Faster Prepared Statement Execution by Using Raw SQL for Caching

[harshit-gangal]

Description

This PR changes the way Vitess handles prepared statements by modifying the plan cache lookup mechanism. Instead of normalizing and parsing the SQL string before looking up a cached execution plan, we now use the raw SQL string as provided in the PREPARE statement. This eliminates unnecessary parsing, improves lookup efficiency, and reduces overhead in query execution.

Problem Statement

Currently, when a client issues a PREPARE statement, vtgate:

1. Parses the SQL query.
2. Normalizes it to a canonical form.
3. Uses the normalized query string to check for an existing cached execution plan.

What to do About it

Instead of parsing and normalizing the query before checking the plan cache, we will:
• Directly store the original SQL string from the PREPARE statement in the cache.
• Use this unmodified SQL string for cache lookups.
• Only parse the query when the plan is not found in the cache.

Benchmark

goos: darwin
goarch: arm64
pkg: vitess.io/vitess/go/test/endtoend/preparestmt
cpu: Apple M1 Max
                                        │ pfp-main.txt │            pfp-new.txt            │
                                        │    sec/op    │   sec/op     vs base              │
PreparedStmt/Insert-4                      494.3µ ± 3%   482.4µ ± 1%  -2.41% (p=0.002 n=6)
PreparedStmt/Select-4                      301.6µ ± 2%   288.4µ ± 3%  -4.35% (p=0.004 n=6)
PreparedStmt/Join_Select:Simple_Route-4    292.8µ ± 1%   269.9µ ± 1%  -7.81% (p=0.002 n=6)
PreparedStmt/Update-4                      344.9µ ± 2%   336.1µ ± 2%  -2.54% (p=0.009 n=6)
PreparedStmt/Delete-4                      311.4µ ± 0%   302.6µ ± 1%  -2.80% (p=0.002 n=6)
geomean                                    342.1µ        328.4µ       -4.01%

Related Issue(s)

• Closes Prepare statement execution optimization #17621

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

[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 83.56164% with 84 lines in your changes missing coverage. Please review.

Project coverage is 67.44%. Comparing base (b05df12) to head (7949d15).

Files with missing lines	Patch %	Lines
go/vt/vtgate/executor.go	93.23%	9 Missing ⚠️
go/vt/vtgate/planbuilder/plan_test_vindex.go	0.00%	8 Missing ⚠️
go/vt/vtgate/fakerpcvtgateconn/conn.go	0.00%	7 Missing ⚠️
go/cmd/vtgateclienttest/services/errors.go	0.00%	6 Missing ⚠️
go/mysql/conn.go	57.14%	6 Missing ⚠️
go/vt/vtgate/plugin_mysql_server.go	0.00%	6 Missing ⚠️
go/cmd/vtgateclienttest/services/echo.go	0.00%	5 Missing ⚠️
go/vt/sqlparser/analyzer.go	37.50%	5 Missing ⚠️
go/cmd/vtgateclienttest/services/fallback.go	0.00%	4 Missing ⚠️
go/vt/vtgate/executorcontext/vcursor_impl.go	89.47%	4 Missing ⚠️
... and 15 more

Additional details and impacted files

@@            Coverage Diff             @@
##             main   #17777      +/-   ##
==========================================
- Coverage   67.45%   67.44%   -0.01%     
==========================================
  Files        1594     1594              
  Lines      259064   259042      -22     
==========================================
- Hits       174760   174721      -39     
- Misses      84304    84321      +17     

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

[vitess-bot]

Hello! 👋

This Pull Request is now handled by arewefastyet. The current HEAD and future commits will be benchmarked.

You can find the performance comparison on the arewefastyet website.