Local request call in helper loop says module stopped

[Zach_May]

I created a timer module for an application, but the users found that it would frequently stop counting but the main application continued working as expected. Sometimes this would happen after just a few hours, and sometimes after a couple of days. I added some status logging to track what was happening, and it looks like the timer module continued running as expected but the helper loop in the module stopped sending the message to check the timer like it should. After further debugging, it looks like the local request I created to trigger the "Check Timer" event was throwing an error saying that the module was stopped even though the main loops of the module were still running.

I'm at a bit of a loss regarding what's happening. The main DQMH loops continue running and responding to events, it's just the local request in the helper loop that appears to think that the module has stopped. The helper loop and local request were both created through the DQMH scripting tools though, so I'm not sure what the issue could be. It seems like enqueuing the "Check Timer" case directly to the module queue fixes the issue though, so I know it's something with the user events being trigger by the local requests.

I've attached the source code for the timer module in case it helps someone find the problem. I've thrown a disable diagram structure in the helper loop. The currently enabled case should be with the local request calls and the disabled case replaces those calls with direct enqueues for the events so both options can be compared.

Any ideas or suggestions what might be happening? I can make it work for now with the direct enqueues, but I like using the local requests since it's easy to track where those events are getting called. It's much easy to search for those than to look at all of the message enqueues and try to find the specific message I'm looking for.

[ChrisFarmerWIS]

Had a quick look.

Looks ok to me!

How many clones are you running at a time?  

Are you stopping and starting them regularly?

How is the memory of your overall application?  Perhaps the issue is something else entirely.

[Zach_May]

I only have two clones running, and I'm bundling the moduleIds into a cluster after starting the modules so there should be no risk of confusion. I only start the modules when the application starts and stop the modules when the application stops. There's no dynamic starting or stopping which is what confuses me since the local requests think the module was stopped.

Memory of the application is great, no concerns there at all. The application continues to behave and respond appropriately. It's only the timer modules that aren't working as expected, but again, the module itself is still running. It's only the local requests called by the helper loop that seems to think the module has stopped.

It's also odd because both instances of the module appear to stop working at the same time. So I think the issue is related to the Timer.lvlib:Local Instance Events SEQ.lvlib:Obtain SEQ.vi or possibly Read SEQ.vi since those two are responsible for returning the correct local instance user event references. I just can't find what could be going wrong or how.