Pick an option on each card and press Done; your answer goes straight
back to the session. The raiser's pick is marked. "I'm not sure" is a real answer.
1Raise the merge queue's check timeout from 30 to 45 minutes?
In plain terms. When a batch of PRs is being tested together it now takes a bit over 30 minutes, and GitHub gives up at exactly 30 and throws the PRs out even when they passed. Should we give it 45 minutes instead?
Why, and what was tried
The session raising this recommends: 45 min Big queue trains now take 30-35 min; #17468 was kicked out twice, once at 16:42Z for a train that went green at 16:44Z. 45 covers that with room; 60 just lets a truly stuck train hold the line longer. Tried first, unsuccessfully: Read the live ruleset 'Require CI' (id 14925316): check_response_timeout_minutes=30, 8 PRs per train. Spot reclaims no longer red a train (#16996/#17231), so the remaining ejections are pure slowness, not failures.
bacda5a9 is parked on this.
2Should PRs armed from Fede's side also get the pre-queue type check?
In plain terms. Before a PR gets in line to merge, Gera's merge command runs a quick type check, but Fede's side skips it, and those skipped ones cause about a third of the failed test batches. Should Fede's side run the same quick check?
Why, and what was tried
The session raising this recommends: Yes, add it About 30% of red queue trains in the last 24h were type errors on Fede-armed PRs against newer main. Those cost a full train run each; a 2-3 min type check before arming would catch them. Tried first, unsuccessfully: auto-merge-all.yml is disabled_manually since 10-01 20:48Z, and that was on purpose (the 'retire auto-merge-all' series, #14592-#14788). But it took the pre-queue check with it for every arm path except Gera's pr-merge command, which dispatches prequeue-typecheck.yml. Fix is to make Fede's arm path dispatch the same workflow, not to re-enable auto-merge-all.