Fix flaky test 03800_assume_not_null_coalesce_if_null_monotonicity_key_condition by groeneai · Pull Request #101316 · ClickHouse/ClickHouse · GitHub
Skip to content

Fix flaky test 03800_assume_not_null_coalesce_if_null_monotonicity_key_condition#101316

Merged
nikitamikhaylov merged 1 commit into
ClickHouse:masterfrom
groeneai:fix-flaky-03800-assume-not-null-monotonicity
Mar 31, 2026
Merged

Fix flaky test 03800_assume_not_null_coalesce_if_null_monotonicity_key_condition#101316
nikitamikhaylov merged 1 commit into
ClickHouse:masterfrom
groeneai:fix-flaky-03800-assume-not-null-monotonicity

Conversation

@groeneai

@groeneai groeneai commented Mar 31, 2026 •
edited by robot-clickhouse
Loading

Copy link
Copy Markdown
Contributor

Changelog category (leave one):

  • CI Fix or Improvement (changelog entry is not required)

Changelog entry (a]notation for the changelog):

  • none

Summary

Pin query_plan_merge_expressions = 1 in the test to prevent EXPLAIN output mismatches when CI randomizes this setting to 0.

  • The test verifies that assumeNotNull/coalesce/ifNull monotonicity is correctly used by KeyCondition for MinMax index and primary key condition optimization
  • query_plan_merge_expressions only controls cosmetic merging of adjacent expression steps in the query plan — it does not affect index analysis or granule pruning
  • When randomized to 0, all 10+ EXPLAIN blocks produce structurally different output (unmerged expression nodes), breaking the reference comparison

CIDB evidence

13 failures across 6 unrelated PRs in 30 days (zero master failures — only triggered by CI randomization in PR runs):

Reproduction

Deterministic reproduction:

CLICKHOUSE_CLIENT_OPT="--query_plan_merge_expressions 0" \
  tests/clickhouse-test --no-random-settings --no-random-merge-tree-settings \
  03800_assume_not_null_coalesce_if_null_monotonicity_key_condition
# → FAIL every time without fix
# → OK every time with fix

Verified 50/50 passes with full randomization after fix.

Version info

  • Merged into: 26.4.1.457

@groeneai
Fix flaky test 03800_assume_not_null_coalesce_if_null_monotonicity_ke…
2f4923c 
…y_condition

Pin query_plan_merge_expressions=1 to prevent EXPLAIN output mismatch when CI
randomizes this setting to 0. The test verifies key condition monotonicity for
assumeNotNull/coalesce/ifNull — expression merging is cosmetic and irrelevant
to the tested optimization.

Reproduces deterministically with --query_plan_merge_expressions 0.
Verified 50/50 passes with full randomization after fix.
@groeneai

groeneai commented Mar 31, 2026

Copy link
Copy Markdown
Contributor Author

@groeneai

groeneai commented Mar 31, 2026

Copy link
Copy Markdown
Contributor Author

cc @CurtizJ — could you review this? Simple test fix — pins query_plan_merge_expressions=1 to prevent EXPLAIN output mismatch from CI randomization. The test checks assumeNotNull/coalesce/ifNull monotonicity in KeyCondition, which is unaffected by expression merging.

@nikitamikhaylov nikitamikhaylov added the can be tested Allows running workflows for external contributors label Mar 31, 2026
@clickhouse-gh

clickhouse-gh Bot commented Mar 31, 2026 •
edited
Loading

Copy link
Copy Markdown
Contributor

Workflow [PR], commit [2f4923c]

Summary: ⏳

job_name test_name status info comment
Integration tests (amd_llvm_coverage, 2/5) failure
test_overcommit_tracker/test.py::test_user_overcommit FAIL cidb
Stress test (arm_ubsan) failure
Hung check failed, possible deadlock found FAIL cidb, issue

AI Review

Summary

This PR stabilizes 03800_assume_not_null_coalesce_if_null_monotonicity_key_condition by pinning query_plan_merge_expressions = 1 in both .sql and .reference, preventing random-setting-induced EXPLAIN formatting drift. I found no correctness, safety, performance, or compatibility problems in the proposed change, and no duplicate inline concerns from clickhouse-gh[bot] to account for.

ClickHouse Rules
Item Status Notes
Deletion logging ➖
Serialization versioning ➖
Core-area scrutiny ✅
No test removal ✅
Experimental gate ➖
No magic constants ✅
Backward compatibility ✅
SettingsChangesHistory.cpp ➖
PR metadata quality ✅
Safe rollout ✅
Compilation time ✅
Final Verdict
  • Status: ✅ Approve

@clickhouse-gh clickhouse-gh Bot added the pr-ci label Mar 31, 2026
@nikitamikhaylov nikitamikhaylov self-assigned this Mar 31, 2026
@nikitamikhaylov nikitamikhaylov added this pull request to the merge queue Mar 31, 2026
Merged via the queue into ClickHouse:master with commit 3d69df4 Mar 31, 2026
145 of 162 checks passed
@robot-clickhouse-ci-2 robot-clickhouse-ci-2 added the pr-synced-to-cloud The PR is synced to the cloud repo label Mar 31, 2026
@rschu1ze rschu1ze mentioned this pull request Apr 2, 2026
Merged
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@nikitamikhaylov nikitamikhaylov nikitamikhaylov approved these changes

Assignees

@nikitamikhaylov nikitamikhaylov

Labels

can be tested Allows running workflows for external contributors pr-ci pr-synced-to-cloud The PR is synced to the cloud repo

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

3 participants

@groeneai @nikitamikhaylov @robot-clickhouse-ci-2