Public: anyone can read this; the crew can change it.

The no-release alarm reaches the owner when production stands still for 90 minutes while main moves: a test replays 2026-11-11 #11673

shipped production is serving it opened this has it

Landed in pull #11680 · 2 runs · run

opened this

Part of #11507 (v1 list, item 2). On 2026-11-11 production stood on 0.7.535 for 5 h 50 min while main merged dozens of pulls, and nothing said so — yet src/flow/alarm_no_release.ts (#5930) and src/release/behind.ts (#2613) both exist. Find why they stayed silent and fix that, nothing more.

Done looks like:

Files: src/flow/alarm_no_release.ts, src/release/landings.ts or the notifier that tells the owner, one new test. Risks: do not touch the release workflow or the chart; do not add a second alarm beside the existing one.

the crew's run was interrupted: restart, resumes on the next pass — cancelled
commented

Reviewed pull #11680 at its head: ticked The cause is named in the pull (for example: waitingSince in src/release/landings.ts answering undefined, behind.ts needing a tag that was never cut, or the notifier telling the owner once per start of an alarm that never ended).; A new in-process test (temp dir, createApp, no servers) builds the 11-11 shape: a fixed running commit, merged pulls after it, no keeper rows, no new tag. At 89 minutes after the oldest unreleased merge no alarm fires; at 91 exactly one no_release alarm fires in flowAlarms and the owner is told exactly once; when the running commit contains the merges it clears. It fails without the fix.; The alarm's words name production's version and short commit, how many merged pulls are waiting, and how long.; bun x tsc --noEmit and the touched tests pass in the sandbox before the push; a new test file joins scripts/gate.ts if the gate does not find it by itself..

Sign in to join this conversation.

↑↓ navigate ↵ open ⌘↵ its app esc close

Keyboard shortcuts

Docs