Issues - go-kv - Gitea: Git with a cup of tea

dailz/go-kv

Labels Milestones New Issue

[Important] publishedSequence 需要明确的内存序约束

#14 by dailz was closed 2026-06-11 20:36:15 +08:00

[Important] Batch 校验缺少资源上限

#13 by dailz was closed 2026-06-11 20:27:18 +08:00

[Important] 尾部截断后缺少持久化步骤

#12 by dailz was closed 2026-06-11 20:17:59 +08:00

[Important] CURRENT 文件权威性与实际恢复模型不一致

#11 by dailz was closed 2026-06-11 20:09:55 +08:00

[Important] MemTable 写入失败（Arena 满）在 WAL 写入后未覆盖

#10 by dailz was closed 2026-06-11 19:53:25 +08:00

[Critical] Go 内存模型：lock-free read 需要原子发布机制

#9 by dailz was closed 2026-06-11 18:30:07 +08:00

[Critical] 恢复后未确认 Batch 可能变成已发布

#8 by dailz was closed 2026-06-11 18:15:21 +08:00

[Critical] 预创建 segment 可破坏尾部截断逻辑

#7 by dailz was closed 2026-06-11 17:52:35 +08:00

[Critical] Segment 边界跨 Batch 行为未定义

#6 by dailz was closed 2026-06-11 17:33:52 +08:00

[Critical] WAL write failure 语义过于简化：write() 失败后 bytes 可能已落盘

#5 by dailz was closed 2026-06-09 13:38:17 +08:00

3.2 WAL：补充 ErrCommitUnknown 的 API 层提交状态确认约束

#4 by dailz was closed 2026-06-09 11:39:14 +08:00

3.2 WAL：评估 Value Log 先 fsync 再 WAL fsync 对写吞吐的影响

#3 by dailz was closed 2026-06-09 11:39:14 +08:00

3.2 WAL：明确 write-stopped 后 pending/aborted entry 生命周期

#2 by dailz was closed 2026-06-09 11:39:14 +08:00

3.2 WAL：澄清 WAL write failure 与 MemTable pending entry 的时序

#1 by dailz was closed 2026-06-09 11:39:14 +08:00