Configuration Job Pod

[TartanLeGrand]

What would you like to be added?

Add a init container for wait redpanda to get ready.

Why is this needed?

Because the job crash loop while waiting for redpanda .

JIRA Link: K8S-629

[chrisseto]

Hey @TartanLeGrand could you provide a bit more detail? I'm entirely certain what you mean here. I was under the impression that helm would wait for the STS to be ready before deploying any post install jobs. Even then, the job should be configured to retry until it eventually succeeds. Are you observing otherwise?

[TartanLeGrand]

Hello 👋 @chrisseto ,

In our environment, the init pod keeps crashing until the StatefulSets become ready resulting in about five failures before it finally succeeds, since the StatefulSets only become operational at that point.

As a quick workaround, my first thought would be to introduce an init container whose sole responsibility is to watch for the StatefulSets to be fully Ready before kicking off the rest of the pod.