From 511aafb25291ba95326349a667423859089d2bd7 Mon Sep 17 00:00:00 2001 From: Joshua Temple Date: Sat, 27 Jun 2026 10:54:31 -0400 Subject: [PATCH] fix(fleet): raise dispatch-suite watch cap 75->180 min Under full-fleet load a passing suite can queue plus run past the 75-minute watch window (callbacks hit this: its suite succeeded but the watch gave up), reding the lane on a flake. Raise MAX_ATTEMPTS 75->180 (3h at 60s) for headroom, still well under GitHub's 6h per-job execution limit and short enough that a genuinely hung suite still fails in reasonable time. Part of #373. Signed-off-by: Joshua Temple --- .github/actions/dispatch-suite/action.yaml | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/.github/actions/dispatch-suite/action.yaml b/.github/actions/dispatch-suite/action.yaml index d9cb74bb..93975e95 100644 --- a/.github/actions/dispatch-suite/action.yaml +++ b/.github/actions/dispatch-suite/action.yaml @@ -110,11 +110,11 @@ runs: # This loop retries on transient API errors while still failing closed on # real run failures and on timeout. # - # MAX_ATTEMPTS * WATCH_INTERVAL = wall-clock cap (~75 min at 60s each). + # MAX_ATTEMPTS * WATCH_INTERVAL = wall-clock cap (~180 min (3h) at 60s each, headroom for queue+run under full-fleet load, well under GitHub's 6h job cap). # The 4env suite runs the full lifecycle plus a chained multi-env hotfix # and the Step 13 conflict probe, each a serial cascade run, so the cap # must clear its 60-minute internal job timeout with headroom. - MAX_ATTEMPTS=75 + MAX_ATTEMPTS=180 CONSEC_ERRORS=0 MAX_CONSEC_ERRORS=5 attempt=0