480772c108
spec(ha-commit-durability): 6.1 answered — durability loss + liveness are REAL (reproduce off overlay), not an fdatasync artifact; Phase 6 now at 6.2 root-cause
2026-06-20 03:19:57 +00:00
c205b5b25d
spec(ha-commit-durability): sharpen bug-795 — durability loss reproduces in rolling failover (trustworthy soak) + >60s election liveness finding
2026-06-20 01:28:26 +00:00
3ad646ad0e
spec(ha-commit-durability): REOPEN for bug-795 total-outage cold-restart-all committed-job loss
...
Phase 6 added (6.1 reproduce-on-real-fs gates 6.2 root-cause / 6.3 fix / 6.4 revalidate).
Surfaced by the xce-test-e2e-primitives cold-restart-all soak step; Phase 5 only
validated rolling single-node failover. CLAUDE.md durability claim qualified pending 6.x.
2026-06-20 00:33:18 +00:00
d9b43c2d2f
chore(ha-commit-durability): mark 5.1 + all phase groups complete (all 18 sub-tasks done)
2026-06-18 11:21:10 +00:00
f139e1c08e
docs(ha-commit-durability): flip bug-661 KNOWN GAP to GUARANTEED + re-verify protocol checklist (task 5.3)
2026-06-18 11:15:13 +00:00
d128cc96ef
test(ha-commit-durability): verify zero acked-job loss + cluster availability across failover on no-self-exit build (task 5.2)
2026-06-18 11:12:15 +00:00
59d9df246c
fix(ha-commit-durability)!: controller never self-terminates — loud non-promotable fence replaces recovery FATAL/abort
...
Per design invariant (a controller exits only on crash or admin kill), replace the
self-abort recovery terminal (tasks 3.3/4.3/4.4) with a loud, permanent, non-promotable
fence that stays alive and keeps retrying catch-up (self-heals on peer return; no manual
admin recovery). Records the never-self-terminate invariant as project-wide guidance in
src/controller/CLAUDE.md + .kiro/steering/ha-async-io.md.
2026-06-18 11:10:19 +00:00
addf53e3b4
feat(ha-commit-durability): bounded-window catch-up before FATAL — fence no longer FATALs synchronously (task 4.4)
2026-06-18 10:35:13 +00:00
e14bd62115
test(ha-commit-durability): end-to-end durable-majority commit + liveness integration tests (task 5.1)
2026-06-18 09:56:26 +00:00
e9e2a6d703
feat(ha-commit-durability): pre-promotion eligibility gate + catch-up fence/release (task 4.3)
2026-06-18 09:40:32 +00:00
3049b620f3
feat(ha-commit-durability): durable-backed vote/PreVote freshness floor + bare-durable candidate advertise (task 4.2)
2026-06-18 09:02:11 +00:00
4b7457144c
docs(ha-commit-durability): pin durable-backed freshness-floor formula + election safety/liveness verification (task 4.1)
2026-06-18 08:44:42 +00:00
c1c35ab1b3
feat(ha-commit-durability): publish durable floor from recovered WAL tip (task 3.4)
2026-06-18 08:24:46 +00:00
e3fc42c11e
feat(ha-commit-durability): loud bounded FATAL on unrecoverable recovery (task 3.3)
2026-06-18 08:15:22 +00:00
6483ee8afd
docs(ha-commit-durability): pin boot-policy decision table + pending skeleton tests (task 3.2)
2026-06-18 08:00:52 +00:00
dd4d70863b
feat(ha-commit-durability): WAL recovery integrity verdict (CLEAN/TRUNCATED) + recovery_truncated flag (task 3.1)
2026-06-18 07:43:01 +00:00
526bca9679
feat(ha-commit-durability): gate legacy sync-append commit store on durable_lsn + commit<=durable invariant (task 2.4)
2026-06-18 07:09:47 +00:00
e4c4968331
feat(ha-commit-durability): gate standby ACK on fdatasync success + persist-before-ACK regression test (task 2.3)
2026-06-18 06:45:21 +00:00
c41edc515e
feat(ha-commit-durability): cap commit self-contribution at durable_lsn + durable-advance commit wake (task 2.2)
2026-06-18 06:29:47 +00:00
cc83caf5a0
feat(ha-commit-durability): demand-driven group-commit flusher + durable_lsn watermark (task 2.1)
2026-06-18 06:06:55 +00:00
7e6f09f804
test(ha-commit-durability): deterministic strace-based ack-before-durable gap repro (task 1.3)
2026-06-18 05:39:06 +00:00
f90395282c
feat(ha-commit-durability): add XCE_TESTING durable-LSN-on-disk verification seam (task 1.2)
2026-06-18 04:56:50 +00:00
1ca2d640cb
feat(ha-commit-durability): add --wal-sync-ms flush-cadence override (task 1.1)
2026-06-18 04:46:01 +00:00
