pingqiu 943000ae8e fix: RebuildSourceDecision returns FullBase when CommittedLSN=0 ...

When CommittedLSN=0 (sync_all mode, replica degraded), snapshot-tail
rebuild was chosen because IsRecoverable(checkpoint, 0) is vacuously
true (0 <= HeadLSN always). But snapshot-tail requires a valid committed
endpoint for tail-replay. Without it, ExecuteRebuildPlan calls
TransferSnapshot which RemoteRebuildIO doesn't support → immediate fail.

Fix: if CommittedLSN=0, force RebuildFullBase. This is the correct
source when the primary has data but no replica has confirmed durability.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

2026-04-09 15:46:00 -07:00

..

runtime

feat: Phase 20 acceptance fixes + sw-test-runner suite mode

2026-04-06 11:30:54 -07:00

adapter.go

fix: add full-base pin and clean up session on WAL pin failure

2026-03-30 12:20:24 -07:00

budget.go

feat: add V2 engine replication core (Phase 05 Slice 1)

2026-03-29 20:51:01 -07:00

command.go

feat: rebuild MVP — dual-lane session with bitmap protection

2026-04-07 14:30:34 -07:00

doc.go

feat: start Phase 14 V2 core shell

2026-04-03 16:11:38 -07:00

driver_test.go

fix: derive WAL pin from actual replay need, PlanRebuild fails closed

2026-03-30 12:51:38 -07:00

driver.go

fix: PlanRebuild targetLSN=0 when replica is degraded (CommittedLSN fallback)

2026-04-09 15:35:31 -07:00

engine.go

feat: rebuild hardening A1-A5 + session-controlled execution path

2026-04-08 14:39:11 -07:00

event.go

feat: rebuild MVP — dual-lane session with bitmap protection

2026-04-07 14:30:34 -07:00

executor_test.go

fix: rebuild executor consumes bound plan, fix catch-up timing

2026-03-30 13:33:52 -07:00

executor.go

feat: Phase 09 — V2 execution primitives and production closure

2026-04-02 16:25:23 -07:00

go.mod

feat: add V2 engine replication core (Phase 05 Slice 1)

2026-03-29 20:51:01 -07:00

history.go

fix: RebuildSourceDecision returns FullBase when CommittedLSN=0

2026-04-09 15:46:00 -07:00

integration_test.go

fix: orchestrator owns full catch-up contract (budget + truncation)

2026-03-30 11:04:34 -07:00

observe.go

feat: add integration closure and observability (Phase 05 Slice 4)

2026-03-30 00:15:46 -07:00

orchestrator.go

fix: orchestrator owns full catch-up contract (budget + truncation)

2026-03-30 11:04:34 -07:00

outcome.go

feat: add V2 engine replication core (Phase 05 Slice 1)

2026-03-29 20:51:01 -07:00

ownership_test.go

fix: separate stable ReplicaID from Endpoint in registry

2026-03-29 21:06:11 -07:00

phase14_boundary_test.go

refactor: aggregate bounded multi-replica catch-up conservatively

2026-04-04 09:27:03 -07:00

phase14_command_test.go

refactor: close bounded recovery drain and invalidation seams

2026-04-04 11:01:12 -07:00

phase14_core_test.go

feat: Phase 14A+14B — V2 core publication ownership + command semantics

2026-04-03 16:52:55 -07:00

phase14_sync_test.go

feat: rebuild hardening A1-A5 + session-controlled execution path

2026-04-08 14:39:11 -07:00

projection.go

feat: rebuild MVP — dual-lane session with bitmap protection

2026-04-07 14:30:34 -07:00

rebuild.go

fix: correct catch-up entry counting and rebuild transfer gate

2026-03-29 21:35:03 -07:00

recoverability_test.go

fix: connect recovery decisions to RetainedHistory, fix rebuild source

2026-03-29 23:55:31 -07:00

recovery_test.go

fix: correct catch-up entry counting and rebuild transfer gate

2026-03-29 21:35:03 -07:00

registry.go

fix: separate stable ReplicaID from Endpoint in registry

2026-03-29 21:06:11 -07:00

sender.go

fix: add RecoveryOrchestrator as real integrated entry path

2026-03-30 00:25:58 -07:00

session.go

fix: correct catch-up entry counting and rebuild transfer gate

2026-03-29 21:35:03 -07:00

state.go

feat: rebuild MVP — dual-lane session with bitmap protection

2026-04-07 14:30:34 -07:00

types.go

feat: add V2 engine replication core (Phase 05 Slice 1)

2026-03-29 20:51:01 -07:00

validation_test.go

fix: true mid-executor invalidation test via OnStep hook

2026-03-30 15:51:21 -07:00