How to Use Field Selectors with Controller-Runtime to Filter CRD Events on Api-Server side?

[jimmyyuhao]

Body:
Hello,

I am working with the controller-runtime package in Kubernetes and I need to implement a functionality where the API server only notifies my controller about specific Custom Resource Definitions (CRDs). Specifically, I want to filter events so that I only receive notifications for CRDs where metadata.name is equal to a specific value, such as nodeName.

I would like to know:

How can I configure the manager initialization to achieve this?
Is there a way to set up a field selector to filter these events at the API server level?
If direct support for field selectors is not available, what would be the best approach to implement this functionality in my controller?
Any guidance or examples would be greatly appreciated!

Thank you!

[guettli]

same here. I could not find that in the docs. I found that design doc:

https://github.com/kubernetes-sigs/controller-runtime/blob/main/designs/use-selectors-at-cache.md

Which is implemented in #1435

In my case: I want to watch one configMap. But I don't want to filter the cache for that, because I want to also read other configMaps. Afaik I can't get objects via the default (cached) client, if the cache is filtered.

@jimmyyuhao maybe this conversation helps: https://kubernetes.slack.com/archives/C02MRBMN00Z/p1731692196792769

[MuneebAijaz]

		Watches(&corev1.Namespace{}, handler.EnqueueRequestsFromMapFunc(func(ctx context.Context, object kubeclient.Object) []reconcile.Request {
			return mapNamespaceToCustomResource(object, r.Log)
		})).

Here's an example from my setup in SetupWithManager. In the last function, you can filter objects based on any of their fields, and tell the reconciler about which resource u want to be reconciled using reconcile.Request.

You can also filter the cache if you are sure that your objects are of particular selector/namespaces. But then you wont be able to get events outside of those options.

[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

[guettli]

@MuneebAijaz afaik your example filters on the client-side. This question is about filtering at the server-side.

[MuneebAijaz]

right in that case @guettli your answer covers it. Cache can be filtered based on namespace and object level, so only those objects are stored in informer cache and will be returned when queried.

[k8s-triage-robot]

The Kubernetes project currently lacks enough active 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 rotten
• Close this issue with /close
• Offer to help out with Issue Triage

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

/lifecycle rotten

[k8s-triage-robot]

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

This bot triages 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:

• Reopen this issue with /reopen
• Mark this issue as fresh with /remove-lifecycle rotten
• Offer to help out with Issue Triage

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

/close not-planned

[k8s-ci-robot]

@k8s-triage-robot: Closing this issue, marking it as "Not Planned".

In response to this:

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

This bot triages 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:

• Reopen this issue with /reopen
• Mark this issue as fresh with /remove-lifecycle rotten
• Offer to help out with Issue Triage

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

/close not-planned

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository.