Working for a customer recently on BizTalk ESB Exception management, ran into issue when creating a generic orchestration for handling exceptions. This generic orchestration will be used by all other applications in the enterprise. I ran into issue when an exception message is sent to this generic orchestration, we had 100% CPU utilization and found out to be a issue with the BizTalk ESB Exception management.
Check the article:
Improving the ESB Toolkit: Fixing the "endless loop" bug when creating Fault messages http://www.bizbert.com/bizbert/2011/05/06/Improving+The+ESB+Toolkit+Fixing+The+Endless+Loop+Bug+When+Creating+Fault+Messages.aspx
But I don’t want to go and change the Exception management framework; I did a simple fix by passing the exception and fault message to the generic orchestration from the main process. In the generic orchestration have an expression shape to throw the exception and have an exception block which I used the ESB exception framework to send out the message to the BizTalk ESBExceptionDb.
Even though it took additional step for this, I prefer this to be clean and simple solution until we have a fix in the BizTalk ESB toolkit. Hope this helps…