In flow designer,facing 60 sec timeout and wait for condition issues while executing NowAssist skill

MDBilalI · a month ago

Hello All,

In ServiceNow Flow designer I am getting this error when I am executing a Now Assist skill - "An execution timed out with timeout of 60 SECONDS", does anyone know how to fix this issue. (before upgrade)

Also, some more issues we are facing after our instance upgraded to Australia version:

1. Flow designer is taking so much time to execute, some time it is in "Queued" state and completed after some time but sometimes it never runs.
2. If there is any wait for condition in the Flow, for example - wait for condition is of 1 minute but once flow is reached to that step and even after 1 minute flow did not start, sometimes it starts after 10 minutes, 30 minutes or even more up to hours.
3. The 60 sec timeout issue is still there, mostly happening with Now assist doc intel use cases.

If someone know how to fix these issues or any resources you have please share, please provide your inputs if you have any idea.

Thanks,

Bilal

Tanushree Maiti · a month ago

Hi @MDBilalI

Could you please confirm which version of Now Assist you are currently using, along with the package name?

Additionally, please ensure that your instance is running the latest version of the Now Assist Skill package. You can check for available updates in the ServiceNow Store or on your instance’s plugin page.

sanjaygupta · a week ago

These three issues share one root: after an upgrade, the Flow Designer queue can fall behind, and Now Assist skill calls are slow. I couldn’t find an official fix that applies to all of them, so I’d treat this as a case for ServiceNow Support once you have the evidence below.

For the 60 second timeout, the same thing was reported in the community: a skill runs fine when tested on its own but times out when called from Flow Designer with Execute Skill. The advice there was to move to the latest patch, but the poster still saw it on a recent Zurich patch with several different LLMs. Another post suggests that very large prompts and inputs can cause timeouts. I found no documented way to raise the 60 second limit, so I wouldn’t promise one. Try these steps:

Update Now Assist and Flow Designer to the latest store versions and patches in your instance.
Pass only the fields the skill needs, not the whole record, to keep the prompt small.
For Doc Intel, test with a smaller or shorter document to see whether size is the trigger.
Compare a run from the skill’s own test screen with a run from the flow.

For flows stuck in Queued, the way it works is that when a flow’s trigger fires, the system creates a flow.fire event, and out-of-box jobs pick up that event and run the flow. If those jobs are busy, flows wait in Queued. Check this:

Open the sysevent list in your instance and filter the name to start with flow.fire. See whether events are piling up in the Ready state.
Open the sys_trigger list and look at the Flow Engine event handler jobs. Check they are running and not stuck.
Open All > Flow Designer > Operations > Executions and look at which flows are Queued and since when.
Repair or update the Flow Designer plugin.
Don’t edit the out-of-box jobs yourself.

A support article called “Flow Designer is not executing as expected” is also worth reading for this one.

The Wait for condition delays are likely the same problem. A community post describes flows staying in a waiting state for hours even though the condition was met, and the same event backlog could cause it. If the delay always happens after the upgrade and across many flows, it is probably an instance-wide queue problem and not the flow itself.

All three started after the Australia upgrade, so open a Support case. Include the flow execution IDs, the queued event counts and a screenshot of the job list. Test any change in a sub-production instance first.

To sharpen this, share your Now Assist and Flow Designer plugin versions, the patch level, and whether the delays happen at specific times of day.

Sources: