Skip to content

Stable 25.8 Backport of #123436 - Fix ALIAS row policies on legacy distributed reads - #2496

Merged
mkmkme merged 2 commits into
stable-25.8from
backports/stable-25.8/123436
Oct 9, 2026
Merged

mkmkme merged 2 commits into
stable-25.8from
backports/stable-25.8/123436

Conversation

@mkmkme

@mkmkme mkmkme commented Oct 7, 2026

Copy link
Copy Markdown
Collaborator

Fix ALIAS row policies on legacy distributed reads

Changelog category (leave one):

  • Bug Fix (user-visible misbehavior in an official stable release)

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

Fix a bug where using aliases and row policies with enable_analyzer=0 could error out (ClickHouse#123436 by @mkmkme)

Documentation entry for user-facing changes

...

CI/CD Options

Exclude tests:

  • Fast test
  • Integration Tests
  • Stateless tests
  • Stateful tests
  • Unit tests
  • Performance tests
  • Aarch64 tests
  • All with ASAN
  • All with TSAN
  • All with MSAN
  • All with UBSAN
  • All with Coverage
  • All Regression
  • Disable CI Cache

Regression jobs to run:

  • Fast suites (mostly <1h)
  • Aggregate Functions (2h)
  • Alter (1.5h)
  • Benchmark (30m)
  • CAS (content-addressed storage; Antalya only)
  • ClickHouse Keeper (1h)
  • Iceberg (2h)
  • LDAP (1h)
  • OAuth (5m)
  • Parquet (1.5h)
  • RBAC (1.5h)
  • SSL Server (1h)
  • S3 (2h)
  • S3 Export (2h)
  • Swarms (30m)
  • Tiered Storage (2h)

…-distributed-26.8

26.8: Fix `ALIAS` row policies on legacy distributed reads
@mkmkme mkmkme added stable backport Backport 25.8 25.8 Altinity Stable stable-25.8 labels Oct 7, 2026
@github-actions

github-actions Bot commented Oct 7, 2026 •

Copy link
Copy Markdown

Workflow [PR], commit [83d5877]

@mkmkme

mkmkme commented Oct 9, 2026

Copy link
Copy Markdown
Collaborator Author

@blau-ai

@blau-ai

blau-ai commented Oct 9, 2026

Copy link
Copy Markdown
Collaborator

CI triage — #2496 (backport of ClickHouse#123436, ALIAS row policies on legacy distributed reads)

Verdict: 4 failing checks, 0 caused by this PR. 2 are base-image CVE scans, 2 are the stress-test runner failing before the server ever started (Azure emulator unreachable). Every test suite that actually exercises this PR's code — including the suites running the new 05316_row_policy_alias_column_distributed test — is green. No code change is needed from you; the reds are a known base-image CVE plus a transient infra flake, safe to re-run.

Head SHA: 83d5877


1–2. Grype Scan (keeper + server -alpine) — NOT PR-caused (base-image CVE)

Both fail with "Grype Scan Completed with 1 high/critical vulnerabilities". The keeper report names it:

/docker vulnerabilities/CVE-2026-85091 @ alpine: distro: alpine: 3.21, High

This is an OS-package CVE in the Alpine base image, surfaced by scanning the built Docker image — it has nothing to do with a C++ source backport. The diff touches only src/Interpreters/InterpreterSelectQuery.cpp and a stateless test; it cannot add or remove an Alpine package CVE. These scans fail fleet-wide until the base image / Alpine CVE fix lands (cf. the recent 25.8-stable-alpine-cve-fix merge on this branch) and are tracked separately by the release team.
Next step: none for this PR. Rebase once the Alpine CVE bump reaches stable-25.8, or let the release owners clear it.

3–4. Stress test (amd_debug) and Stress test (amd_ubsan) — NOT PR-caused (infra)

Both report check_status.tsv doesn't exists, i.e. the job produced no results because the server never started. From run.log, the harness tried clickhouse start 120 times, each time hitting:

Code: 210. DB::NetException: Connection refused (localhost:9000). (NETWORK_ERROR)

and application_errors.txt shows why the server kept dying on boot — it was crashing during metadata load, before any query runs:

<Error> Application: Caught exception while loading metadata: std::exception. Code: 1001,
type: Azure::Storage::StorageException, e.what() = 500 Poco::Exception. Code: 1000,
e.code() = 111, Connection refused

The stress config enables Azure/S3 storage policies, and the Azure Blob emulator (Azurite) was unreachable (Connection refused), so the server couldn't load disk metadata and never came up. Identical root cause on both amd_debug and amd_ubsan. This is an environment failure at startup; this PR only changes row-policy ALIAS handling during query interpretation — code that isn't reached until a server is running and a query executes.
Next step: re-run the two Stress test jobs. No code change required.


Health check (the parts that matter for this PR are green)

All builds, unit tests, and the full stateless matrix passed, including the suites that run the new test 05316_row_policy_alias_column_distributed (tagged distributed, enable_analyzer=0):

  • Stateless tests (amd_debug, parallel) / (amd_debug, sequential) ✅
  • Stateless tests (amd_asan, distributed plan, parallel 1/2, 2/2, sequential) ✅
  • Stateless tests (amd_ubsan, parallel/sequential) ✅
  • Stateless tests (amd_binary, old analyzer, …) and ParallelReplicas variants ✅
  • All Build (*) and Unit tests (asan/ubsan) ✅

So the backport's behavioral change is validated by CI. The only reds are the Alpine base-image CVE (release-owned) and the Azure-emulator stress-test infra flake (re-runnable). Nothing here blocks on the author.

Analysis from CI reports only — I can't build or run ClickHouse in this container.

@mkmkme mkmkme added the verified Approved for release label Oct 9, 2026
@mkmkme
mkmkme merged commit 6f83247 into stable-25.8 Oct 9, 2026
226 of 235 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

25.8 25.8 Altinity Stable 25.8.33.10001 backport Backport stable stable-25.8 verified Approved for release

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants