Skip to content

fix(t27b-lab): a clone that cannot check out is cloned again - #6220

Merged
gHashTag merged 1 commit into
masterfrom
claude/t27b-lab-checkout-heal
Oct 4, 2026
Merged

gHashTag merged 1 commit into
masterfrom
claude/t27b-lab-checkout-heal

Conversation

@gHashTag

@gHashTag gHashTag commented Oct 4, 2026

Copy link
Copy Markdown
Owner

Closes #6219. Refs #6063.

After the owner's redeploy on 2026-10-04 ~17:55Z, the lab's first run at bae81ae failed at checkout with fatal: could not fetch 737282229a... from promisor remote. lab.checkout() had no second try, so a broken partial clone on the volume would fail every later poll.

Change

  • checkout() tries once. On a failed fetch or checkout it removes the clone and clones again, once.
  • steps.checkout now records clone: kept or clone: recloned plus first_error, so a heal is visible in the run.

Evidence

  • scripts/ci/test_the_t27b_lab_heals_its_clone.py: 4/4 ok. It uses local file:// repos and no network.
  • Negative control: the same test against master's lab.py fails (the checkout raises git fetch failed).
  • Census re-blessed in this PR (rule Q12): shell.txt run steps 284 -> 285, from the one new step.

Note

This takes effect only after the next lab redeploy, which is the owner's decision.

🤖 Generated with Claude Code

After the 2026-10-04 redeploy the lab's checkout of bae81ae failed on
its partial clone ("could not fetch ... from promisor remote") and
checkout() had no second try. It now removes the clone and clones once
more; steps.checkout records clone kept/recloned and the first error.
New CI step in loop-tools-gate; master's checkout fails the same case.

Census: shell.txt run steps 284 -> 285 (runner bash 263 -> 264) -- the
one new loop-tools-gate step above (rule Q12).

Closes #6219

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@github-actions

github-actions Bot commented Oct 4, 2026

Copy link
Copy Markdown
Contributor

PR Dashboard

Generated at: 2026-10-04 18:00:20 UTC

Summary

Status Count
Total Open PRs 42
PRs with Failing Checks 29
PRs with All Checks Green 13
READY 12
FAILING 29
PENDING 0
NO CHECKS YET 0

These columns do not partition: 12 + 29 + 0 + 0 = 41, and there are 42 open PRs. A PR is being counted twice or not at all.

Seal Status

  • ⚠️ STALE -- sha256(compiler.rs)=8597b6ded596 != manifest seal=87e5cbd3ad94.
    The committed NMSE numbers were certified against an older compiler.rs.
    Run scripts/reseal-check.sh locally for the two-step reseal command (advisory; not a merge gate).

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.

t27b lab: a clone that cannot check out is cloned again

1 participant