Add another tablet load balancer algorithm: random

[nickvanw]

Description

This PR implements #18786 and adds a cell-agnostic "random" tablet load balancer algorithm.

At PlanetScale, we often find that traffic into VTGates is very uneven - either because of individual applications that might be single-AZ, a lack of AZ overlap, or a specific piece of infrastructure is singular, but produces a lot of load on replicas.

When this happens, it ends up that an individual cell's worth of VTTablets can get very overloaded, causing humans to hae to step in and manually scale components, etc. If the particular overloaded cell changes, it requires continued manual access.

Many times, we'd rather pay for the cross-zone latency and bandwidth costs and just spread it evenly across every replica. That is what this PR implements.

Related Issue(s)

RFC: #18786

Checklist

• "Backport to:" labels have been added if this change should be back-ported to release branches
• If this change is to be back-ported to previous releases, a justification is included in the PR description
• Tests were added or are not required
• Did the new or modified tests pass consistently locally and on CI?
• Documentation was added or is not required

AI Disclosure

This PR was written by a human and Claude Code - all changes were reviewed by a human.

[vitess-bot]

Review Checklist

Hello reviewers! 👋 Please follow this checklist when reviewing this Pull Request.

General

• Ensure that the Pull Request has a descriptive title.
• Ensure there is a link to an issue (except for internal cleanup and flaky test fixes), new features should have an RFC that documents use cases and test cases.

Tests

• Bug fixes should have at least one unit or end-to-end test, enhancement and new features should have a sufficient number of tests.

Documentation

• Apply the release notes (needs details) label if users need to know about this change.
• New features should be documented.
• There should be some code comments as to why things are implemented the way they are.
• There should be a comment at the top of each new or modified test to explain what the test does.

New flags

• Is this flag really necessary?
• Flag names must be clear and intuitive, use dashes (-), and have a clear help text.

If a workflow is added or modified:

• Each item in Jobs should be named in order to mark it as required.
• If the workflow needs to be marked as required, the maintainer team must be notified.

Backward compatibility

• Protobuf changes should be wire-compatible.
• Changes to _vt tables and RPCs need to be backward compatible.
• RPC changes should be compatible with vitess-operator
• If a flag is removed, then it should also be removed from vitess-operator and arewefastyet, if used there.
• vtctl command output order should be stable and awk-able.

[arthurschreiber]

Neat! ❤️

[codecov]

Codecov Report

❌ Patch coverage is 67.56757% with 36 lines in your changes missing coverage. Please review.
✅ Project coverage is 69.72%. Comparing base (dbcfdd9) to head (b142edb).
⚠️ Report is 1 commits behind head on main.

Files with missing lines	Patch %	Lines
go/vt/vtgate/balancer/random_balancer.go	65.85%	14 Missing ⚠️
go/vt/vtgate/balancer/balancer.go	71.79%	11 Missing ⚠️
go/vt/vtgate/tabletgateway.go	66.66%	10 Missing ⚠️
go/vt/vtgate/vtgate.go	0.00%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main   #18787      +/-   ##
==========================================
+ Coverage   69.70%   69.72%   +0.01%     
==========================================
  Files        1607     1608       +1     
  Lines      214608   214698      +90     
==========================================
+ Hits       149592   149689      +97     
+ Misses      65016    65009       -7     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[nickvanw]

I noticed the previous strategy wasn't covered by end to end tests, so I'm going to write up a suite of them that should cover everything.

[mattlord]

This is looking good! Just some minor comments. Happy to chat about them. I can then go through the tests.

[mattlord]

LGTM! My only issues are with the tests as I think they will prove to be flaky over time. Let me know what you think about my comments so far, as I think they will apply throughout the others as well.

[mattlord]

I would recommend using topodata.TabletType_REPLICA.String() and similar throughout.

[nickvanw]

It looks like there's a caps vs lowercase mixup, so I added a test-level variable that should work

[nickvanw]

@mattlord thanks for the reviews! I've cleaned up the test suite significantly - there's no more time.Sleep present anywhere, and many things are broken out into helper methods so it should be straightforward to see what we're doing.

I think I've addressed all of the naming feedback as well, but let me know if you think there's anything else missing!

[mattlord]

Nice work on this, @nickvanw ! I had a couple of very minor comments that you can deal with as you choose.

[nickvanw]

Thanks for the further feedback @mattlord! All accepted in 3246c08 😄