ci: run the image job on mac2, and make the staging gate say what failed #253

2026-10-14T14:59:48Z

extant2000git commented

2026-10-14 14:59:48 +00:00

The image job has been red on PR #252 and green on main, and nothing in that PR touches CI. The runner is the whole difference:

run	image job ran on	result
434	mac2	success, deploy success
441	mac1	failure, "staging postgres never became ready"

Both runs were of the same branch and the same gate script. mac1 has the smaller OrbStack allocation, and the gate starts a throwaway Postgres, a Redis and the candidate relay as sibling containers on the host daemon.

Two changes:

The image job names mac2 instead of taking whichever runner is free, so the result does not depend on which machine picks the job up.
The gate prints the Postgres container's state and its log when it does not come up. "Never became ready" covered both a Postgres still starting and a container that never ran at all, and those need different fixes.

The second change is the one that matters longer term: mac1 said nothing further, so 90 seconds of waiting read as a relay bug.

🤖 Generated with Claude Code

The image job has been red on PR #252 and green on main, and nothing in that PR touches CI. The runner is the whole difference: | run | image job ran on | result | |---|---|---| | 434 | mac2 | success, deploy success | | 441 | mac1 | failure, "staging postgres never became ready" | Both runs were of the same branch and the same gate script. mac1 has the smaller OrbStack allocation, and the gate starts a throwaway Postgres, a Redis and the candidate relay as sibling containers on the host daemon. Two changes: 1. The image job names `mac2` instead of taking whichever runner is free, so the result does not depend on which machine picks the job up. 2. The gate prints the Postgres container's state and its log when it does not come up. "Never became ready" covered both a Postgres still starting and a container that never ran at all, and those need different fixes. The second change is the one that matters longer term: mac1 said nothing further, so 90 seconds of waiting read as a relay bug. 🤖 Generated with [Claude Code](https://claude.com/claude-code)

extant2000git added 1 commit 2026-10-14 14:59:48 +00:00

ci: run the image job on mac2, and make the staging gate say what failed

ci / checks (pull_request) Successful in 14s

Details

ci / swift (pull_request) Successful in 11m51s

Details

ci / image (pull_request) Successful in 51s

Details

ci / deploy (pull_request) Skipped

Details

15737e05b4

The image job failed on PR #252 and passed on main with no code difference
between them that touches CI. The runner is the difference: run 434 built on
mac2 and passed, run 441 built on mac1 and failed at "staging postgres never
became ready", on the same two commits. mac1 has the smaller OrbStack
allocation, and the gate starts three sibling containers on the host daemon.
The job now names mac2 rather than taking whichever runner is free.

The gate also prints the container's state and log when Postgres does not come
up. "Never became ready" covered both a Postgres still starting and a
container that never ran, which need different fixes, and mac1 said nothing
further, so 90 seconds of waiting read as a relay bug.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

extant2000git merged commit ebdfc87eb9 into main

2026-10-14 15:13:17 +00:00

extant2000git referenced this pull request

2026-10-14 15:14:15 +00:00

Relay review wave 2: in-process wake, signin and tokens, scale, streaming attachments, federation pins, report digest, metrics #252

extant2000git referenced this pull request

2026-10-14 15:14:15 +00:00

Relay review wave 2: in-process wake, signin and tokens, scale, streaming attachments #254

extant2000git referenced this issue from a commit

2026-10-14 16:21:53 +00:00

Merge pull request 'Relay review wave 2: in-process wake, signin and tokens, scale, streaming attachments' (#254) from release/relay-review-wave2-2026-10-14 into main

extant2000git referenced this issue from a commit