Record the real-rig run of Slice 3 Task 5 #1477

Merged

forgejo-admin merged 1 commit from explore/s3-rig-run into main

2026-09-30 21:56:49 +00:00

forgejo-admin commented

2026-09-30 21:53:58 +00:00

Owner

Problem

Slice 3's task list records Task 5 (run seed S1 end to end on the rig, then prove cleanup.sh) as blocked on celilo#1471, and its status as Ready. The block has since been fixed on main (532e8716), and the task has now been run.

Solution

Docs only: tick Task 5 in tasks-slice-3.md with the evidence, keep one line on the earlier block, and set the list's status to Complete (all seven tasks are done). Plus an empty changeset.

Evidence (run 2026-09-30 on `main` `a40d5cb7`, rig images rebuilt, celilo 5.3.0)

The hold suite collected and paused at net.debug(). The earlier Cannot find module '@celilo/e2e' failure is gone.
Baseline healthy: /foo/ and /bar/ served their own content; public DNS returned the firewall's address, the internal view the zone-side one.
Seed S1 behaved as expected: deploy exit 0, /foo-bar/ served "hello bar" and /foo/ still "hello foo", and the snapshot diff added the foo-bar web roots and a /foo-bar/* Caddy block.
cleanup.sh exited 0 against the live, paused suite: lock free, 0 per-test containers, 0 suite processes (the suite exited by its resume file, log ends hold-suite exit=0), hold suite file deleted. No signals were sent.
Side observation, not a verdict: /bar/ still serves after the path move, so issue #1468 is unfixed on this build.

Not exercised: the firewall, module-host and hairpin probe rows, and the IPAM control (seed S6 needs the simulator mode). Issue #1471 is still open; I have not commented on or closed it.

🤖 Generated with Claude Code

## Problem Slice 3's task list records Task 5 (run seed S1 end to end on the rig, then prove `cleanup.sh`) as blocked on celilo#1471, and its status as Ready. The block has since been fixed on `main` (`532e8716`), and the task has now been run. ## Solution Docs only: tick Task 5 in `tasks-slice-3.md` with the evidence, keep one line on the earlier block, and set the list's status to Complete (all seven tasks are done). Plus an empty changeset. ## Evidence (run 2026-09-30 on `main` `a40d5cb7`, rig images rebuilt, celilo 5.3.0) - The hold suite collected and paused at `net.debug()`. The earlier `Cannot find module '@celilo/e2e'` failure is gone. - Baseline healthy: `/foo/` and `/bar/` served their own content; public DNS returned the firewall's address, the internal view the zone-side one. - Seed S1 behaved as expected: deploy exit 0, `/foo-bar/` served "hello bar" and `/foo/` still "hello foo", and the snapshot diff added the `foo-bar` web roots and a `/foo-bar/*` Caddy block. - `cleanup.sh` exited 0 against the live, paused suite: lock free, 0 per-test containers, 0 suite processes (the suite exited by its resume file, log ends `hold-suite exit=0`), hold suite file deleted. No signals were sent. - Side observation, not a verdict: `/bar/` still serves after the path move, so issue #1468 is unfixed on this build. Not exercised: the firewall, module-host and hairpin probe rows, and the IPAM control (seed S6 needs the simulator mode). Issue #1471 is still open; I have not commented on or closed it. 🤖 Generated with [Claude Code](https://claude.com/claude-code)

forgejo-admin added 1 commit

2026-09-30 21:53:58 +00:00

Record the real-rig run of Slice 3 Task 5

npm-consumer-smoke / npm-consumer-smoke (push) Successful in 2s

Details

changeset-check / require-changeset (pull_request) Successful in 10s

Details

ci / validate (pull_request) Successful in 10m6s

Details

pr-guards / provisional (pull_request) Successful in 2s

Details

pr-guards / migration-journal (pull_request) Successful in 3s

Details

pr-guards / openspec-validate (pull_request) Successful in 3s

Details

5afa9e597a

Task 5 was blocked on celilo#1471, fixed on main as 532e8716. It now
ran on main a40d5cb7: the hold suite collects, seed S1 behaves as
expected, and cleanup.sh leaves the machine clean with no signals sent.

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>

forgejo-admin merged commit 3c8d79b77a into main

2026-09-30 21:56:49 +00:00

forgejo-admin referenced this pull request from a commit

2026-09-30 21:56:50 +00:00

Merge pull request 'Record the real-rig run of Slice 3 Task 5' (#1477) from explore/s3-rig-run into main