Performance testing in PRs

[rikatz]

We need a conformance test that can provide us information like “amount of requests, time taken, errors” for our PRs so we can know how improved or degraded our performance was.

Imo there’s no need to know how to code here but mostly we need to establish:
set of important tests (like single http, long running http, tls, 100 simultaneous connections, etc)

• We don’t need to run all the e2e and corner cases here, but we need to run tests for scenarios that impacts 95% of our use cases
• Think in a way to test and export those metrics (like a gh action that runs, exports the metrics and post them somewhere we can consume later) we probably want to use prow for this
• implement this test as “optional” (non blocking) but in a way we can have a performance history of ingress

Who’s up to doing that? Required knowledge/opportunity to learn:

• learn how to test an http server
• Learn how to export metrics (prom gateway? Json? Google docs? :P)
• Optionally prow if we run this in test-infra

Suggested approach:

Using k6 + https://k6.io/blog/load-testing-using-github-actions/

/help
/triage accepted
/priority important-soon
/kind feature

[tao12345666333]

It doesn't look complicated. However, we need consider the results reporting and persistence.

Maybe test-infra already has some feasible solutions.

[longwuyuan]

I am interested but i commented in thread on slack that I have some idea about https://locust.io but no idea of k6.io or js (k6.io uses js)

[tao12345666333]

I tried k6 and it didn't seem to be difficult. Need to design a suitable test plan.

[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

[rikatz]

/remove-lifecycle stale
News on this folks?

[rikatz]

/lifecycle frozen
/assign @tao12345666333
/assign @longwuyuan

[longwuyuan]

/remove-help

[longwuyuan]

k6 implemented by me and used by @strongjz already

[longwuyuan]

/close

[k8s-ci-robot]

@longwuyuan: Closing this issue.

In response to this:

/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.