need to add disk IO aware scheduling to avoid the PODs competing for disk IO resource

[cathyhongzhang]

Area

• Scheduler
• Controller
• Helm Chart
• Documents

Other components

No response

What happened?

The current Kubernetes scheduler only supports cpu, memory, and storage aware scheduling. It does not support disk IO resource-aware scheduling. It could happen that the pods scheduled onto a node compete for the disk IO resource, resulting in performance degradation (the noisy neighbor problem). There is an increasing demand for adding disk IO resource-aware scheduling to Kubernetes to avoid or mitigate the noisy neighbor problem.

A disk IO scheduler plugin can be developed to support disk IO resource-aware scheduling. The plugin would check the new pod's requested disk IO BW against each node's real-time available disk IO BW when making the scheduling decision.

What did you expect to happen?

A disk IO scheduler plugin KEP proposal and corresponding implementation

How can we reproduce it (as minimally and precisely as possible)?

No response

Anything else we need to know?

No response

Kubernetes version

$ kubectl version
# paste output here

Scheduler Plugins version

[zwpaper]

/remove-kind bug
/kind feature

[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

[cathyhongzhang]

A KEP has been proposed to address this issue. https://github.com/kubernetes-sigs/scheduler-plugins/pull/628

[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

[ichbinblau]

/remove-lifecycle rotten