PaaS with Java

PaaS Cloud with Java
Eberhard Wolff, Principal Technologist, SpringSource – A division of VMware
ewolff@vmware.com
Twitter: @ewolff
Blog: http://ewolff.com

© 2009 VMware Inc. All rights reserved

Agenda
§  A Few Words About Cloud
§  PaaS – Platform as a Service
§  Google App Engine
§  Vmforce
§  Amazon Beanstalk

2

A Few Words About Cloud

3

Three types of Clouds

Infrastructure Platform Software
as a Service as a Service as a Service
•  Virtual Servers •  Virtual Application •  Software or Service
•  Virtual Storage Server that you use

•  Similar to •  Handles Scale-Out •  Components that you
Virtualization •  Everything is add/integrate into
your app
•  Custom Solutions Services and APIs

•  Manage Everything •  Usually Larger Buy-In
Yourself •  Mostly Managed by
Provider

4

Cloud might be...

•  Private /internal “on premise”
–  In your data center
–  Useful in particular for larger organizations
•  Public “off premise”
–  Hosted and Operated by a third Party
–  Ideal for Startups etc

•  Even smaller enterprises can create private clouds
•  Based on virtualization

5

Cloud at the Core is a Business Model

§  Customer pays only for what he uses
•  Per CPU hour or cycles, per GB of storage, per MB of network bandwidth
•  Per user and year
•  Not for having stand-by capacity

§  Flexibility: More capacity is available on demand
•  Instantly available
•  Customer uses GUI or API to expand capacity
•  Cloud provider is responsible for having capacity ready

§  Makes IT just another service

6

Why Cloud?

§  Compelling Economics
•  Lower CapEx
•  Cheaper handling of peak loads
•  Better resource utilization
•  Simple to achieve benefits – direct ROI

§  Flexibility and better productivity for
development
•  Much easier to set up environments
•  Feasible to have production-like environments
quickly and cheaply available
•  Indirect: Better productivity leads to cost
saving / better time to market

7

OK, so let me get started

8

So, let me get started

§  Get an account at an IaaS provider
§  …or virtualize your data center and create a self service portal

§  Install your (Java EE) environment
§  Install your (Java) application
§  Done

§  Wow, that was easy!

9

That is not enough

§  How do you deal with peaks? Need more app server instances
§  The server instances must be shut down after the peak
§  …otherwise you would pay for them
§  Traditional middleware does not allow for that
§  Elastic scaling

§  RBMS prefer scale up (larger server)
§  In the cloud it is easier to scale out (more server)
§  That is why Amazon and Google use NoSQL / key-value stores

10

That is not enough: Handling Lot of Date

§  Traditional approach: Batch on a few nodes
§  Alternative approach: Map / Reduce on
many nodes
§  Map each value using some function
•  Runs on many nodes in parallel
•  E.g. given a text file emit a word count each time a
word is found
§  Reduce takes the result of the map step and
reduces it
•  E.g. add up all word counts
§  Distributed algorithm running on potentially
many nodes
§  Cloud: Can easily use a lot of nodes to run
Map Reduce

11

That is not enough: More Than Just a Virtualized Computer

§  Additional services
•  Amazon Simple Queue Service (SQS)
•  Amazon Simple Notification Service (SNS)
•  Amazon Simple Storage Service (S3)
•  Amazon Elastic Block Store (EBS)

12

Cloud Influences the Programming Model

§  Some Jave EE technologies are not a good fit
•  Non-JMS messaging technologies
•  AMQP / RabbitMQ
•  Amazon Simple Queue Service (SQS)
•  Amazon Simple Notification Service (SNS)
•  Two Phase Commit / JTA leads to long locking / strong coupling
•  What do you do about Map / Reduce?
•  What do you do about NoSQL?

§  You might be better off with a different programming model
•  Spring can handle non-Java-EE environments (e.g. Tomcat)
•  Projects for NoSQL, AMQP …

13

More Flexibility Needed

§  And remember: You pay the resources you consume
§  You need to use more or less resources based on load

§  You need to be able to shut down servers if load is low
§  You need to be able to start new servers if load is high

14

What you will eventually come up with
§  A tool to take an Application
§  …and create a VM with all needed
infrastructure etc

§  Need tools to
•  Install software
•  Manage infrastructure
•  Configure infrastructure
•  Set up user etc
•  Puppet, Chef etc.

§  Like a factory for VMs

15

So…

§  Can we automate this?
§  Developers just want a platform to run applications on

§  Also: Developers need other non-Java-EE services

16

Platform as a service (PaaS) the delivery of a
computing platform and solution stack as a
service.

19

Not just automated…

20

PaaS: Advantages and Disadvantages

§  Advantages
•  More useful than IaaS: You would need to install a server anyway
•  Automatic scaling
•  Resources automatically added
•  Can offer additional service
•  Tuned for Cloud
•  Technical e.g. data store, messaging, GUI elements
•  Domain e.g. predefined data model
§  Disadvantages
•  Less flexible
•  Pre-defined programming model
•  Defines environment
•  Programming model might be different
•  Hard to port existing code
•  Need to learn

22

Google App Engine

23

Google App Engine

§  Infrastructure offered by Google
§  Supports Java and Python

§  Infrastructure completely hidden

§  GAE sandbox more restrictive than normal JVM sandbox
§  So GAE can handle your application better
§  Java classes white list
•  i.e. some Java classes must not be used
•  no Thread
•  no file system
•  parts of System class (e.g. gc(), exit()…)
•  Reflection and ClassLoader work

24

Google App Engine: Storage

§  Relational database only in App Engine for Business
§  Based on BigTable
•  Google's storage technology
•  Key/value i.e. objects stored under some key
•  No joins
•  Simple / simplistic (?)
•  Scalable
•  Example of NoSQL
•  Principles will be discussed further in NoSQL

§  Max. 1 MB per entity (and other limitations)

25

Google App Engine: Storage APIs
§  API: Low level com.google.appengine.api.datastore
•  Not compatible to JDBC or the like
•  Queries, transactions, entities etc.
•  Should only be used by framework
§  API: JDO (Java Data Objects)
•  Standard for O/R Mapper and OODBMS
•  Unsupported: Joins, JDOQL grouping and aggregates, polymorphic queries
§  API: JPA (Java Persistence API)
•  Well established standard for O/R Mappers
•  Unsupported: Many-to-many relationships; join, aggregate and polymorphic
queries, some inheritance mappings
§  Problem: JPA / JDO based on RDBMS, but this is key/value
§  So maybe use the low level API instead?
§  JPA and JDO actually implemented by Data Nucleus library
•  Byte code must be enhanced in an additional compile step
26

Google App Engine: Additional services

§  Memcache
•  Implements JCache (JSR 107)
•  Fast access to data in memory
§  URL Fetch based on java.net.URL to read data via HTTP
§  EMail support using JavaMail
§  XMPP instant messages (proprietary API)
§  Image Manipulation (proprietary API)
§  Authentication / Authorization based on Google accounts
(proprietary API)
§  Task queuing (proprietary API, experimental)
§  Blob store (proprietary API, experimental) for data >1MB
§  Numerous other services available (not tied to App Engine)
§  Google Predict, Google BigQuery …

27

Google App Engine for Business

§  Centralized administration.
§  99.9% uptime SLA
§  Premium developer support available.
§  Sign on for users from your Google Apps domain
§  Announced
•  SSL on your company’s domain for secure communications
•  Access to advanced Google services.

28

Google App Engine

§  Documentation + SDK + Eclipse Plug In available
§  SDK contains environment for local tests

§  Spring / Spring Roo is the preferred programming model
§  …as it is popular and works with the limited model of the Google
App Engine

§  http://code.google.com/appengine/

29

VMforce

§  Joined offerings by Salesforce and VMware
§  Salesforce
•  Leading provider for SaaS / CRM
§  Salesforce offers force.com environment
•  Scalable PaaS
•  Database + data model
•  Charting
•  Reporting
•  Chat etc
§  VMforce = force.com + VMware virtualization + SpringSource
programming model

32

vmforce Vision

§  VMforce = force.com
§  + VMware virtualization
•  Offers IaaS foundation
§  + SpringSource tc Server / Hyperic
§  + Spring programming model

force.com
§  i.e. run your standard Platform
Spring Apps SpringSource
Services tc Server
§  Use force.com data (charts,
model / database reports)
VMware
§  …and charts / reports Force.com vSphere
Database

§  Apply for the beta at http://www.vmforce.com/

33

Amazon BeanStalk

§  Based on the Amazon EC2 infrastructure
§  Add Linux, OpenJDK, Tomcat
§  Currently in beta
§  …and only in US-East
§  Eclipse Plug In available
§  Supports version handling of applications
§  Supports scaling depending on load indicators
§  Access to Tomcat logs etc.

§  Videos to get started
§  Demo application based on Spring
•  Uses also S3 (storage) and Simple Notification Service (SNS)

34

PaaS: Conclusion

§  PaaS offer a runtime environment in the cloud
§  Usually based on IaaS
§  Makes Cloud easier usable for a developer
§  …but less customizable and less choice
§  Technical challenges (scalability) solved by platform
§  Spring can be used as a universal model

§  Each PaaS differentiate themselves with their added services
§  VMforce: database model, charts, reports, GUI elements
(functional components)
§  GAE: Big Table, caching, security, image processing, blob storage
(technical platform)
§  Amazon Beanstalk: Standard Java stack
§  Choose the right tool for the job!
35

PaaS with Java

Recommended

More Related Content

What's hot (16)

Similar to PaaS with Java (20)

More from Eberhard Wolff (20)

Recently uploaded (20)

PaaS with Java