executor: skip analyze flush broadcast in test binaries

[0xPoe]

What problem does this PR solve?

Issue Number: close #68416

Problem Summary:
make bench-daily (CI job gobench4) fails on BenchmarkNonPartitionPointGetPlanCacheOn: analyze table t hits context deadline exceeded after 10s. PR #67939 added flushStatsDeltaForAnalyze, which broadcasts FLUSH STATS_DELTA CLUSTER via a TiDB-type cop request before each ANALYZE. The local-dump fallback was gated on intest.InTest, but that variable is only set under the intest build tag. make bench-daily runs go test without the tag, so the broadcast hits the mockstore session's empty :10080 and hangs.

What changed and how does it work?

Replace the gate with a runningUnderGoTest helper backed by testing.Testing(). The Go linker sets testing.Testing() to true for any binary produced by go test (covering both unit tests and make bench-daily) and leaves it false for go build of cmd/tidb-server. The gate is therefore a no-op in production regardless of host or flag configuration, and no longer depends on the intest build tag.

Check List

Tests

• Unit test
• Integration test
• Manual test (add detailed scripts or steps below)
• No need to test

  • I checked and no code files have been changed.

Reproduction and verification (same machine, same command):

#68417 (comment)

Side effects

• Performance regression: Consumes more CPU
• Performance regression: Consumes more Memory
• Breaking backward compatibility

Documentation

• Affects user behaviors
• Contains syntax changes
• Contains variable changes
• Contains experimental features
• Changes MySQL compatibility

Release note

None

Summary by CodeRabbit

• Chores
  • Improved test execution efficiency by optimizing internal statistics handling during testing. Test processes now use a more efficient local flushing approach instead of cluster-level broadcasts, reducing unnecessary background operations during test runs.

[pantheon-ai]

@0xPoe I've received your pull request and will start the review. I'll conduct a thorough review covering code quality, potential issues, and implementation details.

⏳ This process typically takes 10-30 minutes depending on the complexity of the changes.

_{ℹ️ Learn more details on Pantheon AI.}

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro

Run ID: 9afd1f7e-bb6f-48c7-9b77-d9e0cd875c12

📥 Commits

Reviewing files that changed from the base of the PR and between dcea072 and 46c8ff8.

📒 Files selected for processing (1)

• pkg/executor/analyze.go

🚧 Files skipped from review as they are similar to previous changes (1)

• pkg/executor/analyze.go

---

📝 Walkthrough

Walkthrough

Replace the intest.InTest build-tag gate with a runtime runningUnderGoTest() helper (uses testing.Testing()) so flushStatsDeltaForAnalyze flushes stats deltas locally when running under Go test binaries.

Changes

Analyze stats delta flushing topology detection

Layer / File(s)	Summary
In-process test detection with testing.Testing() pkg/executor/analyze.go	Add testing package import, introduce runningUnderGoTest() helper that returns testing.Testing() and documents in-process test domains lacking a real TiDB RPC listener, and update flushStatsDeltaForAnalyze to use this helper instead of intest.InTest.

Sequence Diagram

sequenceDiagram
  participant Analyze
  participant runningUnderGoTest
  participant FlushAction
  Analyze->>runningUnderGoTest: call testing.Testing()
  runningUnderGoTest-->>Analyze: bool result
  Analyze->>FlushAction: [true] local dump
  Analyze->>FlushAction: [false] cluster FLUSH STATS_DELTA

Loading

Estimated code review effort

🎯 3 (Moderate) | ⏱️ ~20 minutes

Possibly related PRs

• pingcap/tidb#67939: Introduced the flushStatsDeltaForAnalyze function with the original intest.InTest gate; this PR fixes its topology detection for go test execution without build tags.

Suggested labels

sig/planner, component/statistics, ok-to-test

Suggested reviewers

• henrybw
• qw4990
• terry1purcell

Poem

🐰 I sniff the test-time breeze,
No build tag hides the tease,
testing.Testing() lights the way,
Local flush keeps broadcasts at bay,
Hooray — no mockstore delays! 🎉

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 40.00% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title 'executor: skip analyze flush broadcast in test binaries' clearly and concisely describes the main change: modifying executor behavior to skip the broadcast in test binaries.
Description check	✅ Passed	The PR description includes the required issue reference (close #68416), explains the problem and solution, and documents manual test verification.
Linked Issues check	✅ Passed	The changes fully address issue #68416 by replacing the intest.InTest gate with runningUnderGoTest() backed by testing.Testing(), which correctly detects Go test binaries and prevents broadcast hangs.
Out of Scope Changes check	✅ Passed	All changes are within scope of fixing the benchmark timeout by improving test detection for conditional stats-delta flushing; no unrelated modifications are present.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[codecov]

Codecov Report

❌ Patch coverage is 66.66667% with 1 line in your changes missing coverage. Please review.
✅ Project coverage is 75.6786%. Comparing base (8c17ce1) to head (46c8ff8).
⚠️ Report is 4 commits behind head on master.

Additional details and impacted files

@@               Coverage Diff                @@
##             master     #68417        +/-   ##
================================================
- Coverage   77.2761%   75.6786%   -1.5975%     
================================================
  Files          2010       2008         -2     
  Lines        555473     564521      +9048     
================================================
- Hits         429248     427222      -2026     
- Misses       125305     137195     +11890     
+ Partials        920        104       -816     

Flag	Coverage Δ
integration	41.1463% <66.6666%> (+1.3522%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Components	Coverage Δ
dumpling	60.4679% <ø> (ø)
parser	∅ <ø> (∅)
br	49.9725% <ø> (-13.0354%)	⬇️

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[coderabbitai]

🧹 Nitpick comments (1)

pkg/executor/analyze_utils_test.go (1)

74-81: ⚡ Quick win

Avoid replacing the entire global config in this test.

Using config.NewConfig() and StoreGlobalConfig(...) here swaps all global config fields, which increases cross-test coupling risk. Prefer preserving the existing config and changing only AdvertiseAddress for each case.

As per coding guidelines, keep test changes minimal and deterministic; avoid broad churn unless required.

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@pkg/executor/analyze_utils_test.go` around lines 74 - 81, The test currently
replaces the entire global config using config.NewConfig() +
config.StoreGlobalConfig(...), which is risky; instead, read the existing global
config, copy it and only change the AdvertiseAddress field for the two cases,
call config.StoreGlobalConfig with that modified copy, and ensure you restore
the original config with defer after the test; update the instances where
AdvertiseAddress is set (referencing AdvertiseAddress, config.StoreGlobalConfig,
and isInProcessTopology) so only AdvertiseAddress is mutated and the rest of the
global config remains unchanged.

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Nitpick comments:
In `@pkg/executor/analyze_utils_test.go`:
- Around line 74-81: The test currently replaces the entire global config using
config.NewConfig() + config.StoreGlobalConfig(...), which is risky; instead,
read the existing global config, copy it and only change the AdvertiseAddress
field for the two cases, call config.StoreGlobalConfig with that modified copy,
and ensure you restore the original config with defer after the test; update the
instances where AdvertiseAddress is set (referencing AdvertiseAddress,
config.StoreGlobalConfig, and isInProcessTopology) so only AdvertiseAddress is
mutated and the rest of the global config remains unchanged.

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro

Run ID: 2a92f652-e0af-42b3-a403-759f857dd18b

📥 Commits

Reviewing files that changed from the base of the PR and between 7510a45 and 0c2cbfd.

📒 Files selected for processing (3)

• pkg/executor/BUILD.bazel
• pkg/executor/analyze.go
• pkg/executor/analyze_utils_test.go

🚧 Files skipped from review as they are similar to previous changes (1)

• pkg/executor/analyze.go

[0xPoe]

Tested locally:

go test -run '^$' -bench '^BenchmarkNonPartitionPointGetPlanCacheOn$' \
        -benchtime 1x -timeout 60s ./pkg/planner/core/tests/partition

Before

make bench-daily builds without the intest tag, so the gate stays closed, the broadcast targets the mockstore session's empty :10080, and the analyze setup hits the 10s context deadline:

[FATAL] mustExecute error sql="analyze table t" error="context deadline exceeded"
  flushStatsDeltaForAnalyze   pkg/executor/analyze.go:128
  broadcast                   pkg/executor/simple.go:3034
  (*copIterator).Next         pkg/store/copr/coprocessor.go:1445
  preparePointGet             pkg/planner/core/tests/partition/bench_test.go:200
  BenchmarkNonPartitionPointGetPlanCacheOn   pkg/planner/core/tests/partition/bench_test.go:313

exit status 1
FAIL    github.com/pingcap/tidb/pkg/planner/core/tests/partition    12.414s

After

BenchmarkNonPartitionPointGetPlanCacheOn-14        1    202458 ns/op
PASS
ok  github.com/pingcap/tidb/pkg/planner/core/tests/partition    2.333s

[0xPoe]

🔢 Self-check (PR reviewed by myself and ready for feedback)

• Code compiles successfully

• Unit tests added

• No AI-generated elegant nonsense in PR.

• Comments added where necessary

• PR title and description updated

• Documentation PR created (or confirmed not needed)

• PR size is reasonable

/cc @terry1purcell @mjonss

[mjonss]

LGTM

[ti-chi-bot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: mjonss, terry1purcell

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Details

Needs approval from an approver in each of these files:

• pkg/executor/OWNERS [mjonss,terry1purcell]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[ti-chi-bot]

[LGTM Timeline notifier]

Timeline:

• 2026-05-18 12:14:24.789139781 +0000 UTC m=+179994.293270457: ☑️ agreed by mjonss.
• 2026-05-18 14:39:43.358006949 +0000 UTC m=+188712.862137625: ☑️ agreed by terry1purcell.

[0xPoe]

/retest

[ti-chi-bot]

In response to a cherrypick label: new pull request created to branch release-8.5: #68471.

[0xPoe]

Thanks for your review! 💚 💙 💜 💛 ❤️