Hey guys,
I'm stumped.
So I have a scheduled flow that basically grabs (up to 6) new incoming e-mails, parses their body in a 'for each' action and post relevant information into my SharePoint list.
Normal execution time is between 20~70seconds (depending on the amount of emails I pass to the 'for each' action) -- not great but this is acceptable (was going to optimize it later).
Unfortunately, ~ once per day, the flow's execution simply hangs as "Running". It's execution time balloons up from the previous 20~70 seconds to hours per single flow run. What's worse, it does not fail (wish it would); it simply remains stuck as "Running".
Also, if I cancel the flow, "Run History" never specifies which action caused this sudden delay. The cards where it stopped -- usually (but not always) the 'for each' action -- remains grayed out (won't let me expand it).
I disabled all parallelism (both in the the flow trigger, and all my for each actions) hoping this would solve it, nope. Tried lowering timeouts/retries on all API calls to (max) under 2 minutes - still no joy.
Only way to fix it is to cancel the hanged/running flow, "Turn off" the entire flow, and then turn it back on. *Poof* everything is back to normal until the next time it hangs at some random point in the future.
I wish there was a flow run-time timeout and/or better performance monitoring on execution (for time optimization), I don't like mucking around in the dark without proper debugging tools - "Run history" is ok, but still lacks in many areas. Was hopeful I could take this solution to production, it was looking great in testing, but now with this issue, I simply can't risk my flows hanging up like that without advising me or my team.
