Unattended flow keeps running

(0) Share

Report

Posted on by YM-26051441-0

Hi!

I've got a unattended Power Automate desktop flow running on a VM. The flow has been running for about 3 weeks without issues. Yesterday, all of a sudden, without changing anything to the flow, the flow was stuck after succeeding an action. It seems like it didn't even try the next action. In the logging of Power Automate action #17 is the last action recorded.

After action #17 of the subflow was succeeded it should move on to action #18.

I also have error-handling on action #18 so if it tried to execute action #18 it couldn't be stuck running for hours. It should only try 3 times with an interval of 5 seconds. So it seems like it never tried to execute the next action.

Does anyone know what could be the cause of this? I could only think of a bug or some lost connection on the VM or something like that.

Extra info: in the 3 weeks before this it never happened, in the same run the bot succesfully executed the same actions 1 minute before (for another work queue item). And after me cancelling the flow manually and restarting the flow it succesfully ran again. Without changing anything.

Thanks in advance!

Categories:

Power Automate Desktop

I have the same question (0)

All responses (2)

Answers (0)

Suggested answer

Inogic 1,265 Moderator on at

Like (0)

Report
Copy link

Link copied!
Hi,

From what you described, this doesn’t look like a logic issue in your flow — it’s more of a runtime/environment problem.

In unattended mode, Power Automate actually creates and manages a remote desktop session on the VM to run your flow. If anything goes wrong in that session (like disconnect, lock, or resource issue), the flow can just hang without throwing an error.

Also, Microsoft mentions that disconnected/locked sessions or machine issues can block or interrupt execution, especially in unattended runs.

That’s why:

Your flow didn’t move to the next step

Retry logic never triggered (because next action was never reached)

Restart fixed it (fresh session = clean execution)

What you can do

Add logs before/after key actions to track where it stops

Set timeouts on actions (to avoid infinite waiting)

Ensure VM is stable (no disconnects)

Hope this helps!

Thanks!

Inogic

Was this reply helpful? Yes No
YM-26051441-0 13 on at

Like (0)

Report
Copy link

Link copied!

@Inogic Thanks for the reply. Yes, that is what I was thinking. Just not sure how it happens. Ever since I posted this question it ran smoothly every day. So, to me it's weird that this happened 1 out of 30+ days. Probably a glitch then I'm assuming?

Was this reply helpful? Yes No