Hacktoberfest 2026: los issues que los mantenedores marcaron para octubre, abiertos y aptos para principiantes. Explorar issues de Hacktoberfest

[Bug] KV recovery fails after all TabletServer local data is lost

Abierto
#4,566 0 comentarios 0 reacciones 1 asignado Ver en GitHub

Los mantenedores suelen responder en 2 días

@swuferhong ya está trabajando en esto.

Desde el 8/10/2026.

  • #4567 de @swuferhong — abierto

Evaluación

Este issue todavía no se ha evaluado.

Descripción

Search before asking
  • I searched in the issues and found nothing similar.
Fluss version

1.0.0 (latest release)

Please describe the bug 🐞
Description

KV tables may fail to recover after all TabletServer local data is lost, even when ZooKeeper metadata, completed KV snapshots, and the required remote WAL are still available.

An empty local log is initialized at offset 0. However, KV recovery may start from a snapshot at a higher offset. Without restoring the local log boundary from remote durable state, recovery can attempt to read an unavailable local offset and fail.

Failed recovery attempts may also leave an initialized KV tablet registered. Subsequent retries can then fail with a Duplicate kv tablet directories error, obscuring the original failure.

Steps to reproduce
  1. Create a primary-key table with multiple replicas and write some records.
  2. Complete a KV snapshot and upload subsequent WAL segments to remote storage.
  3. Write additional records that remain outside both the snapshot and the uploaded WAL.
  4. Stop the cluster and remove all TabletServer local data, including KV files, WAL, checkpoints, and clean-shutdown markers. Preserve ZooKeeper and remote storage.
  5. Restart the cluster and access the table.
Actual behavior

KV recovery fails because the snapshot recovery offset is inconsistent with the empty local log. Recovery retries may additionally fail because partially initialized KV state was not cleaned up.

Expected behavior

When the snapshot and required remote WAL are available, the table should recover to a consistent durable boundary and support subsequent reads, writes, replication, snapshots, and log tiering.

Writes beyond that boundary may be lost, including previously acknowledged writes that existed only on the lost local replicas.

Missing or truncated WAL required for recovery should cause an explicit failure rather than silently produce incomplete KV state.

Solution

No response

Are you willing to submit a PR?
  • I'm willing to submit a PR!
Lenguaje dominante
Java
Estrellas
2.2k
Forks
658
Merge medio
3 d 4 h
PR fusionados (30 d)
106

Preparar el entorno

  • Sin Dockerfile ni archivo de Docker Compose
  • Tiene una plantilla de pull request
  • Sin guía de contribución

Primeros pasos

  1. Lee el issue completo y luego la guía de contribución del proyecto.
  2. Comenta en el issue que vas a ocuparte — evita que dos personas hagan lo mismo.
  3. Haz un fork del repositorio y trabaja en una rama.
  4. Abre un pull request que haga referencia al número del issue.

Más de apache/fluss

Todos los issues de apache/fluss

Issues similares

Más issues de Java

Recibe los nuevos issues en tu correo

Un resumen breve de issues de GitHub para principiantes.