benchmark: TwoPC commit mode

[harshit-gangal]

Description

This PR benchmarks two Commit Modes:

1. Best Effort Commit (multi) – commits sequentially across shards.
2. Atomic Commit (TwoPC) – ensures atomicity through a multi-step process with parallel phases.

The benchmarks measure performance across transactions involving varying numbers of shards.
Results are provided for 4, 8, and 16 shard configurations.

4 shards

                          │ multi_v4.txt │             twopc_v4.txt             │
                          │    sec/op    │    sec/op     vs base                │
TwoPCCommit/SingleShard-8    756.5µ ± 1%   764.6µ ±  2%         ~ (p=0.180 n=6)
TwoPCCommit/TwoShards-8      1.545m ± 4%   3.386m ±  4%  +119.08% (p=0.002 n=6)
TwoPCCommit/ThreeShards-8    2.393m ± 2%   4.877m ± 14%  +103.82% (p=0.002 n=6)
TwoPCCommit/FourShards-8     3.309m ± 4%   6.098m ±  6%   +84.29% (p=0.002 n=6)
geomean                      1.744m        2.962m         +69.82%

8 shards

                          │ multi_8shards.txt │          twopc_8shards.txt           │
                          │      sec/op       │    sec/op     vs base                │
TwoPCCommit/SingleShard-8        750.3µ ±  1%    754.0µ ± 1%         ~ (p=0.818 n=6)
TwoPCCommit/TwoShards-8          1.528m ±  8%    3.448m ± 4%  +125.63% (p=0.002 n=6)
TwoPCCommit/FourShards-8         3.269m ±  3%    6.149m ± 5%   +88.10% (p=0.002 n=6)
TwoPCCommit/SixShards-8          5.176m ±  6%    8.577m ± 7%   +65.71% (p=0.002 n=6)
TwoPCCommit/EightShards-8        7.140m ± 11%   10.968m ± 5%   +53.62% (p=0.002 n=6)
geomean                          2.681m          4.320m        +61.12%

16 shards

                            │ multi_16shards.txt │          twopc_16shards.txt          │
                            │       sec/op       │    sec/op      vs base               │
TwoPCCommit/EightShards-8           7.611m ±  6%   11.261m ± 11%  +47.97% (p=0.002 n=6)
TwoPCCommit/TwelveShard-8           12.19m ± 10%    17.21m ± 11%  +41.18% (p=0.002 n=6)
TwoPCCommit/SixteenShards-8         17.35m ± 15%    24.06m ± 10%  +38.69% (p=0.002 n=6)
geomean                             11.72m          16.71m        +42.56%

Summary

• Best Effort Commit scales poorly as the number of shards increases due to its sequential execution of commits.
• Atomic Commit (TwoPC) incurs higher overhead but offers atomicity guarantees. Its parallel phases help reduce the performance gap as the shard count grows.

Related Issue(s)

• Closes RFC: Atomic Distributed Transactions #16245

Checklist

• 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

Deployment Notes

[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

All modified and coverable lines are covered by tests ✅

Project coverage is 67.51%. Comparing base (45192d2) to head (70d90f5).
Report is 6 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main   #17397      +/-   ##
==========================================
- Coverage   67.52%   67.51%   -0.01%     
==========================================
  Files        1581     1581              
  Lines      253948   253945       -3     
==========================================
- Hits       171480   171456      -24     
- Misses      82468    82489      +21     

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