I see that I can configure the

interval_seconds

of a

KubernetesJob

so that the workers poll for Kubernetes job status more or less frequently. I have set this to

300s

(default is 5s) in the

job_variables

argument from

Flow.deploy.

This value now shows up in my Deployment configuration: However, I do not see any change in behavior. I would expect that a pod that runs for only 1 minute would now show in the Prefect UI as running for longer (in most cases) because Prefect is only polling for the status every 5 minutes. This makes me think the configuration is not setup correctly. Does anyone know how to correctly setup

interval_seconds

to change the polling speed of Kubernetes job status?

Hey @Scott Walsh! What version of

prefect-kubernetes

are you using? There were some pretty big changes between 0.5.x and 0.6.x that affected job/pod watching