The efficiency of synchronizing Pod status too slow.

[FengyunPan2]

What would you like to be added?

Use a single List api instead of multiple Get api.

Why is this needed?

My k8s node is a big machine(500C CPU & 2000GB Memory), the maxPods of kubelet could be set to 200+.
When I delete 200+ pods or create 200+ pods, the synchronization of Pod status by the kubelet becomes too slow(about 230s).
Making a large number of GET API calls can indeed increase the load on the API server.

[k8s-ci-robot]

This issue is currently awaiting triage.

If a SIG or subproject determines this is a relevant issue, they will accept it by applying the triage/accepted label and provide further guidance.

The triage/accepted label can be added by org members by writing /triage accepted in a comment.

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/test-infra repository.

[FengyunPan2]

/assign

[neolit123]

/sig node

[ffromani]

do we have more precise numbers about the (in)efficiency? for example, a graph of pod count vs execution time

[matthyx]

do we have more precise numbers about the (in)efficiency? for example, a graph of pod count vs execution time

I agree with @ffromani it would be nice to see how the new patch improves things.