Search code examples
biztalkbiztalk-2016

BizTalk Direct-Bound Send Port Slow to Trigger

I have a BizTalk 2016 Enterprise Dev Environment with latest FP2 appropriate CU applied.

I have a "Gatekeeper" orchestration bound to a WebHttp receive location. This simply publishes an incoming xml message to a direct-bound port called "MvcFormsPort". From here, a physical request-response port calls a WCF web service and the response flows back to the orchestration and is returned to the caller of the WebHttp receive location.

Functionally, this is working fine. However one one particular server (integration test), there is a problem with performance which seems to be down to the time it takes for the physical request-response port to wake up to the fact that a message has been published to which it has a subscription. This can be seen in the following screen grab. The logical direct-bound port "MvcFormsPort" received the message at 11:43:47 but the subscribing physical send port doesn't collect this until 11:44:02; 15 seconds later!

The same process on on my dev vm takes about 1.5 seconds end-to-end.

enter image description here

The polling settings for the hosts are still at the default of 500ms.

Any ideas what may be causing the delay in the problem environment?

Solution

  • My bad! The problem was a connection string missing in the SSO of the problem server.