BizTalk Server Deep Dive Tips and Tricks For Developers and Admins at TUGA IT

Editor's Notes

• #7: It’s critical to use good tools and techniques to produce working solutions as quickly as possible and at the same time, given the increase the requirements and number of applications organizations develop today. But at the same time, it's also critical to maintain the health of the entire platform. In this session, which I'll try to be a very interactive session (so be prepare to participate), I'll address and share some useful BizTalk Server Tips and Tricks (and Workarounds) both for developers and administrators. Covering some topics like RosettaNet, SAP, database maintenance, debatching, out-of-the-box pipelines vs custom pipelines and many more.
• #8: Make a proper effort estimation
• #16: It is possible to extend the framework by writing your own: MetaInstruction and Instruction classes in assemblies which should be deployed to the GAC Creating vocabulary definitions that correspond to public methods on your MetaInstructions, and then using the AddMetaInstruction vocabulary definition in the BREPipelineFramework vocabulary in InstructionLoaderPolicy BRE Policies which will allow you to use the your custom vocabularies in the ExecutionPolicy.
• #17: It is for these type of person… and I will start with this tip…
• #18: I love this tip… you all remember my dear friend Tord Glad Nordahl complaining every year in his session about Developers writing unnecessary information Application Log… and for they do not use the Event Viewer… well I partially agree with him… you shouldn’t write custom application errors, warnings of information in the Application Log… but you can use a custom log for that I really don’t care about that since you do not use the Application Log to write these stuff What the Admin does normally? Ask the developer to change is code, that already is deployed in all the environments, to not write in the application log And they will try to find a thousand excuses for explaining why such information is important My advice: Let the developer by happy by writing in the Event Viewer But take back the control of your environment by easily creating or using PowerShell
• #19: Have you had the time to think about what is some features like RosettaNet, ESB or UDDI have in common? Well, all of them have custom databases and all of them are optional features The important question here, because they have custom databases, is: Do you think that these databases are being backuped? And the data inside are saved? And the response is NO… by default the BizTalk jobs don’t perform any operations in this databases. Fortunately for us, Microsoft provides two SQL Scripts that you need to run against this databases in other to extend the standard backup mechanisms but you also need to modify the adm_OtherBackupDatabases table to include a row for each of your custom databases.
• #20: All the BizTalk database which is being backed up by the 'Backup BizTalk Server' job, so all databases with the exception of the BAM Star Schema database (BAMStarSchema), has one table called “MarkLog”. By default each 15 minutes, a string is stored on that table, created by the 3th step (MarkAndBackUpLog) of the ‘Backup BizTalk Server’ job. This tables are holding all the transaction marks And there isn’t any job that cleans these tables so this tables are always growing unnecessarily… the solution provide by Microsoft is using Terminator Tool… which means stop your environment for what… to purge all the data from these tables Luckily for us…. You don’t need to do that… I decide to create a stored procedure to extend the delete history capabilities of BizTalk to include this tables according to some best practices Credits also to Tord Glad Nordahl, Rui Romano, Pedro Sousa and Mikael Sand that helped me to create and improve this script I really think that Microsoft should take this script, review it and include in the next version or as a hotfix
• #21: Normally Developers have in their environments Full tracking Enable because it's easier to debug and see what is happening with your new applications Does Developer’s remember to disable it before they create the resources to put in production? No! And actually… they don’t care about that -- is not there task to do it And sometimes they don't know the right configurations that should be apply into production This can be an annoying and time consuming operation and asking developers to change their My advice is… take back the control of your environment These tasks can be easily automated and configured by easily creating or using PowerShell To Disabling all Tracking or enable just the important ones at the application level or addressing all of them You lose 1 day developing these scripts and then you do not need to worry any more Ok enough of admin tips… And now Do you know what I mean about Developer Tips?...
• #25: It is for these type of person… thinking in debatching mode and in unrecoverable interchange mode too
• #26: What developers normally do? Create an Envelope Schema Create a Schema Create a Custom pipeline that is usually a replica of XMLReceive pipeline The important question here is: Do I really need to do all of this? Is there any other way? Normally you don’t need to create a custom pipeline to accomplish this. The standard XMLReceive pipeline will perform this task is for you: You just need to define the envelope and the Body XPath property And you already did that! So you don’t need to developer additional artifacts Important considerations When Debatching messages, Recoverable Interchange Processing is an important feature it will allow you to process completely even if one or more messages in the interchange fail at the following stages/phases: Disassembly stage of a receive pipeline XML validation stage of a receive pipeline Map execution phase of a receive port We will see this in more details in the upcoming demos
• #27: Same topic: Pipelines - Validate XML Messages or converting Flat Files  What developers normally do? Create a Custom pipeline with the “XML Disassembler” component or the “XML validator” component And for Flat Files we create a custom pipeline with a “Flat file disassembler” component I had a project with almost 100 flat file schemas so imagine how many pipelines were created in that project Again: Do I really need to do all of this? Is there any other way? Again… you don’t need to create again and again in every project a custom pipeline to accomplish this. You can Create a generic custom pipeline like the “XMLReceive” or a generic “FlatFileReceive”… yes you can create a generic “FlatFileReceive” pipeline using a special trick that we will see in the demos and reuse it whenever you need Or use the standard pipelines and you just need to change the settings at runtime
• #28: Request-Response Content Based Routing with LOB operations is that possible? Yes of course.., it’s easier to create orchestration to accomplish this but You don’t need orchestrations to perform a simple insert, delete, update or select using WCF-SQL or send an IDoc for SAP You don’t need orchestrations for request/response scenarios! Whenever it’s possible avoid using orchestrations The main problem is that using Lob Adapters is that you need to specify an Operation name for a specific action, for example Operation Name=Insert for the action TableOp Insert Delete for the action TableOp Delete And so on That need to be promoted in runtime What you can do/use? Create a generic custom pipeline and pipeline component to promote the BTS.Operation property
• #29: It is possible to extend the framework by writing your own: MetaInstruction and Instruction classes in assemblies which should be deployed to the GAC Creating vocabulary definitions that correspond to public methods on your MetaInstructions, and then using the AddMetaInstruction vocabulary definition in the BREPipelineFramework vocabulary in InstructionLoaderPolicy BRE Policies which will allow you to use the your custom vocabularies in the ExecutionPolicy.
• #31: It is possible to extend the framework by writing your own: MetaInstruction and Instruction classes in assemblies which should be deployed to the GAC Creating vocabulary definitions that correspond to public methods on your MetaInstructions, and then using the AddMetaInstruction vocabulary definition in the BREPipelineFramework vocabulary in InstructionLoaderPolicy BRE Policies which will allow you to use the your custom vocabularies in the ExecutionPolicy.
• #32: Normally when we create a custom functoid they need to be copy to the “Developer Tools\Mapper Extensions” folder under the BizTalk installation path and install in the Global assembly cache (GAC) However this is not entirely true When developer a custom functoid you need to be aware that we have two types of Functoids Normally we use the SetExternalFunctionName expression to define the action of this functoid In this case we are basically saying that we will use an external assembly to perform this action and in this case this assembly needs to be installed in the GAC in all the environments that you will use this functoid However we have available another expression to define the action of the functoid… by using the SetScriptBuffer expression. What’s the difference between both? Well do you know what’s happen when you use an out-of-the-box string Functoid in your map? The C# code will be embedded in the XSLT code that will be deployed… and in this case you don’t need to deploy any assembly to your environment This is basically what happens when you use the SetScriptBuffer expression, this type of functoid will tell the BizTalk Map to embed the function in the XSLT code… The assembly of this Functoid will only need to be copy to “Developer Tools\Mapper Extensions” folder in your developing environment and you don’t need to install nothing in production
• #34: Let’s start with an easy and simple tip: Database Lookup Functoid I always find hard to remember the correct value for the connection string to be used inside this functoid The Easiest way to make sure we are using the correct connection string value and for not having to remember this by head is to create a simple Universal Data Link (.udl) File… set OLE DB provider connection parameters and test the connection to check if everything is correct After data Open the file in notepad and you will find the connection string value that you can copy and use in the functoid Important considerations: You shouldn’t Hard-coding this value directly in the functoid otherwise it will be a nightmare when you deploy this to a different environment. You can and you should store this parameter in a different storage location (SSO, Registry or others) and get this value using a scripting Functoid or custom functoid which can then be linked to the Database Lookup Functoid
• #35: It is possible to extend the framework by writing your own: MetaInstruction and Instruction classes in assemblies which should be deployed to the GAC Creating vocabulary definitions that correspond to public methods on your MetaInstructions, and then using the AddMetaInstruction vocabulary definition in the BREPipelineFramework vocabulary in InstructionLoaderPolicy BRE Policies which will allow you to use the your custom vocabularies in the ExecutionPolicy.
• #40: Understanding RosettaNet can be a challenger believe me so in this tip I will provide some important topics that you need to be aware about it: BTARN require: Microsoft .NET Framework 4.5 If you have multiple versions of .NET Framework installed on your computer, make sure that the BtarnAPP Web application is referencing .NET Framework 2.0. You can configure this by using the Internet Information Services (IIS) Manager. DESPITE the BtarnAPP Web application working properly in .NET 4.0 You need to create a WebService extension for BTSHTTPReceive.dll, configuring the IIS isolation mode. The BizTalk Host Instance Account and the BizTalk Isolated Host Instance Account should be the same. Otherwise, BTARN will not function correctly. If the service account set for the BTARN application pools is different from the Isolated Host account, BTARN will not be able to process incoming messages correctly. When the receive .aspx page calls the pipeline, the pipeline will not have access to the appropriate certificates. Therefore, it will not be able to decrypt the incoming message or validate the signature. It will also not be able to access the MessageBox database. BTARN does not support the use of alias created for SQL instance to configure the BTARN database. RosettaNet requires the IIS_WPG group (group provided by IIS 6.0 that provides the minimum set of user rights and permissions required to run an application) unfortunately for us this group doesn’t exist anymore… so you need to create it manually   Very important: The accelerator requires both in-process and isolated host to be marked as "Authentication Trusted" and "32-bit only“. And the reason why in-process need to be also trusted is that the BizTalk Host Instance Account and the BizTalk Isolated Host Instance Account should be the same and you cannot use the same account for trusted and untrusted… otherwise it was only necessary to mark the isolated host as trusted. Important: Make sure the BTARN in-process host is the default send handler of HTTP Adapter, when you create a Partner it will create to send ports and it will use the default send handler, if the BTARN in-process host you need to stop the BTARN process unbinding the send ports fix the send handler and start everything again… so it induce a small shutting down BTARN supports enhanced tracking using BizTalk Activity Monitoring (BAM). Microsoft provide a Tracking file with the activity definitions however: you need to you BAM views There are no maintenance processes to clean up this database