worker and ui pods do not have liveness probes

[clayzermk1]

What happened?

I was looking into liveness probes for the various pods that get created in my cluster. I noticed that some pods have liveness probes and others don't. In the deployment_builder code for base, admintools, and ui, I can see that:

• The worker pod is explicitly omitted from liveness probe creation (I found the issue and PR).
• admintools runs a different liveness probe configuration from frontend/history/matching.
• The ui pod doesn't have one at all.

What did you expect to happen?

I expected the worker and ui pods to have a liveness probe.

Is there a reason that the worker and ui pods have no liveness probes? I understand that the worker pod doesn't have an RPC endpoint, but would it be OK for it to do something simple like check the filesystem? Similarly for the ui pod, would it be OK for it to perform a basic liveness check? Were these omitted intentionally, or is it something that you'd be open to accepting a PR for?

[alexandrevilain]

Hi @clayzermk1 !

Thanks for reporting this issue !
For sure you can open a PR for this. This would be a great addition! 👍