mirror of
https://github.com/seaweedfs/seaweedfs.git
synced 2026-08-19 05:36:58 +00:00
docs(p15): update G9C post-close durable ack evidence
This commit is contained in:
@@ -1,7 +1,7 @@
|
||||
# V3 Phase 15 G9C — ReplicaReady Feed Continuity / Post-Close ACK Mini-Plan
|
||||
|
||||
**Date**: 2026-05-02
|
||||
**Status**: A-slice landed (`1e36a50`)
|
||||
**Status**: B-slice landed (`ca5d4f6`)
|
||||
**Code branch**: `p15-g9c/replica-ready-feed-continuity`
|
||||
**Predecessors**: G9B replica join lifecycle close
|
||||
|
||||
@@ -88,10 +88,32 @@ Both passed on `p15-g9c/replica-ready-feed-continuity`.
|
||||
|
||||
---
|
||||
|
||||
## 6. Next Slice
|
||||
## 6. B-Slice Evidence
|
||||
|
||||
G9C-B should move from component predicate to integration signal shape:
|
||||
Code: `seaweed_block@ca5d4f6` — `G9C: emit post-close durable ack from dual-lane recovery`
|
||||
|
||||
1. Ensure the real recovery feeder path emits a post-close durable/progress fact after the recovery window closes.
|
||||
2. Add one L2 or subprocess test where a recovered replica remains `recovering` immediately after close, then becomes `ready` only after the real progress callback.
|
||||
What landed:
|
||||
|
||||
1. Receiver `BarrierResp` now carries durable achieved LSN: `min(TryComplete achieved, Sync frontier)`.
|
||||
2. Dual-lane `BlockExecutor` emits `OnDurableAck` after successful `OnSessionClose`.
|
||||
3. The post-close ACK uses the receiver's durable BarrierResp witness, not the primary frontier intent.
|
||||
4. `TestDualLane_BlockExecutor_StartRebuild` now pins event order: close first, post-close durable ACK second.
|
||||
|
||||
Verification:
|
||||
|
||||
```powershell
|
||||
go test ./core/recovery/... ./core/transport/... ./core/engine ./core/adapter ./core/host/volume -count=1
|
||||
go test ./core/authority ./core/recovery/... ./core/transport/... ./core/engine ./core/adapter ./core/host/volume -count=1
|
||||
```
|
||||
|
||||
Both passed on `p15-g9c/replica-ready-feed-continuity`.
|
||||
|
||||
---
|
||||
|
||||
## 7. Next Slice
|
||||
|
||||
G9C-C should move from transport signal shape to host/daemon lifecycle evidence:
|
||||
|
||||
1. Add one L2 or subprocess test where a recovered replica remains `recovering` immediately after close, then becomes `ready` only after the real progress callback.
|
||||
2. Prove the status endpoint reflects the same transition.
|
||||
3. Keep placement and ACK voter eligibility out of this slice unless explicitly ratified.
|
||||
|
||||
Reference in New Issue
Block a user