Performance benchmark is need

[lowang-bh]

What would you like to be added:

1. some benchmark test for scheduling performance.

Why is this needed:

1. supply basic benchmark test, and avoid future's change has a negative effect
2. more easy to observe and compare difference when do some performance improvement

[JamesBrianD]

Hello, I have also been trying to add performance benchmarks for Volcano. The initial plan is that, to first construct a benchmark for a single action in the form of unit tests, building clusters of different scale magnitudes (100 nodes 500 pods; 1000 nodes 5000 pods; 1000 nodes 100000 pods 1000 jobs…). This will test the performance of a single action under various scenarios. In the following step, we will use the actual scheduler operation as a benchmark to evaluate the performance of Volcano from the start of an open session to the end of a close session.

Could you share some suggestions for me?

[lowang-bh]

Yes, your idea sound good and is corresponding to my opinion. There is already a mockSchedulerCache and UTframework commit in https://github.com/volcano-sh/volcano/pull/3269. We can add a performanceTest Framework base on the mockScheduler as the UTFramework. And then do as you said, we can mock various scale pods/nodes to test each actions and even the full-scheduling process with mockScheduler's runOnce.

[stale[bot]]

Hello 👋 Looks like there was no activity on this issue for last 180 days. Do you mind updating us on the status? Is this still reproducible or needed? If yes, just comment on this PR or push a commit. Thanks! 🤗 If there will be no activity for 90 days, this issue will be closed (we can always reopen an issue if we need!).

[hwdef]

still need.

[tanjunchen]

still need.