Feature idea: extend logging by the leader / follower status of replica

[mimowo]

/kind feature

I would like to consider extending the set of injected key-value pairs by controller-runtime to include the "replicaRole" key, indicating "leader" or "follower". This could be either by default, or opt-in when wiring a controller or webhook.

This would be very helpful when filtering logs for debugging to quickly see all logs, and only those from the "leader".

I'm exploring the solution inside controller-runtime as it is already injecting some key-value pairs into the logger, plus it is aware of leader election.

The motivation is this effort in Kueue: kubernetes-sigs/kueue#6772, and probably a workaround is possible using LogConstructor option.

[mimowo]

@alvaroaleman @sbueringer could you help us triage the feature request?

cc @mwysokin @tenzen-y

[alvaroaleman]

The problem with doing this globally is that this can change at runtime and I don't think logr/zap or klog allow passing a closure they call as value for a field, so it is unclear to me how to do this outside of the reconciler log.

The idea as such makes sense.

[austince]

cc @msudheendra-cflt

[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

[mimowo]

For anyone interested we have already implemented this feature in Kueue using the LogXonstructor option. You may find the technical details following the PR links in the issue kubernetes-sigs/kueue#6772.