Forum Replies Created
-
AuthorPosts
-
November 27, 2008 at 11:45 PM in reply to: Custom Pipeline component error : Subscription issue #21263
You should have disassembler component in your custom pipeline. It will promote message type. Or your orchestration should accept XmlDocument.
Theese three messages come from a single port ?
That’s extremely important to know.
No… I will use code blocks for database operations. and transaction context will propogate. (i know that SQL adapter does not support transactions)
I beleive so.
But the behaviour I’m looking for – Is to be able to add new orchestrations for new TradeTypes without recompiling the Broker orchestration.
The broker has [If-Else-Else] Shape. And it needs to be changed when new trade types appear.
All I need – is routing with transaction support… Or maybe some magic in the Broker orchestration which will allow me to call child orchestrations without specifiing it.
November 25, 2008 at 12:55 PM in reply to: Cannot transform a message after debatch in a receive pipeline called from orchestration #21256Hi all,
Just an update. I was able to resolve this issue. Basically the instance of MyMessage generated by the orchestration did not match the schema used in the map, so I modified the message format by modifying the results sent back from the call to a stored procedure (as I use a SQL adapter in my orchestration): in my message, I was using child elements instead of child attributes, as defined in the schema.
Best to all,
Daniel.
I find it very difficult to believe that Microsoft would make people re-map everything after a change on the schema. I work with health care X12 and the maps are complex and take a long time to complete. There has got to be a way to keep the links. It is very impractical to re-map everything every time we make a change to a source or destination table. Has anyone been able to find a way to keep the links. I have worked with other mapping software and they keep the links after changes to the tables. Therefore, I find it hard to believe that Microsoft did not copy that model as well.
Thanks Greg for the response. I was hoping to avoid a custom component but it is looking like it is the only option.
Thanks again,
Geordie
This will still not give you the behaviour you would expect. Although the transaction will propagate to the child orchestration, it won’t go any further. If the child orchestrtaion calls a web service or sends via the SQL adapter the transaction does not propagate to these. The send port is separated from the orchestration by the messagebox, any error will not rollback a web service call or SQL adapter send.
You will need to introduce a compensation block and have undo functionality. Or some other mechanism to provide transactional integrity.
Ok. I see.
I have one broker orchestration, which accepts XML message. The message contains several submessages (1 – *). Each submessage has XML Element TradeType <TradeType>A</TradeType>. I use TradeType in order to route thate submessage to specific orchastration.. I have several orchestrations for each TradeType.
I want to transmitt each submessage from my broker orchestration to specific SubOrchestration (depending on the message TradeType value). But, If some message fails I want to roll back all submessages processings.
I think I have to create a broker Orchestration with Atomic transaction scope, put Loop element into the Scope, and extract submessages in Loop.
Then inside Loop goes If-Then Element which performs routing basing on TradeType. Insede each Branch I need to put Call orchestration element (because call orchestration element will allow me to spread transaction context on to the chil orchestration.)
Hi, I have a similar problem. Could you please tell me how you solved it.
Many Thanks
Eddie
No the transaction context is not passed to the downstream orchestration. It is possible that the downstream orchestration will run in another process or even on another server.
The atomic transaction in an orchestration means that orchestration will persist its state at the beginning of the scope and remain active until the scope exits. If the Biztalk process terminates and restarts, the orchestration will restart at the beginning of the atomic scope. There is no mechanism to advise the downstream orchestration that this has happened. If you throw an exception inside the atomic scope there is no rollback, you cannot undo the send if it has completed. You can however add a compensation block to the scope and perform your own logic to undo what may have been done
I think putting a send/receive pair inside an atomic scope will be counter productive. There is no benefit and several possible problems including resending the message.
You may need to write a custom MIME Decoder pipeline component.
The standard MIME/SMIME Decoder scans the first 1024 bytes of the incoming message to search for the “MIME-Version” header. If this header is not found, the message is treated as a non-MIME message. The HTTP Adapter will strip off the HTTP headers before passing the message up the pipeline so you may lose the things like MIME-Version, etc
November 23, 2008 at 2:27 AM in reply to: Step by step sample tutorials or vedio of biztalk server 2006 #21242Try these books.
Apress First Steps Developing BizTalk Applications
Apress BizTalk 2006 Recipes A Problem Solution Approach
November 22, 2008 at 11:35 PM in reply to: Step by step sample tutorials or vedio of biztalk server 2006 #21241Refer the MSDN and search for Biztalk Labs….There are 4 labs which you can download and practice it religiously. Nothing could help you to learn Biztalk better than these labs.
All The Best!!!
.
thanxs alot for such valuable information …
regards,
preety
Enterprise Single Sign-On (ESSO) is used to store send port and recieve location configuration info for Biztalk. The service is usually installed on your biztalk server when you install biztalk. You should look under Services and make sure the service is started. Here is a link to help explain SSO http://msdn.microsoft.com/en-us/library/aa578564.aspx
-
AuthorPosts