Ensure all topo read calls consider --topo_read_concurrency

[timvaillancourt]

Description

This PR implements the feature request #17275 by ensuring that all topo read calls respect the --topo_read_concurrency flag, specifically Get, GetVersion, List, ListDir operations

As discussed in #17275, there is a separate semaphore/limit for "global" and "cell" reads, ie: by default there can be up to 32 concurrent "global" topo reads and 32 concurrent reads per "local" cell

Related Issue(s)

• Fixes: Feature Request: ensure topo calls use --topo_read_concurrency #17275
• Sub-task of Tracking Issue: Tracking issue: improve VTOrc topo+backend efficiency #17330

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

[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.

[codecov]

Codecov Report

Attention: Patch coverage is 73.17073% with 11 lines in your changes missing coverage. Please review.

Project coverage is 67.53%. Comparing base (63dfb9e) to head (91e9640).
Report is 15 commits behind head on main.

Files with missing lines	Patch %	Lines
go/vt/topo/stats_conn.go	61.53%	10 Missing ⚠️
go/vt/vtorc/logic/tablet_discovery.go	0.00%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main   #17276      +/-   ##
==========================================
+ Coverage   67.51%   67.53%   +0.02%     
==========================================
  Files        1580     1581       +1     
  Lines      253805   253945     +140     
==========================================
+ Hits       171358   171506     +148     
+ Misses      82447    82439       -8     

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

[timvaillancourt]

Moved + renamed from DefaultConcurrency from keyspace.go

DefaultReadConcurrency seems more accurate as this only applies to reads

[deepthi]

The only concern I have is what I said on the linked issue #17275 (comment)
Otherwise the implementation looks fine to me.

[timvaillancourt]

The only concern I have is what I said on the linked issue #17275 (comment) Otherwise the implementation looks fine to me.

@deepthi 👋, makes sense

I'll re-work this so all cells (global or local) get a limit equal to --topo_read_concurrency, no sharing of limits. I think the single flag will mean the global cell with have a limit that is much higher than required, but maybe that's ok for now

[timvaillancourt]

@deepthi updated to use a limit per global and local cell. I also added this a minor change to the v22 changelog 👍

[timvaillancourt]

👋 @deepthi I think I'm done tweaking this PR now

We've backported this to our v19 branch and tested it in a low-volume environment - no issues noticed. Testing on a prod-like environment is to come if we'd like to wait for that

[mattlord]

Nice work on this @timvaillancourt !

The only non-nit that I had was related to the function timing stats. Let me know what you think. I can come back to this quickly and we can get it merged.

Thanks!

[GuptaManan100]

LGTM apart from the outstanding comments

[timvaillancourt]

@mattlord thanks for the review. Responded to some comments and addressed @GuptaManan100's point

I think the outstanding topic is which approach to the operation vs. semaphore-wait stats timings. Thoughts appreciated 🙇

[mattlord]

Using request changes just to be sure that we resolve the stats questions before merging.

[timvaillancourt]

I still think it's probably worth adding something to an existing unit test to confirm the new stats, but it looks good to me! Thanks again!

@mattlord no problem!

Yeah with all those defers crammed together it's easy to mis-read. I usually would add empty lines to separate "concepts" like this, but on big codebases I tend to be a chameleon and copy the existing style like I did here

I still think it's probably worth adding something to an existing unit test to confirm the new stats, but it looks good to me! Thanks again!

I agree, this is some important code. Adding some tests