Change default auto_statistics_types from basic, uniq to basic, uniq_v2

[hanfei1991]

uniq_v2 has less overhead on inserts and memory compared to uniq.

Changelog category (leave one):

• Improvement

Changelog entry (a user-readable short description of the changes that goes into CHANGELOG.md):

Change default auto_statistics_types from basic, uniq to basic, uniq_v2

Version info

• Merged into: 26.7.1.1220 (included in 26.7 and later)

[clickhouse-gh]

Workflow [PR], commit [ca38323]

Summary: ❌

• Performance Comparison: Performance dashboard

job_name	test_name	status	info	comment
AST fuzzer (amd_debug)		FAIL
	Received signal 8 (STID: 4141-5867)	FAIL	cidb

---

AI Review

Summary

This PR changes the default auto_statistics_types for new MergeTree tables from basic, uniq to basic, uniq_v2, and updates the 26.7 compatibility history entry to match. I did not find a concrete correctness bug in the switch itself, but the change is still missing focused regression coverage for the exact default/compatibility boundary, so I cannot approve it as-is.

Tests

• ⚠️ Add a focused regression for the exact implicit statistics set, not just the absence of MinMax. Today the existing upgrade coverage would still pass if new tables kept materializing Uniq instead of uniq_v2, or if compatibility = '26.6' stopped restoring the old default. The smallest useful test is: create a table without an explicit auto_statistics_types, materialize statistics, and assert the exact implicit set under both the current defaults and compatibility = '26.6'.

Final Verdict

Status: ⚠️ Request changes

[clickhouse-gh]

LLVM Coverage Report

Metric	Baseline	Current	Δ
Lines	86.00%	86.00%	+0.00%
Functions	91.80%	91.80%	+0.00%
Branches	78.10%	78.20%	+0.10%

Changed lines: Changed C/C++ lines covered: 14/14 (100.00%) · Uncovered code

Full report · Diff report