Working on adding the new worker pods to the helm chart after the split from the scheduler to address https://github.com/airbytehq/airbyte/issues/6416. Do they have a http healthcheck like the server does with

/api/v1/health

? I'm seeing that they are listening on port 9000 for heartbeating, but I'm not sure if that is valid to use for a liveness or readiness probe

I don’t believe they do

https://github.com/airbytehq/airbyte/blob/master/kube/resources/worker.yaml#L4 is the current worker yamls and it looks like there isn’t any there

we should definitely add a proper check

I found that and verified with curl. I just didn't want to overload it if its not a valid check for "is alive"

that’s intended to be between the workers and the pods. not a bad idea to repurpose as Kube health check probe

I’d say we can run with this for now. Jared and I will discuss with ya on the PR