Skip to content

ci: [branch-1.2] set timeout-minutes on every workflow job (backport #7679) - #7682

Merged
andygrove merged 1 commit into
apache:branch-1.2from
andygrove:backport-7679-branch-1.2
Nov 19, 2026
Merged

andygrove merged 1 commit into
apache:branch-1.2from
andygrove:backport-7679-branch-1.2

Conversation

@andygrove

@andygrove andygrove commented Nov 18, 2026 •

Copy link
Copy Markdown
Member

Which issue does this PR close?

Backport of #7679 to branch-1.2. #7679 is approved but not merged yet, so this should merge after it. There is no issue.

Rationale for this change

No job in the workflows sets timeout-minutes, so each one gets GitHub's six hour default. A hung suite holds a runner, and its pipeline, for six hours before it reports a failure. This came up while investigating CI failures on the branch-1.2 backports, and the branch has the same workflows, so it needs the same limits. CI changes for the branch's own build qualify for a backport (see the backporting guide).

What changes are included in this PR?

The same limits as #7679, applied to the jobs this branch has. The cherry-pick conflicted because branch-1.2 is older than main:

  • ci_fork.yml does not exist here, so its change is dropped.
  • pr_build_linux.yml has no linux-test-native-executors job, so that limit is dropped. The other limits are unchanged.
  • docs.yaml builds and publishes in one job instead of separate build and deploy jobs. That job gets the 60 minute build limit.
  • label_prs.yml also has a job that ci: set timeout-minutes on every workflow job #7679 does not touch, because main no longer has it in this form. It gets a 10 minute limit like the other labelling jobs.

Every job with runs-on now has timeout-minutes. The README section is unchanged from #7679 and its references (pr_build_linux.yml, spark_sql_test_reusable.yml, the merge queue's check_response_timeout_minutes: 300) all exist on this branch.

How are these changes tested?

actionlint, dev/ci/check-ci-config.py and prettier --check on the README pass. A script confirmed that every job with runs-on has timeout-minutes. This pull request's own CI runs under the new limits.

Without timeout-minutes a job runs for GitHub's six hour default. A hung
suite then holds a runner, and its pipeline, for six hours before it
reports. Set a limit on every job that runs on a runner, at about two to
three times its slowest recent successful run, and document how the
limits were chosen in the workflows README.

(cherry picked from commit 4b73842)

Adapted for branch-1.2: ci_fork.yml and the linux-test-native-executors
and deploy-docs jobs do not exist on this branch, so their limits are
dropped. docs.yaml builds and publishes in one job, which gets the 60
minute build limit. label_prs.yml, which main no longer has in this
form, gets a 10 minute limit.
@andygrove
andygrove marked this pull request as draft November 18, 2026 23:43
@andygrove andygrove changed the title ci: set timeout-minutes on every workflow job (backport #7679) ci: [branch-1.2] set timeout-minutes on every workflow job (backport #7679) Nov 18, 2026
@andygrove
andygrove marked this pull request as ready for review November 19, 2026 14:08
@andygrove
andygrove merged commit ce18950 into apache:branch-1.2 Nov 19, 2026
142 of 143 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants