• Joined on 2026-03-08
ltms closed pull request fleet/fleetd#585 2026-09-12 16:42:31 +02:00
fleetd #581: cover completion CAS removals
ltms commented on pull request fleet/fleetd#579 2026-09-12 16:42:25 +02:00
fleetd #562: surface loop health

Already in main — closing as merged.

Head 275ac0d25 is an ancestor of origin/main. It came in through 634d33b50 ("Merge worker/562-loop-health-wiring-test-99611c-5") — the merge of…

ltms commented on pull request fleet/fleetd#580 2026-09-12 16:42:16 +02:00
fleetd #571: add TIMED_OUT_UNCONFIRMED for ATTEMPTED-delivery timeouts

Already in main — closing as merged.

Head d7239ed23 is an ancestor of origin/main, brought in by 4507bc5a7 ("Merge worker/571-attempted-outcome-5739f7-2"). Merged locally and pushed, so…

ltms commented on pull request fleet/fleetd#584 2026-09-12 16:42:06 +02:00
fleetd #562 follow-up: extract loopHealthSource factory, pin its wiring (supersedes #579)

Already in main — closing as merged.

Head 1513d4f26 is an ancestor of origin/main, brought in by 634d33b50 ("Merge worker/562-loop-health-wiring-test-99611c-5"). The merge was done…

ltms commented on pull request fleet/fleetd#585 2026-09-12 16:41:58 +02:00
fleetd #581: cover completion CAS removals

Already in main — closing as merged.

This branch's head dfeb9340b is an ancestor of origin/main, brought in by 1db79bcaa ("Merge worker/581-completionresolver-cas-sites-0542b7-6"). Gitea…

ltms opened issue fleet/fleetd#588 2026-09-12 16:40:37 +02:00
Every async ticket dies at a hardcoded 30 minutes, so any unit longer than that always reports failed while the member is still working — 41 measured at exactly 1800s
ltms commented on issue fleet/fleetd#587 2026-09-12 16:02:58 +02:00
Fleetd.java: sweep every constructor-arg wiring site and report which ones no test would notice being rewired

Delegated, split in two. Both workers are read-only and must open no PR.

Pristine Fleetd.java on main 49a5875 — both workers must restore to this exact value:

shasum -a 256…
ltms opened issue fleet/fleetd#587 2026-09-12 15:57:01 +02:00
Fleetd.java: sweep every constructor-arg wiring site and report which ones no test would notice being rewired
ltms closed issue fleet/fleetd#582 2026-09-12 15:53:02 +02:00
AmqpReplyInbox / LeadMailbox: the pendingByMsgId half of the dual-map cleanup is unasserted at every error-path site (from the #577 sweep)
ltms commented on issue fleet/fleetd#582 2026-09-12 15:52:56 +02:00
AmqpReplyInbox / LeadMailbox: the pendingByMsgId half of the dual-map cleanup is unasserted at every error-path site (from the #577 sweep)

Done. PR #583 merged as 49a5875.

Ten assertions, one per pendingByMsgId cleanup site across AmqpReplyInbox (5) and LeadMailbox (5). **Every one is proven to fail when the line it guards…

ltms deleted branch worker/582-pendingbymsgid-cleanup-ab5382-7 from fleet/fleetd 2026-09-12 15:52:27 +02:00
ltms pushed to main at fleet/fleetd 2026-09-12 15:52:27 +02:00
49a5875586 Merge #583: fleetd #582 — assert pending message-id cleanup at every publish cleanup site
4ca7d72303 #582: assert pending message-id cleanup
Compare 2 commits »
ltms merged pull request fleet/fleetd#583 2026-09-12 15:52:25 +02:00
#582: assert pending message-id cleanup
ltms commented on pull request fleet/fleetd#583 2026-09-12 15:52:17 +02:00
#582: assert pending message-id cleanup

Verified by the lead. All ten assertions are now proven live. Merging.

The PR was held because the implementer wrote ten assertions and proved only three of them red. It then proved three…

ltms commented on issue fleet/fleetd#581 2026-09-12 15:48:52 +02:00
CompletionResolver: 9 sites maintain the CAS-remove invariant, only 2 are asserted (from the #577 sweep)

Follow-up measurement: all NINE CAS sites are pinned, and none of them is an equivalent mutant

The fleet01 lead asked the right question after this merged: *were the survivors checked…

ltms commented on issue fleet/fleetd#578 2026-09-12 15:45:44 +02:00
MessageService.Outcome leaks .name() onto the wire; its sibling ReplyOutcome pins a wireName 400 lines up

CORRECTION 3 — the acceptance must be a PROPERTY, not a SHAPE. Do not implement this ticket until this is read.

Credit: the fleet01 lead found this. I measured it again here on main…

ltms closed issue fleet/fleetd#562 2026-09-12 15:31:21 +02:00
LoopWatchdog.health() is public and nothing reads it — surface it without breaking the two scripts that parse /healthz
ltms closed issue fleet/fleetd#581 2026-09-12 15:31:17 +02:00
CompletionResolver: 9 sites maintain the CAS-remove invariant, only 2 are asserted (from the #577 sweep)
ltms closed issue fleet/fleetd#571 2026-09-12 15:31:13 +02:00
MessageService collapses the new ATTEMPTED state into TIMED_OUT_QUEUED, so a possibly-delivered message is reported as one that will never arrive
ltms commented on issue fleet/fleetd#581 2026-09-12 15:31:00 +02:00
CompletionResolver: 9 sites maintain the CAS-remove invariant, only 2 are asserted (from the #577 sweep)

Merged. Seven sites, seven tests, each red under its own mutation.

Merged to main as part of 634d33b, via PR #585. Test-only, 143 insertions, one file — confirmed by git diff --stat,…