Commit graph

7 commits

Author SHA1 Message Date
d1353360cc test(ha-commit-engine): RED integration tests for ACK feed and sync-append (task 5.1)
Tests 30-34: wired repl_on_ack feeds engine, membership mutators
trigger re-evaluation, per-entry sync-append contract (stale-ACK
rejection, committed_lsn unmoved on timeout, n_peers=1 replaces
-ENODEV). test_repl_majority walAck cases rewired to per-LSN
semantics (configAck byte-identical). Sanctioned scaffolding:
entry_lsn param on repl_wait_majority_ack (legacy body byte-identical,
callers pass 0), inert repl_set_commit_engine stub. 9 RED failures;
baseline signatures unchanged. LSN-0 edge documented for 5.2/5.3.
2026-06-11 23:09:05 +00:00
ab270fab99 test(ha-wal-replication): registration lifecycle, broadcast, ACK tests (tasks 5.1-5.3)
Add test_repl_registration_lifecycle covering the complete 3-step standby
lifecycle: register (nstandbys==1, ep pointer verified), duplicate
(idempotent, same slot), deregister (ep==NULL, nstandbys==0 — the
previously-untested slot-zeroing invariant). Tasks 5.1/5.2 were already
covered by test_broadcast_one_standby_sends_entry, test_wal_wire
round-trip, and test_wal_ack_reply_updates_acked_lsn_and_signals.
18 repl unit tests pass under ASAN.
2026-06-10 11:41:43 +00:00
e2fba08f5d test(ha-wal-replication): verify majority-ack advances committed_lsn (task 4.2)
Fix defective assertion in test_committed_lsn_advances_after_majority_ack
(before/after delta instead of >0 which fails for lsn=0 first write).
Add test_single_node_committed_lsn_unchanged documenting -ENODEV degraded
mode for no-standby path. Fix NACK test expected return from -ETIMEDOUT to
0: with 1 registered standby, quorum=1 is met by primary self-count alone.
308 tests pass, 0 fail under ASAN.
2026-06-10 11:18:25 +00:00
861b746c84 feat(ha-wal-replication): primary WalAckReply handler (task 4.1)
Handle incoming WalAckReply in repl_ack_decode_and_dispatch: extract
lsn+success, call repl_on_ack on success to update standby acked_lsn
and signal majority-ack waiter. NACK is WARN-logged with no state
change. Two unit tests verify positive path (acked_lsn updated, waiter
unblocked) and negative path (NACK leaves acked_lsn=0, wait times out).
2026-06-10 09:46:22 +00:00
81dbc4d7b4 feat(ha-wal-replication): HA→replicator standby registration bridge (tasks 2.1-2.3)
Register standbys with replicator on connect (election promotion +
late-joiner), deregister on disconnect (vote err handler + standby err
handler). repl_add_standby now idempotent (duplicate ep returns
existing slot).

Co-Authored-By: Claude Sonnet 4.6 (1M context) <noreply@anthropic.com>
2026-06-10 08:35:02 +00:00
d326ee1b2e feat: P2-3 repl_wait_ack() with pthread_cond_timedwait + ACK deadlock prevention
Implements sync replication wait: primary blocks until at least one non-catching-up
standby ACKs the target LSN, or times out. Uses pthread_cond_timedwait with
CLOCK_REALTIME deadline. Inner loop rechecks all standbys after each wakeup to
guard against spurious wakeups.

ACK deadlock prevention: repl_on_ack() is called exclusively from the io_worker
read callback (not vtaskmgr), so it only does atomic_store + pthread_cond_signal —
safe to call while vcpu threads block in repl_wait_ack().

4 new TDD tests (all green, 9 total in test_repl_unit):
- test_wait_ack_no_standbys_returns_enodev
- test_wait_ack_already_acked
- test_wait_ack_timeout (verifies >=80ms elapsed)
- test_wait_ack_woken_by_background_ack (pthread thread signals after 30ms)
2026-05-07 17:23:50 +00:00
8ef64a7ee7 feat: P2-1 repl.h/repl.c — replicator lifecycle, slot management, trim_lsn
Add replicator_t structure with per-standby connection slots (standby_conn_t),
lifecycle (repl_init/repl_fini), slot management (repl_add_standby,
repl_remove_standby, repl_expire_disconnect_slots), ACK tracking
(repl_on_ack), WAL trim guard (repl_trim_lsn), and stubs for
repl_broadcast/repl_wait_ack (wired in P2-3/P2-4).

Uses mylock_t (pthread_mutex_t alias) with mylock_lock/mylock_unlock.
snap_interval_s==0 in repl_expire_disconnect_slots forces immediate expiry
to support deterministic unit testing.

5 new Unity tests all pass; 115 pre-existing controller tests unchanged.
2026-05-07 16:25:03 +00:00