Give the login redirect room on a starved host #3

2026-10-03T19:55:49+02:00

mi commented

2026-10-03 19:55:49 +02:00

The trunk build of 98e582d failed in the first Firefox scenario, waiting for the redirect to /login?sent. Its application needed 15.6 seconds to start, where every later scenario needed 1.2 to 2.4, because productionAcceptance was starting its own JVMs in the same seconds. The wait now allows 45 seconds instead of Playwright's 15; the destination is unchanged, so the proof is the same and only a starved host is tolerated.

The overlap of the two acceptance tasks stays: werkjournal is built rarely, and -PproductionThreads=1 would make the production suite the longer branch at a cost of about a minute. The measurement notes now say so, correcting the earlier claim that the knob is free.

Local pipeline green with unchanged counts: 68 operations tests, 257 JVM cases with 46 skipped and no failures, 22 Firefox scenarios, 7 production-JAR scenarios, 20 frontend tests.

🤖 Generated with Claude Code

The trunk build of 98e582d failed in the first Firefox scenario, waiting for the redirect to `/login?sent`. Its application needed 15.6 seconds to start, where every later scenario needed 1.2 to 2.4, because `productionAcceptance` was starting its own JVMs in the same seconds. The wait now allows 45 seconds instead of Playwright's 15; the destination is unchanged, so the proof is the same and only a starved host is tolerated. The overlap of the two acceptance tasks stays: werkjournal is built rarely, and `-PproductionThreads=1` would make the production suite the longer branch at a cost of about a minute. The measurement notes now say so, correcting the earlier claim that the knob is free. Local pipeline green with unchanged counts: 68 operations tests, 257 JVM cases with 46 skipped and no failures, 22 Firefox scenarios, 7 production-JAR scenarios, 20 frontend tests. 🤖 Generated with [Claude Code](https://claude.com/claude-code)

mi added 1 commit 2026-10-03 19:55:49 +02:00

Give the login redirect room on a starved host

werkator/branch build succeeded after 09:36

Details

e6ad2235d5

The trunk build of 98e582d failed in the first Firefox scenario, waiting
for the redirect to /login?sent. Its application needed 15.6 seconds to
start, where every later scenario needed 1.2 to 2.4, because the
production suite was starting its own JVMs in the same seconds. The wait
now allows 45 seconds instead of Playwright's 15. The destination is
unchanged, so the proof is the same; only a starved host is tolerated.

The overlap of the two acceptance tasks stays as it is: werkjournal is
built rarely, and -PproductionThreads=1 would make the production suite
the longer branch at a cost of about a minute. The measurement notes
now say so, correcting the claim that the knob is free.

Local pipeline green with unchanged counts: 68 operations tests, 257 JVM
cases with 46 skipped and no failures, 22 Firefox scenarios, 7
production-JAR scenarios and 20 frontend tests.

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

mi merged commit 4d1f32483d into main

2026-10-03 19:59:29 +02:00

mi deleted branch login-wait-headroom

2026-10-03 19:59:29 +02:00

Sign in to join this conversation.

1 Participants

Notifications

Due Date

No due date set.

Dependencies

No dependencies set.

Reference: mi/werkjournal#3