[fix](be) Fix timestamptz group_array state serde#63827
Merged
Conversation
Contributor
Author
|
/review |
![github-actions[bot]](https://avatars.githubusercontent.com/in/15368?s=60&v=4){kind=small}
Contributor
There was a problem hiding this comment.
I found one regression-test standards issue that should be fixed before merge.
Critical checkpoint conclusions:
- Goal/test coverage: The code change addresses aggregate-state serialization for nested TIMESTAMPTZ and adds a regression case that exercises
group_arrayoverARRAY<TIMESTAMPTZ(6)>. - Scope/focus: The actual PR diff is small and focused.
- Concurrency: No concurrency-sensitive code path was changed.
- Lifecycle: No static/global lifecycle issue was found.
- Config items: No config item was added or changed.
- Compatibility: Aggregate state serialization remains paired write/read logic; no persistent format change was identified in this patch.
- Parallel paths: The change applies to the complex-type collect-list serialization path used by the added nested TIMESTAMPTZ test.
- Conditional checks: No problematic new conditional check was found.
- Test coverage/results: The new test covers the crash path, but it violates the regression-test convention by dropping the test table after execution.
- Observability: Not applicable.
- Transaction/persistence/data write: Not applicable.
- FE/BE variable passing: Not applicable.
- Performance: The added UTC timezone object is local and lightweight; no performance blocker found.
User focus: No additional user-provided review focus was present.
| GROUP BY grp | ||
| """ | ||
|
|
||
| sql "DROP TABLE IF EXISTS tz_group_array_crash" |
Contributor
There was a problem hiding this comment.
Please remove this cleanup. Doris regression-test standards require dropping tables before use, not after, so the table remains available for debugging when the case fails. This test already does DROP TABLE IF EXISTS tz_group_array_crash before creating it at the start of the added case.
Contributor
Author
|
/review |
Contributor
There was a problem hiding this comment.
Code review completed for PR #63827.
No new blocking issues found. The previous inline comment about dropping the regression table after the test is already addressed in the current head diff, so I did not duplicate it.
Critical checkpoint conclusions:
- Goal and proof: The PR targets nested TIMESTAMPTZ values in collect_list/group_array aggregate state JSON serde by providing a deterministic UTC timezone. The added regression case covers ARRAY<TIMESTAMPTZ(6)> through group_array and validates the canonical UTC result.
- Scope: The change is small and focused on the complex collect-list aggregate state serialization path plus one regression case.
- Concurrency and lifecycle: No new shared state, locking, threads, or non-trivial lifecycle management are introduced. The timezone object is local to each serialization/deserialization call and is only used synchronously through FormatOptions.
- Configuration and compatibility: No configuration items are added. The aggregate state binary framing is unchanged; only the text representation for nested TIMESTAMPTZ is made deterministic with UTC.
- Parallel paths: Direct TIMESTAMPTZ collect uses the non-complex path, while nested TIMESTAMPTZ uses the modified complex serde path. The array_agg path shares this collect-list data implementation, so it benefits as well.
- Error handling and data correctness: Existing Status checks in the serialize/deserialize path are preserved. Using UTC keeps TIMESTAMPTZ aggregate-state round trips independent of BE session/local timezone while preserving the instant.
- Tests: Regression output was added for the new nested TIMESTAMPTZ group_array case. The query uses array_sort around group_array output, which makes the aggregate result deterministic enough for the expected output.
- Observability, transactions, persistence, and memory: No new observability, transaction, persistence, or significant memory-accounting concerns apply to this small serde fix.
- User focus: No additional user-provided review focus was specified.
Contributor
Author
|
run buildall |
Contributor
TPC-H: Total hot run time: 31663 ms |
Contributor
TPC-DS: Total hot run time: 171963 ms |
Contributor
BE Regression && UT Coverage ReportIncrement line coverage Increment coverage report
|
zclllyybb
approved these changes
May 29, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What problem does this PR solve?
Problem Summary:
Fix collect_list/group_array on nested TIMESTAMPTZ values when complex aggregate state is serialized through JSON. This keeps the existing state format for compatibility, provides a UTC timezone during serde, and adds regression coverage for the nested group_array case.
Release note
None
Check List (For Author)
Test
Behavior changed:
Does this need documentation?
Check List (For Reviewer who merge this PR)