Gracefully exit when setup fail

[AlmogBaku]

trafficstars

https://github.com/kubernetes-sigs/e2e-framework/blob/d19e8b569fb2f94568f8ed74b00227f879798aa0/pkg/env/env.go#L369

Due to the usage of klog.Fatalf(), it calls internally os.Exit(), and we can't gracefully run the clean ups.

[vladimirvivien]

@AlmogBaku thank you for pointing this out. Can you provide a bit more information or context so that an appropriate fix may be implemented.

[harshanarayana]

@AlmogBaku if the workflow were to fail during the setup, don't you want to retain the setup intact instead of cleaning up ? This might help in digging into what actually caused the setup to fails easily right ?

There is an option similar to this called disable-graceful-teardown that we enabled a while ago that can prevent teardown on the test failure. Are you suggesting for an option similar to that but the one that can handle setup failures ?

[ojhughes]

Hi, I also hit this issue. envfuncs.LoadDockerImageToCluster() was failing in my setup function, which resulted in the kind cluster not being destroyed

[vladimirvivien]

@harshanarayana I went ahead and label this as a bug. The internal logger should not disrupt the execution flow of the users test.

[k8s-triage-robot]

The Kubernetes project currently lacks enough contributors to adequately respond to all issues.

This bot triages un-triaged issues according to the following rules:

• After 90d of inactivity, lifecycle/stale is applied
• After 30d of inactivity since lifecycle/stale was applied, lifecycle/rotten is applied
• After 30d of inactivity since lifecycle/rotten was applied, the issue is closed

You can:

• Mark this issue as fresh with /remove-lifecycle stale
• Close this issue with /close
• Offer to help out with Issue Triage

Please send feedback to sig-contributor-experience at kubernetes/community.

/lifecycle stale

[vladimirvivien]

/remove-lifecycle stale

[codegold79]

I'd like to work on this issue. I'll try to reproduce the problem ofos.Exit after failed setup as my first step.

[k8s-triage-robot]

The Kubernetes project currently lacks enough contributors to adequately respond to all issues.

This bot triages un-triaged issues according to the following rules:

• After 90d of inactivity, lifecycle/stale is applied
• After 30d of inactivity since lifecycle/stale was applied, lifecycle/rotten is applied
• After 30d of inactivity since lifecycle/rotten was applied, the issue is closed

You can:

• Mark this issue as fresh with /remove-lifecycle stale
• Close this issue with /close
• Offer to help out with Issue Triage

Please send feedback to sig-contributor-experience at kubernetes/community.

/lifecycle stale

[codegold79]

/remove-lifecycle stale