[WIP] kubelet: Improved performance of the cadvisor metric collection

[bwplotka]

Fixed #107960 PR.

As in the previous PR, the vendor is 80% of this PR. Life is hard 🙃

Most of the work was done by @ehashman

/cc @ehashman @dgrisonnet

What type of PR is this?

/kind feature
/sig node
/sig instrumentation

What this PR does / why we need it:

Significantly improves kubelet perf of cadvisor metric collection on scrape (e.g by Prometheus).

Which issue(s) this PR fixes:

Fixes #

Special notes for your reviewer:

Two PRs has to be merged upstream, before moving with this PR:

• Added TransactionalGatherer with Cached implementation (+benchmark) prometheus/client_golang#929
• Optimize PrometheusCollector scrape efficiency. google/cadvisor#2974

Also:

• Macro benchmarks are in progress

Does this PR introduce a user-facing change?

cAdvisor uses a new type of registry for container metric collection. Users should see significant improvements in kubelet CPU utilization as a result.

Additional documentation e.g., KEPs (Kubernetes Enhancement Proposals), usage docs, etc.:

[k8s-ci-robot]

@bwplotka: Adding the "do-not-merge/release-note-label-needed" label because no release-note block was detected, please follow our release note process to remove it.

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.

[k8s-ci-robot]

@bwplotka: GitHub didn't allow me to request PR reviews from the following users: bwplotka.

Note that only kubernetes members and repo collaborators can review this PR, and authors cannot review their own PRs.

In response to this:

Fixed #107960 PR.

As in the previous PR, the vendor is 80% of this PR. Life is hard 🙃

Most of the work was done by @ehashman

/cc @ehashman @bwplotka

What type of PR is this?

/kind feature
/sig node
/sig instrumentation

What this PR does / why we need it:

Significantly improves kubelet perf of cadvisor metric collection on scrape (e.g by Prometheus).

Which issue(s) this PR fixes:

Fixes #

Special notes for your reviewer:

Two PRs has to be merged upstream, before moving with this PR:

• Added TransactionalGatherer with Cached implementation (+benchmark) prometheus/client_golang#929
• Optimize PrometheusCollector scrape efficiency. google/cadvisor#2974

Also:

• Macro benchmarks are in progress

Does this PR introduce a user-facing change?

cAdvisor uses a new type of registry for container metric collection. Users should see significant improvements in kubelet CPU utilization as a result.

Additional documentation e.g., KEPs (Kubernetes Enhancement Proposals), usage docs, etc.:

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.

[k8s-ci-robot]

Welcome @bwplotka!

It looks like this is your first PR to kubernetes/kubernetes 🎉. Please refer to our pull request process documentation to help your PR have a smooth ride to approval.

You will be prompted by a bot to use commands during the review process. Do not be afraid to follow the prompts! It is okay to experiment. Here is the bot commands documentation.

You can also check if kubernetes/kubernetes has its own contribution guidelines.

You may want to refer to our testing guide if you run into trouble with your tests not passing.

If you are having difficulty getting your pull request seen, please follow the recommended escalation practices. Also, for tips and tricks in the contribution process you may want to read the Kubernetes contributor cheat sheet. We want to make sure your contribution gets all the attention it needs!

Thank you, and welcome to Kubernetes. 😃

[k8s-ci-robot]

Hi @bwplotka. Thanks for your PR.

I'm waiting for a kubernetes member to verify that this patch is reasonable to test. If it is, they should reply with /ok-to-test on its own line. Until that is done, I will not automatically test new commits in this PR, but the usual testing commands by org members will still work. Regular contributors should join the org to skip this step.

Once the patch is verified, the new status will be reflected by the ok-to-test label.

I understand the commands that are listed here.

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.

[k8s-ci-robot]

[APPROVALNOTIFIER] This PR is NOT APPROVED

This pull-request has been approved by: bwplotka
To complete the pull request process, please assign liggitt after the PR has been reviewed.
You can assign the PR to them by writing /assign @liggitt in a comment when ready.

The full list of commands accepted by this bot can be found here.

Needs approval from an approver in each of these files:

• OWNERS
• staging/src/k8s.io/legacy-cloud-providers/OWNERS

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[dims]

/ok-to-test

[k8s-ci-robot]

@bwplotka: The following tests failed, say /retest to rerun all failed tests or /retest-required to rerun all mandatory failed tests:

Test name	Commit	Details	Required	Rerun command
pull-kubernetes-dependencies	740f2f1	link	true	/test pull-kubernetes-dependencies
pull-kubernetes-unit	740f2f1	link	true	/test pull-kubernetes-unit

Full PR test history. Your PR dashboard. Please help us cut down on flakes by linking to an open issue when you hit one in your PR.

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. I understand the commands that are listed here.

[k8s-ci-robot]

@bwplotka: PR needs rebase.

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.

[fedebongio]

/triage accepted

[bwplotka]

Current status:

• Almost all functional problems fixed: one more to go.
• We see 50% of past CPU usage and 10% of past memory allocations. Yet 2x of memory is being kept in heap, so utilization is overall higher. Investigating further.
• Needs rebase

[dashpole]

/assign @ehashman @dgrisonnet

[bwplotka]

Status update:

• After Dep bump to runc 1.1.0, cadvisor 0.44.0 #109029 this PR will be much smaller and reasoning about it will be much easier.
• There are two caches in cadvisor with this change. The internal cgroup cache and client-go cache. The second cache allows us to have reduced CPU and allocation, but it keeps extra bytes on top of the internal cgroup cache, that's why we see more bytes in memory being kept. It is unsure if we can get rid of the internal cgroup, since it is not only used by cgroup prometheus metric export, but also for ad-hoc kubelet operations like handling summary (GetCPUAndMemoryStatS()). This is why removing internal cache might have unpredictiable consequences (for us, Prometheus client-golang experts, not cadvisor ones). We are looking with @saswatamcode (in our very limited free time) if we are not missing anything else here, if there are means of easy improvements of this situation without rewriting cadvisor OR alternatives to caching logic we added that solves Steady state kubelet CPU usage is high due to excessive allocation in prometheus scraping of cadvisor #104459.

If anyone wants to dig into benchmarked profiles I obtained thanks to https://github.com/bwplotka/kubelet-bench setup:

• Baseline heap profile (23.55 MB) vs This PR heap profile (53 MB)
• Baseline CPU profile (6.1 CPU seconds) vs This PR CPU profile (2.6 CPU seconds),

[k8s-triage-robot]

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

This bot triages issues and PRs 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 or PR as fresh with /remove-lifecycle stale
• Mark this issue or PR as rotten with /lifecycle rotten
• Close this issue or PR with /close
• Offer to help out with Issue Triage

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

/lifecycle stale

[enj]

/remove-sig auth

This appears to belong to SIG Node. Feel free to add back SIG Auth if needed.

[k8s-triage-robot]

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

This bot triages issues and PRs 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 or PR as fresh with /remove-lifecycle rotten
• Close this issue or PR 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 and PRs 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 or PR with /reopen
• Mark this issue or PR as fresh with /remove-lifecycle rotten
• Offer to help out with Issue Triage

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

/close

[k8s-ci-robot]

@k8s-triage-robot: Closed this PR.

In response to this:

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

This bot triages issues and PRs 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 or PR with /reopen
• Mark this issue or PR as fresh with /remove-lifecycle rotten
• Offer to help out with Issue Triage

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

/close

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.