t above the gr
r assemblies
x to the right o
rid containing
under test an
of the Code C
g Code Cover
nd click OK.

37

Coverage row
rage). In the

77
w.
Then
code

Click
doubl
follow

Usin
Manu
unit te
install
editor
click Apply a
coverage but
the Show Co
e-click it to se
wing figure), p
ng PEX to
ually creating t
ests to the ne
ling PEX, you
r, and selectin
nd Close to d
tton. Now you
de Coverage
ee the source
artially tested
o Enhanc
these unit tes
xt level by try
u can have it e
ng Run PEX.
dismiss the Te
ur Code Cove
Coloring icon
e code colored
d or untested
ce Code C
sts is very val
ying paramete
explore a met
est Settings e
erage Results
n, and then d
d to reflect ar
(gray) the uni
Coverage
uable, but PE
er values you
thod simply b
editor. Re-run
should look a
rill down to a
eas that were
it tests just ru
e
EX can help y
may not have
by right clickin
your tests an
as follows:
method in the
e tested (gree
un.
you intelligent
e considered
ng the method

37
nd press the

e grid and
en in the

ly take your
. After
d in the code
78

After a
here i
PEX w
crafte
impor
gener

Try
The s
Use

Moc
Moles
Moq
NMoc
Rhino
EasyM
NSub
FakeI
TypeM
JustM
a short while,
is what we go
was able to c
ed that genera
rtant, in the de
rated tests ca
it out!
sample used i
eful Links
king Fram
s
ck3
o Mocks
Mock.NET
bstitute
tEasy
Mock Isolator
Mock
, PEX will finis
ot when ran P
reate one tes
ated that exce
etails pane yo
an be added to
n this article i
s
eworks
r
sh with its pro
PEX on our wo
st that resulted
eption (a null v
ou can see th
o your projec
is available at
ocessing and
orker role's Tr
d in an excep
value for the
e code PEX w
t simply by se
t sample_link
the results w
ryProcessOrd
ption, and it sh
queue param
was executing
electing them
k.
will stabilize. F
der method. N
hows you the
meter). What is
g. The results
and clicking

37

For example,
Notice that
inputs it
s even more
s of all PEX
Promote.

79

380
Unit Test Related
MSDN Verifying Code by Using Unit Tests
MSDN - Unit Testing in MVC Applications
Building Testable Applications
Pex and Moles
PrivateObject and PrivateType Introduced
CLR Profiling API Blog
Mocks Aren't Stubs
Dependency Injection and IoC
Microsoft Unity
Structure Map
MSDN Magazine - Aspect Oriented Programming, Interception and Unity 2.0
MSDN Magazine - Interceptors in Unity
MSDN Magazine - MEF v.s. IoC


381
Database


382
Data Migration to SQL Azure: Tools and
Techniques
This document provides guidance on migrating data definition (schema) and data to SQL Azure.
It is primarily for one-time migration from SQL Server to SQL Azure. For on-going data sharing
and SQL Azure backup, see SQL Azure Data Sync Overview.
Migration Considerations
Microsoft Windows Azure offers several choices for data storage. You may choose to use one or
many in your projects.
SQL Azure Database is SQL Server technology delivered as a service on the Windows Azure
Platform. The cloud-based SQL Azure database solutions can provide many benefits, including
rapid provisioning, cost-effective scalability, high availability, and reduced management overhead.
SQL Azure supports the same tooling and development practices used for on-premise SQL
Server applications. Therefore, it should be a familiar experience for most developers.
The long term goal for SQL Server and SQL Azure is for symmetry and parity in both features
and capabilities, however there are currently differences in architecture and implementation that
need to be addressed when migrating databases to SQL Azure and developing SQL Azure
solutions.
Before migrating a database to SQL Azure, it is important to understand when to migrate and the
differences between SQL Azure and SQL Server.
When to Migrate
There are three main storage offerings on the Windows Azure platform. Windows Azure Storage
contains Table, Blob, and Queue. When designing a Windows Azure solution, you shall evaluate
different options, and use each storage mechanism to provide the best performance for the part
of the solution.

Storage Offering Purpose Maximum Size
SQL Azure Database Relational database
management system
50GB
Windows Azure
Storage
Blob Durable storage for large
binary objects such as
video or audio
200GB or 1TB
Table Durable storage for
structured data
100TB
Queue Durable storage for inter-
process messages
100TB
Local Storage Per-instance temporary
storage
250GB to 2TB

383

Local storage provides temporary storage for a local running application instance. A local store is
only accessible by the local instance. If the instance is restarted on different hardware, such as in
the case of hardware failure or hardware maintenance, data in the local store will not follow the
instance. If your application requires reliable durability of the data, want to share data between
instances, or access the data outside of Windows Azure, consider using a Windows Azure
Storage account or SQL Azure Database instead.
SQL Azure provides data-processing capabilities through queries, transactions and stored
procedures that are executed on the server side, and only the results are returned to the
application. If you have an application that requires data processing over large data sets, then
SQL Azure is a good choice. If you have an application that stores and retrieves large datasets
but does not require data processing, then Windows Azure Table Storage is a better choice.
Because the SQL Azure size limitation is currently set at 50 GB, and SQL Azure is much more
expensive than Windows Azure storage, you may consider moving the blob data into Windows
Azure Blog Storage. So that you can reduce the pressure on the database size limit and reduce
the operational cost.

For more information, see Data Storage Offerings on the Windows Azure platform.
Compare SQL Azure to SQL Server
Similar to SQL Server, SQL Azure exposes a tabular data stream (TDS) interface for Transact-
SQL-based database access. This allows your database applications to use SQL Azure in the
same way that they use SQL Server.
Unlike SQL Server administration, SQL Azure abstracts the logical administration from the
physical administration; you continue to administer databases, logins, users, and roles, but
Microsoft administers and configures the physical hardware such as hard drives, servers, and
storage. Since Microsoft handles all of the physical administration, there are some differences
between SQL Azure and SQL Server in terms of administration, provisioning, Transact-SQL
support, programming model, and features.
The following list provides a high-level overview of some of the major differences:
Database Size
SQL Azure currently offers two editions:
Web Editions, in 1GB and 5GB sizes.
Business Editions, in 10, 20, 30, 40 and 50GB sizes.
It is important to check the size of your database and how it fits within the database
allowances used by SQL Azure. If your database is larger than SQL Azure size limitation,
then you must examine your database and see if it can be broken down into smaller
databases (i.e. sharding), or moving large data to Window Azure blob storage. For more
information on database sharding, see Federation: Building Scalable, Elastic, and Multi-
tenant Database Solutions with SQL Azure.
Authentication

384
SQL Azure supports only SQL authentication. You must consider whether changes are
needed to the authentication scheme used by your application. For more information on the
security limitations, see Security Guidelines and Limitations.
SQL Server Database Version
SQL Azure is based on SQL Server 2008 (level 100). If you want to migrate your SQL Server
2000 or SQL Server 2005 databases to SQL Azure, you must make sure your databases are
compatible with SQL Server 2008. You will find the best path is to migrate from a SQL Server
2008 to SQL Azure. You can go through an on-premises upgrade to SQL Server 2008 before
you migrate to SQL Azure. Here are some great resources to help you with migrating from
older versions of SQL Server: Upgrading to SQL Server 2008 R2 and Microsoft SQL Server
2008 Upgrade Advisor.
Schema
SQL Azure does not support heaps. ALL tables must have a clustered index before data can
be inserted. For more information on the clustered index requirement, see Inside SQL Azure.
Transact-SQL Supportability
SQL Azure Database supports a subset of the Transact-SQL language. You must modify the
script to only include supported Transact-SQL statements before you deploy the database to
SQL Azure. For more information, see Supported Transact-SQL Statements, Partially
Supported Transact-SQL Statements, and Unsupported Transact-SQL Statements.
The Use Statement
In Microsoft SQL Azure database, the USE statement does not switch between databases.
To change databases, you must directly connect to the database.
Pricing
Pricing for your SQL Azure subscription is per database, and it is based on the edition. There
are also additional charges for data transfer volume, any time data comes into or out of the
data center. You have your choice of running your application code on your own premises
and connecting to your SQL Azure database in the data center, or running your application
code in Windows Azure, which is hosted in the same data center as your SQL Azure
database. Running application code in Windows Azure avoids the additional data transfer
charges. In either case, you should be aware of the Internet network latency that cannot be
mitigated using either model. For more information, see Pricing Overview.
Feature Limitations
Some of the SQL Server features are not currently supported by SQL Azure. They include:
SQL Agent, Full-text search, Service Broker, backup and restore, Common Language
Runtime, and SQL Server Integration Services. For a detailed list, see SQL Server Feature
Limitations.
Connection Handling
When using a cloud based database like SQL Azure, it requires connections over the internet or
other complex networks. Because of this, you should be prepared to handle unexpected dropping
of connections.

385
SQL Azure provides a large-scale multi-tenant database service on shared resources. In order to
provide a good experience to all SQL Azure customers, your connection to the service may be
closed due to several conditions.
The following is a list of causes of connection terminations:
Network latency
Latency causes an increase in time required to transfer data to SQL Azure. The best way to
mitigate this effect is to transfer data using multiple concurrent streams. However, the
efficiency of parallelization is capped by the bandwidth of your network.
SQL Azure allows you to create your database in different datacenters. Depending on your
location and your network connectivity, you will get different network latencies between your
location and each of the data centers. To help reducing network latency, select a data center
closest to majority of your users. For information on measuring network latency, see Testing
Client Latency to SQL Azure.
Hosting your application code in Windows Azure is beneficial to the performance of your
application because it minimizes the network latency associated with your application's data
requests to SQL Azure.
Minimizing network round trips can also help with reducing network related problems.
Database Failover
SQL Azure replicates multiple redundant copies of your data to multiple physical servers to
maintain data availability and business continuity. In the case of hardware failures or
upgrades, SQL Azure provides automatic failover to optimize availability for your application.
Currently, some failover actions result in an abrupt termination of a session.
Load Balance
Load balancer in SQL Azure ensures the optimal usage of the physical servers and services
in the data centers. When the CPU utilization, input/output (I/O) latency, or the number of
busy workers for a machine exceeds thresholds, SQL Azure might terminate the transactions
and disconnect the sessions.
Throttling
To ensure that all subscribers receive an appropriate share of resources and that no
subscriber monopolizes resources at the expense of other subscribers, SQL Azure may close
or throttle subscriber connections under certain conditions. SQL Azure Engine Throttling
service continually monitors certain performance thresholds to evaluate the health of the
system and may initiate varying levels of throttling to particular subscribers depending on the
extent to which these subscribers are impacting system health.
The following performance thresholds are monitored by SQL Azure Engine Throttling:
Percentage of the space allocated to a SQL Azure physical database which is in use, soft
and hard limit percentages are the same.
Percentage of space allocated for SQL Azure log files that is in use. Log files are shared
between subscribers. Soft and hard limit percentages are different.
Milliseconds of delay when writing to a log drive, soft and hard limit percentages are
different.

386
Milliseconds of delay when reading data files, soft and hard limit percentages are the
same.
Processor usage, soft and hard limit percentages are the same.
Size of individual databases relative to maximum size allowed for database subscription,
soft and hard limit percentages are the same.
Total number of workers serving active requests to databases, soft and hard limit
percentages are different. If this threshold is exceeded, the criteria of choosing which
databases to block are different than in the case of other thresholds. Databases utilizing
the highest number of workers are more likely to be throttling candidates than databases
experiencing the highest traffic rates.
For more information, see SQL Azure Connection Management, and SQL Azure Performance
and Elasticity Guide.
The best way to handle connection loss it to re-establish the connection and then re-execute the
failed commands or query. For more information, see Transient Fault Handling Framework.
Optimize Databases for Data Import
There are a few things you can do to the databases to improve the migration performance:
Delay creation of non-clustered indexes or disable non-clustered indexes. Addition indexes
created before loading the data can significantly increase the final database size and the time
taken to load the same amount of data.
Disable triggers and constrain checking. The triggers may fire when a row is inserted into a
table causing the row to be reinserted into another table. The trigger can cause delays and
you might not want those types of inserts to be reinserted.
Bulk import performance is improved if the data being imported is sorted according to the
clustered index on the table. For more information, see Controlling the Sort Order When Bulk
Importing Data.
Transfer Large Data to SQL Azure
SQL Server Integration Services (SSIS) and bcp utility perform well with large data migration.
When loading large data to SQL Azure, it is advisable to split your data into multiple concurrent
streams to achieve the best performance.
By default, all the rows in the data file are imported as one batch. To distribute the rows among
multiple batches, specify a batch size whenever it is available. If the transaction for any batch
fails, only insertions from the current batch are rolled back. Batches already imported by
committed transactions are unaffected by a later failure. It is best to test a variety of batch size
settings for your particular scenario and environment to find the optimal batch size.
Choose Migration Tools
There are various tools available for migration database to SQL Azure. In general, database
migration involves schema migration and data migration. There are tools supporting one of each
or both. You could even use the Bulk Copy API to author your own customized data upload
application.

387
Migrate from SQL Server


Tools Schema SQL Azure
Compatibility
Check
Data Data Transfer
Efficiency
Note
DAC Package Yes Yes No N/A Entity
containing all
database
objects, but no
data
Full SQL Azure
support
DAC BACPAC
Import Export
Yes Yes Yes Good Export/import
of DAC plus
data with DAC
framework
Service for
cloud-only
support
available
SQL DAC
Examples
available on
CodePlex
SSMS
Generate
Scripts Wizard
Yes Some Yes Poor Has explicit
option for SQL
Azure scripts
generation
Good for
smaller
database
bcp No N/A Yes Good Efficient
transfer of data
to existing table
Each bcp
command
transfer one
database
SQL Azure
Migration
Wizard
Yes Yes Yes Good Great
capabilities,
e.g. evaluate
trace files

388
Open source
on CodePlex
Not supported
by Microsoft
SQL Server
Integration
Services
No N/A Yes Good Most flexibility
SQL Server
Import and
Export Wizard
No N/A Yes Good Simple UI on
top of SSIS;
also available
in SSMS

Migrate from Other RDMSs

SQL Azure Migration Assistant can be used to migration database from Access, MySQL, Oracle,
Sybase to SQL Azure.
Microsoft Codename Data Transfer can transfer data in CSV or Excel file to SQL Azure.
Migrate between SQL Azure Databases

To migrate data from one SQL Azure database to another SQL Azure database, you can use
SQL Azure database copy and SQL Azure Data Sync.
SQL Azure supports a database copy feature. This creates a new database in SQL Azure which
is a transactionally consistent copy of an existing database. To copy database, you must be
connected to the master database of the SQL Azure server where the new database will be
created, and use the CREATE DATABASE command:
CREATE DATABASE destination_database_name AS COPY OF
[source_server_name.]source_database_name

The new database can be on the same server, or on a different server. The user executing this
statement must be in the dbmanager role on the destination server (to create a new database)
and must be dbowner in the source database. For more information see Copying Databases in
SQL Azure .
SQL Azure Data Sync enables creating and scheduling regular synchronizations between SQL
Azure and either SQL Server or other SQL Azure databases. For more information, see SQL
Azure Data Sync Overview.
Using Migration Tools


389
Data-tier Application DAC Package
Data-tier Applications (DAC) were introduced in SQL Server 2008 R2, with developer tool support
in Visual Studio 2010. They are useful for packaging the schema, code and configuration of a
database for deploying to another server. When a DAC is ready to deploy, it is built into a DAC
package (.bacpac), which is a compressed file that contains the DAC definitions in the XML
format. From SQL Server Management Studio, you can export database schema to a DAC
package, and then deploy the package to SQL Azure.
The DACPAC format is different from the BACPAC format. The BACPAC format extends
the DACPAC format to include a metadata file and JavaScript Object Notation (JSON)
encoded table data, in addition to the standard .dacpac file contents. The BACPAC
format is discussed in the DAC Import Export section.
You have the option to modify DAC package using Visual Studio 2010 before deployment. Within
the DAC project, you can specify pre-deployment and post-deployment scripts. These are
Transact-SQL scripts that can perform any action, including inserting data in the post-deployment
scripts. However it is not recommended to insert a large amount of data using a DAC package.
Installation and Usage
DAC is shipped with SQL Server 20008 R2. There are two main steps involved migrating SQL
Server database schema to SQL Azure:
1. Extract a DAC package from a SQL Server database:
The Extract Data-tier Application Wizard can be used to build a DAC package based on an
existing database. The DAC package contains the selected objects from the database, and
associated instance-level objects such as the logins that map to database users.
Here is a screenshot of opening the wizard:
Note


T
a
b
c.
A
P
su
su
he wizard inv
. Set the DA
package fi
. Validate th
. Build the p
A DAC can on
ack 4 (SP4) o
upported in a
upported in a
volves the follo
AC properties
le location.
hat all the data
package.
ly be extracte
or later. You c
DAC, or con
DAC, see DA
owing main s
, including DA
abase objects
ed from a data
cannot extrac
tained users.
AC Support F
steps:
AC applicatio
s are support
abase in SQL
ct a DAC if the
For more info
For SQL Serve
n name, vers
ed by a DAC
L Azure, or SQ
e database ha
ormation abo
er Objects an
ion, descriptio
.
QL Server 20
as objects tha
out the types o
nd Versions.

39
on, and the
05 Service
at are not
of objects
90


391
2. Deploy the DAC package to SQL Azure:
The Deploy Data-tier Application Wizard can be used to deploy a DAC package. You must
first connect to the SQL Azure server from SQL Server Management Studio. The wizard
creates the database if the database doesnt exist. The wizard deploys the DAC package to
the instance of the Database Engine associated with the node you selected in the Object
Explorer hierarchy. For example, in the following screenshot, it deploys the package to the
SQL Server called maqqarly23.database.windows.net:


It is a good practice to review the contents of a DAC package before deploying it in
production, especially when the deployed package was not developed in your
organization. For more information, see Validate a DAC Package.
The wizard involves the following main steps:
a. Select the DAC package.
b. Validate the content of the package.
c. Configure the database deployment properties, where you specify the SQL Azure
database.
d. Deploy the package.
Other than using the wizard, you can also use PowerShell with the dacstore.install() method to
migration schema to SQL Azure.
Resources
Understand Data-tier Applications.
Extract a DAC From a Database
Deploy a Data-tier Application
Important

392
Data-tier Application BACPAC Package
A data-tier application (DAC) is a self-contained unit for developing, deploying, and managing
data-tier objects. DAC enables data-tier developers and database administrators to package
Microsoft SQL Server objects, including database objects and instance objects, into a single
entity called a DAC package (.dacpac file). The BACPAC format extends the DACPAC format to
include a metadata file and JavaScript Object Notation (JSON)encoded table data in addition to
the standard .dacpac file contents. You can package your SQL Server database into a .bacpac
file, and use it to migrate the database to SQL Azure.
DACPAC & BACPAC are similar but actually target very different scenarios. DACPAC is
focused on capturing and deploying schema. Its primary use case is for deploying to
development, testing, and then production environment.
BACPAC is focused on capturing schema and data. It is the logical equivalent of a
database backup and cannot be used to upgrade existing databases. BACPACs
primary use cases are moving a database from one server to another (or to SQL Azure)
and archiving an existing database in an open format.
The Import and Export Service for SQL Azure is currently available as a public CTP. The service
can directly import or export BACPAC files between a SQL Azure database and Windows Azure
Blob storage. The Import and Export Service for SQL Azure provides some public REST
endpoints for the submission of requests.
The Windows Azure Platform Management Portal has an interface for calling the Import and
Export Service for SQL Azure.
Note

SQL S
file. Y
The S
datab
their s
D
D
S
S
O
D
T
fr
su
N
Server Manag
You can levera
SQL DAC Exa
bases from SQ
source code:
DAC Import a
DAC Import a
QL Azure to i
QL Azure dat
One way to co
Data Transfer
he ability to im
amework is c
upported by t
Note
gement Studi
age the DAC
amples shows
QL Server to S
and Export C
and Export Se
import and ex
tabase.
opy a bacpac f
r. For more in
mport and exp
currently only
he communit
o currently do
API to import
s how to use t
SQL Azure. T
lient-side too
ervice client
xport bacpac
file to Window
nformation, se
port data to S
available as C
y.
oes not suppo
t and export d
the Data-tier
The package
ols can be us
can be used
files between
ws Azure Blob
ee the Micros
SQL Azure us
CodePlex exa
ort exporting a
data.
Application F
provides two
sed to export
to call the Im
n Windows Az
b storage is to
soft Codenam
sing the Data-
amples. The t
a database in
Framework AP
command lin
and import ba
mport and Exp
zure Blob stor
o use Microso
me Data Trans
-tier Applicatio
tools are only

39

nto a BACPAC
PI to migrate
ne utilities and
acpac files.
port Service fo
rage and a
oft Codename
sfer section.
on (DAC)
y
93
C
d
or
e

394
Installation and Usage
This section demonstrates how to use the client tools of the SQL DAC Examples for migrating
database from SQL Server to SQL Azure.
The SQL DAC Examples can be downloaded from CodePlex. To run the sample, you must also
install Data-Tier Application Framework on your computer.
Before using the tools to migrate database, you must create the destination SQL Azure database
first. There are two steps involved using the tool for migration:
1. Export a SQL Server Database
Assume a database exists that is running on SQL Server 2008 R2, which a user has
integrated security access to. The database can be exported to a .bacpac file by calling the
sample EXE with the following arguments:
DacCli.exe -s serverName -d databaseName -f
C:\filePath\exportFileName.bacpac -x -e
2. Import the Package to SQL Azure
Once exported, the export file can be imported to a SQL Azure database with the following
arguments:
DacCli.exe -s serverName.database.windows.net -d databaseName -f
C:\filePath\exportFileName.bacpac -i -u userName -p password
Resources
How to Use Data-Tier Application Import and Export with SQL Azure
DAC Framework Client Side Tools Reference
Generate Scripts Wizard
The Generate Scripts Wizard can be used to create Transact-SQL scripts for SQL Server
database and/or related objects within the selected database. You can then use the scripts to
transfer schema and/or data to SQL Azure.
Installation and Usage
The Generate Scripts Wizard is installed with SQL Server 2008 R2. The wizard can be opened
from SQL Server Management Studio 2008 R2. The following screenshot shows how to open the
wizard:


The w
1. C
2. S
w
3. S
B
co
an
wizard involve
Choose object
et scripting o
window; or pub
et advanced
y default, the
onfiguration, y
nd then set th
es the followin
ts to export.
ptions. You h
blish it to a W
scripting optio
script is gene
you must clic
he Script for
ng main steps
have the optio
Web service.
ons.
erated for sta
k the Advanc
the database
s:
ons to save th
and-alone SQ
ced button fro
e engine typ
e script to file
L Server insta
om the Set Sc
pe property to
e, clipboard, n
ance. To cha
cripting Opti
SQL Azure.

39

new query
ange the
ions dialog,

95

396



You can also set the Types of data to script to one of the following based on your
requirements: Schema only, Data only, Schema and data.
After the script is created, you have the option to modify the script before running the script
against a SQL Azure database to transfer the database.
Resources
How to: Migrate a Database by Using the Generate Scripts Wizard (SQL Azure Database)
bcp
The bcp utility is a command line utility that is designed for high performing bulk upload to SQL
Server or SQL Azure. It is not a migration tool. It does not extract or create schema. You must
first transfer the schema to SQL Azure database using one of the schema migration tools.
You can use bcp to backup and restore your data on SQL Azure.
The SQL Azure Migration Wizard uses bcp.
Installation and Usage
The bcp utility is shipped with SQL Server. The version shipped with SQL Server 2008 R2 is fully
supported by SQL Azure.
There are two steps involved using bcp:
Note
Note

397
1. Export the data into a data file.
To export data out of SQL Server database, you can run the following statement at command
prompt:
bcp tableName out C:\filePath\exportFileName.dat S serverName
T n -q
The out parameter indicates copying data out of SQL Server. The -n parameter performs the
bulk-copy operation using the native database data types of the data. The -q parameter
executes the SET QUOTED_IDENTIFIERS ON statement in the connection between the bcp
utility and your SQL Server instance.
2. Import the data file into SQL Azure
To import data into your SQL Azure database, you must first create the schema in the
destination database, and then run the bcp utility on from a command prompt:
Bcp tableName in c:\filePath\exportFileName.dat n U
userName@serverName S tcp:serverName.database.windows.net P
password b batchSize
The b parameter specifies the number of rows per batch of imported data. Each batch is
imported and logged as a separate transaction that imports the whole batch before being
committed. Identify the best batch size and using the batch size is a good practice for
reducing connection lose to SQL Azure during data migration.
The following are some best practices when you use bcp to transfer a large amount of data.
1. Use the N option to transfer data in the native mode so that no data type conversion is
needed.
2. Use the b option to specify the batch size. Be default, all the rows in the data file are
imported as one batch. In case of a transaction failure, only insertions from the current batch
are rolled back.
3. Use the h TABLOCK, ORDER() option. The h TABLOCK specifies that a bulk update
table-level lock is required for the duration of the bulk load operation; otherwise, a row-level
lock is acquired. It can reduce lock contention on the table. The h ORDER() option
specify the sort order of the data in the data file. Bulk import performance is improved if the
data being imported is sorted according to the clustered index on the table.
You can use the F and L options to specify the first and last rows of a flat file for the upload.
This was useful to avoid having to physically split the data file to achieve multiple stream upload.
Resources
bcp utility
SQL Azure Migration Wizard
The SQL Azure Migration Wizard is an open source UI tool that helps migrating SQL Server
2005/2008 databases to SQL Azure. Other than migrating data, it can also be used to identify any
compatibility issues, fix them where possible and notify you of all issues it knows about.
The SQL Azure Migration Wizard has built-in logic for handling connection lose. It breaks down
the transactions into smaller bunches and runs until SQL Azure terminates the connection. When
the w
and p
to upl
logic t
has b
T
co
Instal
The S
Unzip
the ap
N
izard encount
picks up proce
oad the data
to figure out t
cp restart the
he SQL Azur
ommunity.
llation and U
SQL Azure Mi
p the package
pplication:
Note
ters the conn
essing after th
to SQL Azure
the last succe
e data upload
e Migration W
Usage
igration Wizar
e to your local
ection error, i
he last succes
e, the wizard
essful record u
with the next
Wizard is an o
rd can be dow
l computer, an
it reestablishe
ssful comman
chunks the d
uploaded befo
t set of record
open source to
wnloaded from
nd run SQLA
es a new con
nd. In the sam
ata into smal
ore the conne
ds.
ool built and s
m http://sqlaz
zureMW.exe.
nection with S
me manner, w
ler sections a
ection was clo
supported by
uremw.codep
. Here is a sc

39
SQL Azure
when using bc
and uses retry
osed. Then it
the
plex.com.
creenshot of

98
cp
y

399

The wizard involves the following steps:
1. Select the process you want the wizard to talk you through.
2. Select the source you want to script.
3. Select database objects to script.
4. Generate the script. You have the option to modify the script afterwards.
5. Enter information for connecting to target server. You have the option to create the
destination SQL Azure database.
6. Execute script against destination server.
Resources
Using the SQL Azure Migration Wizard (video)
SQL Server Integration Services
SQL Server Integration Services (SSIS) can be used to perform a broad range of data migration
tasks. It is a powerful tool when operating on multiple heterogeneous data sources and
destinations. This tool provides support for complex workflow and data transformation between
the source and destination. Even though SSIS is not currently supported by SQL Azure, you can
run it on an on-premise SQL Server 2008 R2 to transfer data to SQL Azure Database.
The SSIS Import/Export Wizard lets the user create packages that move data from a single data
source to a destination with no transformations. The Wizard can quickly move data from a variety
of source types to a variety of destination types, including text files and other SQL Server
instances.
Installation and Usage
You must use the SQL Server 2008 R2 version of SSIS to connect to SQL Azure.
The ADO.NET adapters have the required support for SQL Azure. It provides an option to bulk
load data specifically for SQL Azure. Use ADO.NET Destination adapter to transfer data to SQL
Azure. Connecting to SQL Azure Database by using OLEDB is not supported.
The following is a screenshot for configuring the ADO.NET Connection to SQL Azure:

Becau
the pa
the pa
When
possib
perfor
One w
In SS
then c
check
Reso
D
SQL
The S
Serve
conne
use your pack
ackages in a w
ackage.
n you configur
ble option. T
rmance.
way to improv
IS Designer,
connected to
ked.
urces
Designing and
Server Im
SQL Server Im
er Integration
ections, sourc
kage might fa
way so that it
re the ADO.N
hat allows yo
ve the perform
you can refer
an ADO .Net
Implementin
mport and E
mport and Exp
Services pac
ce, and destin
ail due to throt
t can be resum
NET destinatio
ou to use bulk
mance is to sp
rence the files
t Component
g Packages (
Export Wiz
port Wizard o
ckage for a sim
nation, and ad
ttling or netwo
med at the po
on, make sure
k load capabil
plit source da
s using the Fl
that has the U
(Integration S
ard
offers the simp
mple import o
dds any data t
ork issue, you
oint of failure,
e to use the U
ities to improv
ta into multip
lat File Comp
Use Bulk Inse
Services)
plest method
or export. This
transformatio
u might consid
as different t
Use Bulk Inse
ve the transfe
le files on the
ponent. Each
ert when poss
to create a S
s wizard confi
ons that are re

40

der designing
to restarting
ert when
er
e file system.
input file is
sible flag
SQL
igures the
equired to let
00
g

401
you run the import or export immediately. After a package is created, you have the option to
modify the package in SSIS Designer.
The wizard supports the following data sources:
.NET Framework Data Provider for ODBC
.NET Framework Data Provider for Oracle
.NET Framework Data Provider for SQL Server
Flat File Source
Microsoft OLE DB Provider for Analysis Services 10.0
Microsoft OLE DB Provider for Search
Microsoft OLE DB Provider for SQL Server
SQL Native Client
SQL Server Native Client 10.0
The SQL Server Import and Export Wizard can only transfer data. Before using the wizard, you
must transfer the schema using one of the schema migration tools, Generate Scripts Wizard or
DAC package.
On a 64-bit computer, Integration Services installs the 64-bit version of the SQL Server
Import and Export Wizard (DTSWizard.exe). However, some data sources, such as
Access or Excel, only have a 32-bit provider available. To work with these data sources,
you might have to install and run the 32-bit version of the wizard. To install the 32-bit
version of the wizard, select either Client Tools or Business Intelligence Development
Studio during setup.
Installation and Usage
In SQL Server 2008 R2 or later, the SQL Server Import and Export Wizard supports for SQL
Azure. There are several ways to start the wizard:
1. On the Start menu, point to All Programs, point to Microsoft SQL Server 2008, and then
click Import and Export Data.
2. In Business Intelligence Development Studio, right-click the SSIS Packages folder from
Solution Explorer, and then click SSIS Import and Export Wizard.
3. In Business Intelligence Development Studio, on the Project menu, click SSIS Import and
Export Wizard.
4. In SQL Server Management Studio, connect to the Database Engine server type, expand
Databases, right-click a database, point to Tasks, and then click Import Data or Export
data.
5. In a command prompt window, run DTSWizard.exe, located in C:\Program Files\Microsoft
SQL Server\100\DTS\Binn.
The migration involves the following main steps:
1. Choose a data source from which to copy data.
2. Choose a destination where to copy data to.
To export data to SQL Azure, you must choose the .NET Framework Data Provider for
SQLServer as the destination:
Note

402


3. Specify table copy or query.
4. Select source objects.
5. Save and run the package.
After the SQL Server Import and Export Wizard has created the package, you can optionally save
the package to run it again later, or to refine and enhance the package in SQL Server Business
Intelligence (BI) Development Studio.
If you save the package, you must add the package to an existing Integration Services
project before you can change the package or run the package in BI Development
Studio.
Note

403
Resources
How to: Run the SQL Server Import and Export Wizard
Using the SQL Server Import and Export Wizard to Move Data
Microsoft Codename Database Transfer
Microsoft Codename Data Transfer is a cloud service that lets you transfer data from your
computer to a SQL Azure database or Windows Azure Blob storage. You can upload any data
format to Windows Azure Blob storage, and data that is stored in comma-separated value (CSV)
or Microsoft Excel format (.xlsx) to a SQL Azure database. When you upload data to a SQL Azure
database, it is transformed into database tables.
Usage
The Data Transfer service is accessible at https://web.datatransfer.azure.com/. From the home
page, you have the options to import data, manage your datasets and your stores.
The data import to SQL Azure process involves the following steps:
Enter your SQL Azure credentials.
Choose a file to transfer.
Analyze the data file and then transfer the data.
Resources
Microsoft Codename Data Transfer Tutorial
SQL Server Migration Assistant
SQL Server Migration Assistant (SSMA) is a family of products to reduce the cost and risk of
migration from Oracle, Sybase, MySQL and Microsoft Access databases to SQL Azure or SQL
Server. SSMA automates all aspects of migration including migration assessment analysis,
schema and SQL statement conversion, data migration as well as migration testing.
Installation and Usage
SSMA is a Web download. To download the latest version, see the SQL Server Migration Tools
product page. As of this writing, the following are the most recent versions:
Microsoft SQL Server Migration Assistant for Access v5.1
Microsoft SQL Server Migration Assistant for MySQL v5.1
Microsoft SQL Server Migration Assistant for Oracle v5.1
Microsoft SQL Server Migration Assistant for Sybase v5.1
Microsoft SQL Server Migration Assistant 2008 for Sybase PowerBuilder Applications v1.0
SSMS is installed by using a Windows Installer-based wizard. SSMA is free, but you must
download a registration key. After you install and run the application, the application prompts you
to register and download the registration key.
The migration process of the SSMA for Access involves the following steps:
1. Create a new migration wizard. Make sure selecting SQL Azure in the Migrate To box.
2. Add Access databases.

404
3. Select the Access objects to migrate.
4. Connect to SQL Azure.
5. Link tables. If you want to use your existing Access applications with SQL Azure, you can link
your original Access tables to the migrated SQL Azure tables. Linking modifies your Access
database so that your queries, forms, reports, and data access pages use the data in the
SQL Azure database instead of the data in your Access database.
6. Convert selected objects.
7. Load converted objects into SQL Azure database.
8. Migrate data for selected Access objects.
Resources
SQL Server Migration Assistant
Migrating Microsoft Access Applications to SQL Azure (video)
SQL Server: Manage the Migration
See Also
Migrating Database to SQL Azure


405
Using the ReportViewer ASP.NET Control in
Windows Azure
This article was written using: Windows Azure SDK 1.6, Visual Studio 2010 ReportViewer
control for ASP.NET (Microsoft.ReportViewer.WebForms.dll version 10.0.40219.329)
The MSDN articles Getting Started Guide for Application Developers (SQL Azure Reporting) and
How to: Use ReportViewer in a Web Site Hosted in Windows Azure contain information on using
the ReportViewer control for ASP.NET in a Windows Azure application with SQL Azure
Reporting. This article provides key additional information for successful use of ReportViewer in
a Windows Azure website that uses more than one web role instance.
This article will outline two solutions that will allow you to deploy ReportViewer solutions in a
multi-instance Windows Azure website.
IReportServerCredentials causes ASP.NET Session State to be
used
As noted in the documentation for IReportServerCredentials, when the ReportViewer uses an
implementation of this interface, it will use session state to store the object and the ReportViewer
state.
When using a single web role instance, the ASP.NET ReportViewer control will work correctly
without explicitly setting sessionState in the web.config. It will work just by implementing
IReportServerCredentials and setting ReportViewer.ServerReport.ReportServerCredentials
to an instance of your IReportServerCredentials implementation. This will work because it will
automatically use the default in-process session state provider. With only a single web role
instance, requests will always hit the same instance, and the session state will always exist on
the instance. Of course the instance will be rebooted for Windows patches, and it could be
rebooted to move to another physical host machine in the Azure data center. Reboots of the
instance will cause the in-process session state to be lost.
Multiple web role instances hosting the ReportViewer
When you use multiple instances of the web role (for high availability and scale-out) for your
solution, and you are using the default in-process session state provider, you will intermittently
receive the following error when loading the web page that contains a ReportViewer control:
ASP.NET session has expired or could not be found
This error occurs because your initial request hit a single web role instance and had its session
state stored in-process on that web role instance. A subsequent request from the same browser
might hit one of the other web role instances, where the session state does not exist.
The most direct approach is to use the Azure Caching session state provider to provide session
state across all instances as described in Session State Provider for Winddows Azure Caching.
Unfortunately, the use of Azure Caching session state provider surfaces a bug in the
NetDataContractSerializer when using the ReportViewer control that comes with Visual Studio
2008 or Visual Studio 2010. The ASP.NET Session State Provider for Windows Azure Caching
uses the NetDataContractSerializer to serialize objects into session state.

406
The bug causes an exception to be raised by the web page that contains a ReportViewer control
(along with a stack trace):
Server Error in '/' Application.

Type 'Microsoft.Reporting.WebForms.SyncList`1[[System.String, mscorlib, Version=4.0.0.0,
Culture=neutral, PublicKeyToken=b77a5c561934e089]]' is an invalid collection type since
it does not have a default constructor.

Description: An unhandled exception occurred during the execution of the current web
request. Please review the stack trace for more information about the error and where it
originated in the code.

Exception Details: System.Runtime.Serialization.InvalidDataContractException: Type
'Microsoft.Reporting.WebForms.SyncList`1[[System.String, mscorlib, Version=4.0.0.0,
Culture=neutral, PublicKeyToken=b77a5c561934e089]]' is an invalid collection type since
it does not have a default constructor.

The ReportViewer control that will ship with the next version of Visual Studio after Visual
Studio 2010 should have this issue fixed.
Solutions
There are two possible solutions for successful use of ReportViewer:
1. Make the ReportViewer "session-less" to prevent it from using ASP.NET session state. This
is the preferred method.
2. Install hotfix 2624480 (http://support.microsoft.com/kb/2624480) as a startup task in the role
instance. This hotfix fixes an issue with the NetDataContractSerializer. This approach will
cause all role instance reboots to be slower because the startup task to install the hotfix must
complete before the instance is available.
"Session-less" ReportViewer
You can find a discussion of making ReportViewer session-less in the following articles:
Custom Credentials in the Report Viewer
Specifying Connections and Credentials for the ReportViewer Web Server Control
Instead of implementing IReportServerCredentials and assigning an instance to
ReportViewer.ServerReport.ReportServerCredentials, you can implement
IReportServerConnectionor IReportServerConnection2 and point to this class in your web.config
file.
The following code is an implementation of IReportServerConnection (with error handling
removed for brevity). This implementation simply retrieves the Report Server user name,
password, and server name from the web.config AppSettings:

Note

407
Note that this code is based on the Windows Azure ReportViewer Control Sample linked
to from the SQL Azure Reporting Samples TechNet Wiki page.
namespace ReportViewerRemoteMode
{
public sealed class ReportServerConnection : IReportServerConnection
{
public bool GetFormsCredentials(out Cookie authCookie, out string user,
out string password, out string authority)
{
authCookie = null;
user = ConfigurationManager.AppSettings["RSUSERNAME"];
password = ConfigurationManager.AppSettings["RSPASSWORD"];
authority = ConfigurationManager.AppSettings["RSSERVER_NAME"];
return true;
}

public WindowsIdentity ImpersonationUser
{
get { return null; }
}

public ICredentials NetworkCredentials
{
get { return null; }
}

public Uri ReportServerUrl
{
get
{
return new Uri(String.Format("https://{0}/reportserver",
ConfigurationManager.AppSettings["RSSERVER_NAME"]));
}
}

public int Timeout
{
get { return 60000; }
}
}
Note

408
}
Assuming that your assembly name is ReportViewerRemoteMode, the following entry in the
<appSettings> section in the web.config will cause the ReportViewer control to use this
IReportServerConnection implementation:
<appSettings>
<!-- Custom Report Server Connection implementation-->
<add key="ReportViewerServerConnection"
value="ReportViewerRemoteMode.ReportServerConnection, ReportViewerRemoteMode"/>
</appSettings>
Do not set the ReportViewer.ServerReport.ReportServerCredentials or the
ReportViewer.ServerReport.ReportServerUrl in your web page code-behind class.
This solution will not work for all scenarios because the web.config can only point to one
IReportServerConnection implementation.
Hotfix 2624480 installed as startup task
Another solution is to install hotfix 2624480, http://support.microsoft.com/kb/2624480, as a startup
task in the role instance. This hotfix fixes an issue with the NetDataContractSerializer that will
allow the ReportViewer control to successfully serialize to session state without throwing an
exception.
In order to implement this solution, you will need the "x64" version of the hotfix executable
downloaded from http://support.microsoft.com/kb/2624480. Add the file "NDP40-KB2624480-
x64.exe" to your Visual Studio project, and set the file properties to copy to the output directory.
Then in the ServiceDefinition.csdef file, you need to add a <Startup>element inside the <WebRole>
element as follows. This causes the hotfix to silently install without user prompting or intervention.

Note the "/q" parameter for the .EXE
<Startup>
<Task commandLine="NDP40-KB2624480-x64.exe /q"
executionContext="elevated" />
</Startup>

Now when the role instance starts, it will install the hotfix before the instance starts accepting web
requests.
One drawback of this approach is that it will cause all role instance reboots and deployments to
be slower because the startup task to install the hotfix must complete before the instance is
available. This can cause deploy times and startup times to be significantly slower. A test of this
approach with deploying two "small" web role instances to Windows Azure from within Visual
Studio showed that deployments took twice as long when using this startup task to install the
hotfix.
Note

409
If possible, make the ReportViewer "session-less". If this is not possible, then install the hotfix as
a startup task as a fallback method.
Additional Resources
Custom Credentials in the Report Viewer
Specifying Connections and Credentials for the ReportViewer Web Server Control
Getting Started with Defining Startup Tasks for Roles in Windows Azure
Introduction to Windows Azure Startup Tasks
Author: Suren Machiraju
Reviewers: Jaime Alva Bravo and Steve Wilkins
After designing, coding and deploying your Azure AppFabric solution you may find that it does not
work. This article provides guidance on how to efficiently test your Azure AppFabric applications
through the software development life cycle both at the business logic and the comprehensive
end-to-end scenario tests. In this article, we demonstrate how to:
Develop unit tests for the business logic components while eliminating dependencies on
Azure AppFabric components.
Design integration end-to-end tests and eliminate overhead for setup, initialize, cleanup and
teardown of the Azure AppFabric Service resources (for example, Queues) for every test run.
Further eliminate creating duplicate infrastructures for ACS namespaces, Service Bus
Queues, and other resources.
Additionally, this article provides an overview of the various testing technologies and techniques
for testing your Azure AppFabric applications.
The Types of Tests
In this article, we will focus on the following tests:
Unit Tests are narrowly focused tests designed to exercise one specific bit of functionality.
These tests are often referred to as the Code Under Test or CUT. Any dependencies taken
by the CUT are somehow removed.
Integration Tests are broader tests for exercising multiple bits of functionality at the same
time. In many cases, these resemble unit tests that cover multiple feature areas and include
multiple dependencies.
Overall our tests focus on creating and using Test Doubles. We will use the following types of
Test Doubles:
Mocks are simulated objects that imitate the behavior of real objects in controlled ways. They
are replacements of what is dynamically produced by a framework.
Fakes are simulated objects that implement the same interface as the object that they
represent. Fakes return pre-defined responses. Fakes contain a set of method stubs and
serve as replacements for what you build by hand.
Stubs simulate the behavior of software objects.
Our tests, upon executing can verify both state (for example, after you make a call a specific
value is returned) and behavior (the method is called in a certain order or a certain number of
times).

410
The Azure Dependencies
One of the major goals of unit testing is to eliminate dependencies. For the Azure framework,
these dependencies include the following:
Service Bus Queues
Access Control Service
Cache
Windows Azure Tables, Blobs, Queues
SQL Azure
Cloud Drive
Other Web Services
When building tests for Azure Applications we will replace these dependencies to better focus our
tests on exercising the desired logic.
In this article, we will provide examples of Service Bus Queues with full knowledge that the tools
and techniques demonstrated here will also apply to all other dependencies.
The Testing Framework
To implement the testing framework for your Azure AppFabric applications, you will need:
A unit a testing framework to define and run your tests.
A mocking framework to help you isolate dependencies and build narrowly scoped unit tests.
Tools to help with automatic unit test generation for increased code coverage.
Other frameworks that can help you with testable designs by taking advantage of
dependency injection and applying the Inversion of Control Pattern.
Unit Testing with MS Test
Visual Studio includes a command line utility MS Test that you can use to execute unit tests
created in Visual Studio. Visual Studio also includes a suite of project and item templates to
support testing. You typically create a new test project and then add classes (known as test
fixtures) adorned with [TestClass] attributes that contain methods adorned with [TestMethod].
Within MS Test, various windows within Visual Studio enable you to execute unit tests defined in
the project, and review the results after you execute them.
Visual Studio Express and Test Professional editions do not contain MS Test.
In MS Test, unit tests follow an AAA pattern - Arrange, Act, Assert.
Arrange - build up any pre-requisite objects, configurations, and all other necessary
preconditions and inputs needed by the CUT.
Act - perform the actual, narrowly scoped test on the code.
Assert - verify that the expected results have occurred.
The MS Test framework libraries include PrivateObject and PrivateType helper classes that use
reflection to make it easy to invoke non-public instance members or static members, from within
the unit test code.
Note

411
Premium and ultimate editions of Visual Studio include enhanced unit test tooling that integrate
with MS Test that enable you to analyze the amount of code exercised by your unit tests, and
visualize the result by color coding the source code, in a feature referred to as Code Coverage.
Mocking with Moles
A goal of unit testing is to be able to test in isolation. However, often the code under test cannot
be tested in isolation. Sometimes the code is not written for testability. To rewrite the code would
be difficult because it relies on other libraries that are not easily isolated, such as those that
interact with external environments. A mocking framework helps you isolate both types of
dependencies.
A list of mocking frameworks you might consider is available in the links section at the end of this
article. For our purposes, we will focus on how to use Moles, a mocking framework from Microsoft
Research.
Most mocking frameworks will create mock types dynamically by deriving from the type you
indicate for the members you want to change. Fewer frameworks have support for handling
sealed classes, non-virtual methods, static members, delegates or events that require alternative
techniques (such as using the .NET Profiling API). Moles framework provides this advanced
support.
Most of the mocking frameworks do not enable you to rewrite the logic within a constructor. If you
are creating an instance of a type that you have to isolate, and the constructor of this type takes
dependencies you want to isolate, you will have to creatively mock the calls made by the
constructor instead of the constructor itself. Moles can be used to insert logic before constructor
calls, but it stops short of letting you completely rewrite the constructors logic.
Moles provides all these features at no cost to you. In addition, it integrates well with Visual
Studio 2010 and MS Test, and tests generated with it are fully debuggable within Visual Studio.
Automated Test Creation with PEX
PEX (Program EXploration) is a white-box test generation tool available to MSDN subscribers
from Microsoft Research. You can use PEX to automatically create unit tests with high code
coverage by intelligently selecting input parameters for CUT, and recursively for parameterized
unit tests. PEX is capable of building the smallest number of tests that produce the greatest code
coverage.
PEX installs and runs fully integrated within Visual Studio 2010, and installs Moles as part of the
installation process.
Frameworks for Dependency Injection and Inversion of Control (IoC)
Containers
You can use Microsoft Unity for extensible dependency injection and IoC. It supports interception,
constructor injection, property injection, and method call injection.
Microsoft Unity and similar tools help you build testable designs that let you insert your
dependencies at all levels of your application, assuming that your application was designed and
built with dependency injection and one of these frameworks in mind.
These
some
this a
Test
In this
mess
intere
Unit
Let us
by us

When
contro
[HttpP
public
{
tr
{
e frameworks
what heavy in
rticle.
ting Azur
s section, we
ages to a Qu
ested in testin
Testing Co
s say you hav
ing an AppFa
n the user clic
oller. The acti
Post]
c ActionResu
ry
s are great for
n their up-fron
re Solutio
will describe
eue and a wo
g aspects of a
ontroller A
ve a website t
abric Queue. O
cks Create, it
ion is implem
lt Create(Or
r writing testa
nt design requ
ons
a solution tha
orkers role th
all three.
Actions in a
that creates o
Our example
posts the new
ented as follo
rder order)
ble code, and
uirements. W
at includes a w
at processes
a Web Role
orders, and th
web page loo
w order to the
ows:
d ultimately go
We will not cov
web role host
messages fr
e
ese orders ar
oks like this:
e Create actio
ood code, but
ver DI or IoC c
ted web site t
rom the Queu
re queued for
on on the asso

41
t can be
containers in
that pushes
e. We are
r processing

ociate
12

413
string qName = RoleEnvironment.GetConfigurationSettingValue("QueueName");
string qNamespace =
RoleEnvironment.GetConfigurationSettingValue("QueueServiceNamespace");
string qUsername = RoleEnvironment.GetConfigurationSettingValue("QueueUsername");
string qPassword = RoleEnvironment.GetConfigurationSettingValue("QueuePassword");

//Push order to AppFabric Queue
AppFabricQueue queue = new AppFabricQueue(new QueueSettings(qNamespace, qName,
Username, qPassword));

queue.Send(order);

queue.CloseConnections();

return View("OrderCreated");
}
catch (Exception ex)
{
Trace.TraceError(ex.Message);
return View("Error");
}
}
Notice that the code retrieves settings from configuration through RoleEnvironment and sends a
message that contains the order to the queue. The AppFabricQueue class we use here is a
wrapper class that uses the Service Bus .NET APIs (MessageSender, MessageReceiver, etc.) to
interact with Service Bus Queues.
Let us build a unit test for this method using Moles. Notice that the Create action uses the
following methods:
GetConfigurationSettingsValue (RoleEnvironment class)
Send (AppFabricQueue class)
What we want to do is build detours for these methods using Moles to control their behavior and
remove the dependencies on the real environment. Controlling behavior of these methods and
removing their dependencies on the real environment will eliminate the need to run the tests in
the emulator, on Azure, or to call out to the Service Bus Queue. We will exercise the Create
action on the controller to verify that the Order input is the one sent to the queue (checking that it
has the Order Id and Description as input to the action) and that the OrderCreated view is
displayed as a result.
Accomplishing this with moles is easy. Within the Test project, right click the assembly containing
the types you want to mock and select Add Moles Assembly.
In our example, we select Microsoft.Windows.Azure.ServiceRuntime and select Add Moles
Assembly. An XML file named Microsoft.WindowsAzure.ServiceRuntime.moles will be added to
the test project. Then repeat for the AF.Queues assembly.

414
When you build the Test project, references will be added to the auto-generated moled versions
of these assemblies.
In our example these are Microsoft.Windows.Azure.ServiceRuntime.Moles and
AF.Queues.Moles.
Create a new unit test method and decorate it with [HostType("Moles")]. Within the Unit Test, you
use the "moled" types by accessing them through a Moles namespace. All moled types start with
an M.
For example, the moled type for AppFabricQueue is accessed through
AF.Queues.Moles.MAppFabricQueue.
You define your detours (the alternate logic to execute) by assigning a delegate to the method
you want to replace.
For example, to detour RoleEnvironment.GetConfigurationSetting we define a delegate for
MRoleEnvironment.GetConfigurationSettingValueString.
The complete unit test is as follows. Observe the detours we provide for
RoleEnvironment.GetConfigurationSetting, the QueueSettings constructor, and the Send and
CloseConnection methods of the AppFabricQueue class.
[TestMethod]
[HostType("Moles")]
public void Test_Home_CreateOrder()
{
//Arrange

Microsoft.WindowsAzure.ServiceRuntime.Moles.MRoleEnvironment.GetConfigurationSettingValue
String = (key) =>
{
return "mockedSettingValue";
};

bool wasQueueConstructorCalled = false;
AF.Queues.Moles.MAppFabricQueue.ConstructorQueueSettings = (queue, settings) =>
{
wasQueueConstructorCalled = true;
};

Order orderSent = null;
AF.Queues.Moles.MAppFabricQueue.AllInstances.SendObject = (queue, order) =>
{
orderSent = order as Order;
};

bool wasConnectionClosedCalled = false;
AF.Queues.Moles.MAppFabricQueue.AllInstances.CloseConnections = (queue) =>

415
{
wasConnectionClosedCalled = true;
};

Order inputOrder = new Order()
{
OrderId = "Test123",
Description = "A mock order"
};

HomeController controller = new HomeController();

//Act
ViewResult result = controller.Create(inputOrder) as ViewResult;

//Assert
Assert.IsTrue(wasConnectionClosedCalled);
Assert.IsTrue(wasQueueConstructorCalled);
Assert.AreEqual("OrderCreated", result.ViewName);
Assert.IsNotNull(orderSent);
Assert.AreEqual(inputOrder.OrderId, orderSent.OrderId);
Assert.AreEqual(inputOrder.Description, orderSent.Description);

}
Unit Testing Worker Roles that use AppFabric Queues
The web role took care of adding an order to the queue. Let us consider a workers role that
would process the orders by retrieving them from the queue and see how we can test that. The
core item to test in the worker role is defined in the implementation of RoleEntryPoint, typically in
the OnStart and Run methods.
The Run method in our worker role periodically polls the queue for orders and processes them.
public override void Run()
{
AppFabricQueue queue = null;

try
{
string qName = RoleEnvironment.GetConfigurationSettingValue("QueueName");
string qNamespace =
RoleEnvironment.GetConfigurationSettingValue("QueueServiceNamespace");
string qUsername =
RoleEnvironment.GetConfigurationSettingValue("QueueUsername");

416
string qPassword =
RoleEnvironment.GetConfigurationSettingValue("QueuePassword");

queue = new AppFabricQueue(new QueueSettings(qNamespace, qName,
qUsername, qPassword));
queue.CreateQueueIfNotExists();

while (true)
{
Thread.Sleep(2000);

//Retrieve order from AppFabric Queue
TryProcessOrder(queue);
}
}
catch (Exception ex)
{
if(queue !=null)
queue.CloseConnections();

System.Diagnostics.Trace.TraceError(ex.Message);
}

}
We want a test verifying that the routine correctly retrieves a message.
In this case we add a moled assembly for the worker role. For our specific project, this is the
Mvc3Web assembly because we host the RoleEntryPoint there. The following is the complete
unit test for run method of the worker role.
[TestMethod]
[HostType("Moles")]
public void Test_WorkerRole_Run()
{
//Arrange

Microsoft.WindowsAzure.ServiceRuntime.Moles.MRoleEnvironment.GetConfigurationSettingValue
String = (key) =>
{
return "mockedSettingValue";
};

bool wasQueueConstructorCalled = false;
AF.Queues.Moles.MAppFabricQueue.ConstructorQueueSettings = (queue, settings) =>

417
{
wasQueueConstructorCalled = true;
};

bool wasEnsureQueueExistsCalled = false;
int numCallsToEnsureQueueExists = 0;
AF.Queues.Moles.MAppFabricQueue.AllInstances.CreateQueueIfNotExists = (queue) =>
{
wasEnsureQueueExistsCalled = true;
numCallsToEnsureQueueExists++;
};

bool wasTryProcessOrderCalled = false;
int numCallsToTryProcessOrder = 0;
Mvc3Web.Worker.Moles.MWorkerRole.AllInstances.TryProcessOrderAppFabricQueue =
(actualWorkerRole, queue) =>
{
wasTryProcessOrderCalled = true;
numCallsToTryProcessOrder++;

if (numCallsToTryProcessOrder > 3) throw new Exception("Aborting Run");
};

bool wasConnectionClosedCalled = false;
AF.Queues.Moles.MAppFabricQueue.AllInstances.CloseConnections = (queue) =>
{
wasConnectionClosedCalled = true;
};

Mvc3Web.Worker.WorkerRole workerRole = new Worker.WorkerRole();

//Act
workerRole.Run();

//Assert
Assert.IsTrue(wasConnectionClosedCalled);
Assert.IsTrue(wasQueueConstructorCalled);
Assert.IsTrue(wasEnsureQueueExistsCalled);
Assert.IsTrue(wasTryProcessOrderCalled);
Assert.AreEqual(1, numCallsToEnsureQueueExists);
Assert.IsTrue(numCallsToTryProcessOrder > 0);
}
This t
the de
instan
define
In our
instan
detou
throw
time.
You m
classe
comp
fakes
create
appro
Queu
requir
actua
the de
then t
Servic
Ena
To an
both o
from t

By clic
Test R
Howe
that n
before
Code
Soluti
If
N
N
time we focus
elegate of the
nce you creat
ed delegates.
r example, we
nce methods
urs we have a
w an exception
may be asking
es from the S
letely isolate
that inherit fr
e mock types
oaches, you w
eClient, burd
red overrides,
lly rely on, an
ependency on
that is all you
ce Bus.
abling Co
nalyze what th
of the unit tes
the Test Resu
cking the Sho
Results windo
ever, if we do
no coverage d
e running our
coverage da
ion Items. Op
you have a s
New Item to th
Note
s on detouring
e AllInstances
e of the actua

e want to use
CreateQueue
lready seen f
n so that the i
g yourself, wh
Service Bus S
our code from
rom the abstra
for them all.
will ultimately f
ened with cre
, struggling to
nd recreating
n the real Ser
need to moc
de Cover
hese unit tests
sts with MS Te
ults dialog tha
ow Code Cov
ow), we can s
this now, we
data could be
r tests.
ta collection i
pening this file
solution that d
he solution an
g the instance
s property of t
al type will de
the original in
eIfNotExists a
for RoleEnviro
nfinite loop m
hy not just use
DK directly in
m calling the r
act classes in
The problem
find yourself r
eating derived
o expose the i
their construc
rvice Bus. On
k and detour
rage
s actually test
est, we can s
at appears.
verage Result
see the code c
will not see a
found. We fir
is enabled sim
e brings up th
does not have
d then select
e methods of
he generated
tour through
nstances Ru
and TryProces
onement.GetC
maintained wit
e the Messag
nstead of injec
real-world Se
n the Microsof
with either ap
reflecting into
d types from t
internal types
ctors or factor
the other han
in order to iso
ted, we can e
ee that they p
s button (the
covered by al
any data. Inste
rst have to en
mply by config
e Test Setting
e Local.testse
Test >Test S
the WorkerRo
d mole type. B
any of the me
n method, bu
ssOrder (in ad
Configuration
thin Run is ex
geSender/Mes
cting a helper
rvice Bus, we
ft.ServiceBus
pproach is co
o classes like
hese abstrac
s that real ver
ry methods in
nd, if you inse
olate yourself
examine code
pass and also
rightmost but
ll unit tests ex
ead, we will g
nable collectio
guring the Loc
gs editor.
ettings file, you
Settings.
ole class. You
By using the d
ethods for wh
ut provide deto
ddition to the
nSettings). In
xited at a pred
ssageReceive
r type? In orde
e have two ch
s namespace,
omplexity. Wit
TokenProvid
t types that e
rsions of these
n clever ways
ert your own h
f from the rea
e coverage da
o see related
tton on near t
xecuted in the
get a message
on of code cov
cal.testsetting
u can add it u

41
u should set
delegate, any
ich you have
ours for the
static method
our code we
determined
er and related
er to
hoices - write
, or let moles
th both
er and
xpose all thei
e classes
just to excise
helper type,
al-world
ata. If we run
run details

the top of the
e run.
e indicating
verage data
gs under the
using Add
18
d
d
ir
e

Click
Next,
Code

the Data and
click the Con
Coverage De
Diagnostics
nfigure button
etail window s
tab, and the c
(the one just
select all your
check the box
t above the gr
r assemblies
x to the right o
rid containing
under test an
of the Code C
g Code Cover
nd click OK.

41

Coverage row
rage). In the

19
w.
Then
code

Click
doubl
follow

Usin
Manu
unit te
install
editor
click Apply a
coverage but
the Show Co
e-click it to se
wing figure), p
ng PEX to
ually creating t
ests to the ne
ling PEX, you
r, and selectin
nd Close to d
tton. Now you
de Coverage
ee the source
artially tested
o Enhanc
these unit tes
xt level by try
u can have it e
ng Run PEX.
dismiss the Te
ur Code Cove
Coloring icon
e code colored
d or untested
ce Code C
sts is very val
ying paramete
explore a met
est Settings e
erage Results
n, and then d
d to reflect ar
(gray) the uni
Coverage
uable, but PE
er values you
thod simply b
editor. Re-run
should look a
rill down to a
eas that were
it tests just ru
e
EX can help y
may not have
by right clickin
your tests an
as follows:
method in the
e tested (gree
un.
you intelligent
e considered
ng the method

42
nd press the

e grid and
en in the

ly take your
. After
d in the code
20

After a
here i
PEX w
crafte
impor
gener

Try
The s
Use

Moc
Moles
Moq
NMoc
Rhino
EasyM
NSub
FakeI
TypeM
JustM
a short while,
is what we go
was able to c
ed that genera
rtant, in the de
rated tests ca
it out!
sample used i
eful Links
king Fram
s
ck3
o Mocks
Mock.NET
bstitute
tEasy
Mock Isolator
Mock
, PEX will finis
ot when ran P
reate one tes
ated that exce
etails pane yo
an be added to
n this article i
s
eworks
r
sh with its pro
PEX on our wo
st that resulted
eption (a null v
ou can see th
o your projec
is available at
ocessing and
orker role's Tr
d in an excep
value for the
e code PEX w
t simply by se
t sample_link
the results w
ryProcessOrd
ption, and it sh
queue param
was executing
electing them
k.
will stabilize. F
der method. N
hows you the
meter). What is
g. The results
and clicking

42

For example,
Notice that
inputs it
s even more
s of all PEX
Promote.

21

422
Unit Test Related
MSDN Verifying Code by Using Unit Tests
MSDN - Unit Testing in MVC Applications
Building Testable Applications
Pex and Moles
PrivateObject and PrivateType Introduced
CLR Profiling API Blog
Mocks Aren't Stubs
Dependency Injection and IoC
Microsoft Unity
Structure Map
MSDN Magazine - Aspect Oriented Programming, Interception and Unity 2.0
MSDN Magazine - Interceptors in Unity
MSDN Magazine - MEF v.s. IoC