Service Connect - Stopped workflow but No Errors


Can any Service Connect gurus out there explain to me why a workflow process would just suddenly stop dead in its tracks with no error message at all?

These same workflows that stop suddenly also run dozens of times to completion throughout the day. So it’s not the workflow. But at least every other day we have a few random processes (from varying workflows) that will stop in the midst of processing records and just sit there with no indication (that I can see) as to why they stopped.

I usually have to track down which record the process stopped on, create a new file with the unprocessed records, and resubmit it to Service Connect. And those remaining records always process just fine. So it’s not as if any of the records are an issue.

I do notice that this usually happens on a file that has quite a large amount of records…but most of the time that same number of records runs to completion without stopping. So why would one workflow process 500 records in one instance but then stop in the middle of processing 500 records in another instance.

I’m not real clear on how Service Connect works as far as does it grab up AppServers on our Epicor Server and let go of them throughout a workflow process or does it grab up one or a set of AppServers and hold onto them till it finishes all the records. And maybe that comes down to our setup. Just thinking that maybe it lets go of an AppServer and then tries to grab another one but they are all busy and so it just gives up.

Just wondering if someone could shed some light on why this is happening and what settings I could look at (Service Connect? Epicor? SQL?) to try to help fix the problem.

We are running Epicor 9.05.702 - 64-bit - Progress
Service Connect 10.1.4020 - on a separate VM

Thank you in advance for any help you can provide,

From my experience, Service Connect does not “stop in its tracks” with no error. Sometimes it goes to Finish for some reason, or doesn’t branch or redirect like I thought it should. There are several types of branches; can you upload a snip of the overall workflow and the end portion of the activity trace?

Fortunately/Unfortunately I don’t have any workflow processes that are in that state right now but when it happens again (more than likely sometime tomorrow) I will grab some print screens (which I should have done initially) of the state that the Trace is in and the workflow and send them to you. Thanks for the response! I’ll update as soon as I can grab those things.

Perhaps it is worth looking at how you have imported your assemblies. There are two licensing modes if I recall. Maybe your app-server logs may yield some further information at the point of where it stops. Is there anything in the Service Connect event log when this occurs?

Definitely some great ideas of things to look at. Thank you! I’ll take a look at my AppServer logs when it happens. I can’t say that I’ve ever seen anything in the Service Connect event logs to indicate why this happened. I’ll send a print screen of that as well when this occurs again.
I wonder about creating a separate AppServer that handles just Service Connect calls. (I probably should have done this a long time ago…) Just wondering if that might help fix this issue that I’m running into.
I’ll definitely post more information as soon as I have this problem again. Thanks!