Bring your towel and join one of the fastest growing data communities. Welcome to our second-generation open source orchestration platform, a completely rethought approach to dataflow automation.

Prefect Community

Hey there. We have been seeing this error during some flow runs `No heartbeat detected from the remote task; marking the run as failed.` . Any info on what is possibly causing this?

For more context we are hitting an external API and getting results then at some point the task seems to just stall out

Hi <@U01LHF70VRV>, I think this <https://prefect-community.slack.com/archives/CL09KU1K7/p1632751713328500?thread_ts=1632493440.260000&amp;cid=CL09KU1K7|answer> is a good starting point. You can also see <https://docs.prefect.io/orchestration/concepts/services.html#heartbeat-configuration|this> part of the docs for the configuration

this has been happening to us for awhile too, no fix has worked as of yet

What version are you on? I think 0.15.4 propagates the error better to give us a better idea.

Have you tried the heartbeat config? I tried replicating this by spinning up an API with a very long sleep call and couldn’t. Are you querying an API too?

yeah we tried the hearbeat config, no luck.  we are on 0.15.4 on both the agent and the flows.  not an api, a long running db query.  confirmed its not a memory issue too.

Gotcha. Will bring it up when I chat with the team

Agh. Are you getting any of our heartbeat failure logs?

i think what was happening is we had this in our job templates
```restartPolicy: OnFailure```

switching this to Never to let prefect handle the retries for us, and I think this will resolve our issue

Glad to hear. I think that we should still handle multiple pods attempting to run the same flow better so if you have any heartbeat error logs that'd be helpful.

We are on version 0.15.5 so should be good there

Re-running a flow now to see if it does the same thing