I am monitoring Biztalk. There is an app pool in the biztalk server and from there it talks to Biztalk itself. I have the app pool and the biztalk process monitored by the .net agent. However, I can see how the transaction in the app pool ends and then starts a new purepath for the work it does in biztalk. There is a custom entry point that it is used from the Biztalk fast pack. If I want to see those 2 in one purepath, do I have to select a different entry point? How does the dynatrace header get lost between both processes?
Answer by Andreas G. ·
BizTalk has managed and non-managed (native) implementations - especially when messages are passed through BizTalk, persisted, processed and workflows are executed. In that case we sometimes have issues to pass our PurePath ID from Tier-to-Tier as some of the code along the message processing is native and we cant instrument it. In these scenarios it might also be necessary to define custom entry points - or - make sure that the standard .NET Sensors such as .NET WCF and .NET Remoting are set to "active and start PurePath" in your System Profile. That will make sure you get new PurePaths from each BizTalk tier. "Correlating" these then probably needs to happen through some custom Transaction IDs - using Business Transactions to group PurePaths by this type of BizTalk Transaction ID
Answer by Eddy G. ·
Thanks for you answer.
I had the .Net WCF and .Net Remoting set to active and start purepaths. At the moment, the entry points defined for the Biztalk process is the WorkItem.Execure(), which is the one that starts the orchestration if I am not wrong. Might I have a chance to link the purepath between the app pool I am monitoring in the same server where Biztalk is and the Biztalk process if I define another entry point besides the one that comes with the Biztalk fastpack? If not, then I assume the only way is by either using the Tagging ADK or correlating it through a Business Transaction looking at a common parameter like messageId or some sort of Id right?