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

Are there known issues with the kubernetes integration not correctly detecting crashed pods and/or not correctly propagating these to the API?

For example, I'm seeing in the logs that a flow should be marked as crashed, but in the UI it is still tagged as `running`
```07:16:52.914 | INFO    | prefect.flow_runs.worker - Reported flow run '644692a0-4442-4f25-aa41-51be959c11df' as crashed: Flow run infrastructure exited with non-zero status code -1. ```
```~ $ curl -s '<http://prefect-server/api/flow_runs/644692a0-4442-4f25-aa41-51be959c11df>' | jq '.state.type'
"RUNNING"```

Tangentially related: I've filed <https://github.com/PrefectHQ/prefect/pull/17333>

Our pods are using quite a bit of memory and are getting evicted regularly and/or waiting to find a node with no memory pressure, it feels like Prefect has issues handling these scenarios