MeloMei e3e3e7e23f fix(realtime): bounded replay window for ShardedStreamRelay on restart (#4875) ...

ShardedStreamRelay shard readers previously started with lastID=,
which meant events published while a pod was down were silently lost.

Replace the  cursor with a bounded time-window start ID derived
from (now - ReplayGrace), defaulting to 5 minutes. The timestamp is
clamped to 0 to handle misconfigured clocks gracefully.

Key changes:
- Add ReplayGrace field to ShardedStreamRelayConfig (default 5m)
- Add replayStartID() helper with non-negative clamp
- Extract readShardOnce() from readShard() for testability
- Add REALTIME_RELAY_REPLAY_GRACE env var for runtime tuning
- Add regression tests for bounded cursor and replay behavior

Closes #4797

2026-07-03 12:16:43 +08:00

..

backfill_codex_usage_cache

MUL-3264: add Codex usage cache backfill

2026-06-18 13:34:10 +08:00

backfill_task_usage_hourly

…

migrate

test(migrate): concurrent migration race test using real Postgres (MUL-2956) (#3712)

2026-06-08 13:33:16 +08:00

multica

fix(cli): prevent daemon-managed CLI from silently using user tokens (MUL-3922)

2026-07-02 15:08:06 +08:00

server

fix(realtime): bounded replay window for ShardedStreamRelay on restart (#4875)

2026-07-03 12:16:43 +08:00