Github-Mirror/seaweedfs
1
0 Fork 0
mirror of https://github.com/seaweedfs/seaweedfs.git synced 2026-10-07 06:47:51 +02:00
Code Issues Packages Projects Releases Wiki Activity
13,263 Commits 188 Branches 361 Tags
fbcfe89e2461d42158cdc5e678d64e281ec7d2c8
Commit Graph
2 Commits
Author SHA1 Message Date
pingqiuandClaude Opus 4.6 e0116fc631 fix: three hardware blockers — WAL retention + registry race + shutdown beat 
All 43 actions pass on m01/m02 hardware. Auto-failover PASS.
dd_write: 30s → 123ms. Post-failover write: 33,621 IOPS.

1. WAL retention: remove keepup retention floor (MinShippedLSN).
   WAL cannot be pinned during sustained async writes — any pin
   strategy either fills WAL (blocking writes) or over-recycles
   (breaking catch-up). Flusher recycles freely. Future LBA map
   will provide catch-up without WAL retention.
   MinShippedLSN on ShipperGroup retained as diagnostic surface.

2. Registry stale-cleanup race: add RegisteredAt grace period.
   Race: master registers volume → next VS heartbeat arrives before
   VS discovers the volume → stale cleanup deletes the entry →
   failover finds 0 entries. Fix: skip stale cleanup for entries
   registered within 30s (> 2 heartbeat intervals).
   2 new tests: grace protects new entry, old entry still cleaned.

3. Shutdown heartbeat: VS disconnect heartbeat no longer claims
   block inventory authority. Previously, the shutdown beat's
   empty inventory triggered stale cleanup, deleting the entry
   before failover could use it.

Scenario fix: recovery-baseline-failover.yaml now kills the
correct node (discovered primary, not hardcoded), connects to
the correct new primary for post-failover verification.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
 2026-04-08 22:59:46 -07:00
pingqiuandClaude Opus 4.6 b8c6944e3f feat: V2 MVP milestone — masterv2 + volumev2 + in-process failover 
V2 runtime packages:
- sw-block/runtime/masterv2: identity authority (desired state,
  heartbeat handling, promotion arbitration via SelectPromotionCandidate)
- sw-block/runtime/volumev2: per-volume micro-cluster shell (node,
  orchestrator, control session, iSCSI frontend, takeover gate,
  failover session + driver, replica summary reconstruction)
- sw-block/runtime/purev2: RF1 execution shell (engine + store +
  dispatcher + local boundary observations)
- sw-block/runtime/protocolv2: three-channel separation
  (heartbeat/assignment/query + replica summary)

V2 binaries:
- sw-block/cmd/v2singleblock: single-node RF1 block server
- sw-block/cmd/purev2rf1: minimal RF1 runtime binary

Milestone capabilities:
- RF1 write/read/sync with engine-driven mode projection
- masterv2 ↔ volumev2 heartbeat convergence + assignment reissue
- Promotion query with fresh CommittedLSN/WALHeadLSN evidence
- Replica summary for bounded takeover reconstruction
- Primary-loss reconstruction from peer summaries (fail-closed gate)
- In-process failover driver with session observability
- Local boundary observations feed engine (Committed/Durable/Checkpoint)

Design docs:
- v2-two-loop-protocol.md: identity vs data-control separation
- v2-automata-ownership-map.md: event/command ownership split
- v2-loop1-surface-draft.md: heartbeat/query/assignment field spec
- v2-volumev2-single-node-mvp.md: target layering
- v2-kernel-closure-review.md: per-volume micro-cluster principle
- v2-pure-runtime-rf1-bootstrap.md, v2-capability-map.md,
  v2-proof-and-retest-pyramid.md

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
 2026-04-05 13:08:02 -07:00