kubernetes
Add feature gate for StorageNamespaceIndex in v1.30
[APPROVALNOTIFIER] This PR is NOT APPROVED
This pull-request has been approved by: Once this PR has been reviewed and has the lgtm label, please ask for approval from tengqm. For more information see the Kubernetes Code Review Process.
The full list of commands accepted by this bot can be found here.
Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment
Pull request preview available for checking
| Name | Link |
|---|---|
| Latest commit | 2d5cd443c531e778f0ebedc78d8c61c7a6141cdf |
| Latest deploy log | https://app.netlify.com/sites/kubernetes-io-main-staging/deploys/66a0e6e80852360008fd3ac8 |
| Deploy Preview | https://deploy-preview-47259--kubernetes-io-main-staging.netlify.app |
| Preview on mobile | Toggle QR Code...Use your smartphone camera to open QR code link. |
To edit notification comments on pull requests, go to your Netlify site configuration.
![netlify[bot] avatar](https://avatars.githubusercontent.com/in/13473?v=4 "Published by a pub.dev verified publisher"){kind=small}
- Searched for
StorageNamespaceIndexin Kubernetes code base - Tracked to this code PR 121906.
- In the comments, the decision was not to set up a KEP for this contribution, so there's none.
/sig api-machinery
cc: @ahutsunshine @kubernetes/sig-api-machinery-misc
Can we please get some help contributing some detail to this feature gate doc? We appreciate it, thanks!
- Searched for
StorageNamespaceIndexin Kubernetes code base- Tracked to this code PR 121906.
- In the comments, the decision was not to set up a KEP for this contribution, so there's none.
/sig api-machinery
cc: @ahutsunshine @kubernetes/sig-api-machinery-misc
Can we please get some help contributing some detail to this feature gate doc? We appreciate it, thanks!
@drewhagen I'm sorry for the late response. My account @ahutsunshine has some issues, so I register @ahutsunshine1 account to answer this question. Previously we already did a long time discussion and did a benchmark for the namespace indexer feature that shown excellent performance in terms of latency, CPU, and memory in large clusters (like with 100k+ pods) for listing pods within a single namespace with namespace indexers. This implementation significantly reduces cpu and memory consumption, especially when dealing with a small number of pods.
Discussion:
- https://github.com/kubernetes/kubernetes/issues/120778
Benchmark doc:
- https://docs.google.com/document/d/1bTxGqlH0tqrpH16NxC2SIUTGMN3_1zNKRiRdIW9Y5VU/edit?usp=sharing
We, eBay Cloud team, already applied the namespace indexer feature to our production and run good performance.
@ahutsunshine1: Reiterating the mentions to trigger a notification: @kubernetes/sig-api-machinery-misc
In response to this:
- Searched for
StorageNamespaceIndexin Kubernetes code base- Tracked to this code PR 121906.
- In the comments, the decision was not to set up a KEP for this contribution, so there's none.
/sig api-machinery
cc: @ahutsunshine @kubernetes/sig-api-machinery-misc
Can we please get some help contributing some detail to this feature gate doc? We appreciate it, thanks!
I'm sorry for late response. My account has some issues, so I register https://github.com/ahutsunshine1 account to answer this question. Previously we already did a long time discussion and did a benchmark for the namespace indexer feature that shown excellent performance in terms of latency, CPU, and memory in large clusters (like with 100k+ pods) for listing pods within a single namespace with namespace indexers. This implementation significantly reduces cpu and memory consumption, especially when dealing with a small number of pods.
Discussion:
- https://github.com/kubernetes/kubernetes/issues/120778
Benchmark doc:
- https://docs.google.com/document/d/1bTxGqlH0tqrpH16NxC2SIUTGMN3_1zNKRiRdIW9Y5VU/edit?usp=sharing
We, eBay Cloud team, already applied the namespace indexer feature to our production and run good performance.
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.
Can you explain what enabling the feature gate means @ahutsunshine1 ?
You're explaining this to an audience that has never looked at one line of Kubernetes code, doesn't know Go syntax, and who aren't familiar with the performance challenges from not applying the fix.
Can you explain what enabling the feature gate means @ahutsunshine1 ?
You're explaining this to an audience that has never looked at one line of Kubernetes code, doesn't know Go syntax, and who aren't familiar with the performance challenges from not applying the fix.
@sftim After enabling the feature gate, it benefits the LIST namespace-scoped pods like /api/v1/namespaces/default/pods?resourceVersion=0. Simply:
- Before enabling,
/api/v1/namespaces/default/pods?resourceVersion=0will search through all the pods from all namespaces in the cache and then filter the results based on the namespacedefault. For large clusters with more than 100k+ pods, it will spend lots of cpu and memory with the similar high LIST pods requests even the pod count indefaultnamespace may be small. - After enabling,
/api/v1/namespaces/default/pods?resourceVersion=0will get pods fromdefaultnamespace indexer directly. It avoids too many unnecessary filter that reduces the cpu and memory usage.
If you want to know more, I would like to recommend to read the benchmark doc with the detail https://docs.google.com/document/d/1bTxGqlH0tqrpH16NxC2SIUTGMN3_1zNKRiRdIW9Y5VU/edit. I'm not sure if I explain properly. Feel free to leave more comments.
The Kubernetes project currently lacks enough contributors to adequately respond to all PRs.
This bot triages PRs according to the following rules:
- After 90d of inactivity,
lifecycle/staleis applied - After 30d of inactivity since
lifecycle/stalewas applied,lifecycle/rottenis applied - After 30d of inactivity since
lifecycle/rottenwas applied, the PR is closed
You can:
- Mark this PR as fresh with
/remove-lifecycle stale - Close this PR with
/close - Offer to help out with Issue Triage
Please send feedback to sig-contributor-experience at kubernetes/community.
/lifecycle stale
