主检查点记录中的资源管理器 ID 无效

Question

我已将我的 Airbyte 映像从 更新0.35.2-alpha为0.35.37-alpha. [在 kubernetes 中运行]

当系统推出时，db pod 不会终止，我[一个可怕的错误]删除了 pod。当它恢复时，我收到一个错误 -

PostgreSQL Database directory appears to contain a database; Skipping initialization

2022-02-24 20:19:44.065 UTC [1] LOG:  starting PostgreSQL 13.6 on x86_64-pc-linux-musl, compiled by gcc (Alpine 10.3.1_git20211027) 10.3.1 20211027, 64-bit
2022-02-24 20:19:44.065 UTC [1] LOG:  listening on IPv4 address "0.0.0.0", port 5432
2022-02-24 20:19:44.065 UTC [1] LOG:  listening on IPv6 address "::", port 5432
2022-02-24 20:19:44.071 UTC [1] LOG:  listening on Unix socket "/var/run/postgresql/.s.PGSQL.5432"
2022-02-24 20:19:44.079 UTC [21] LOG:  database system was shut down at 2022-02-24 20:12:55 UTC
2022-02-24 20:19:44.079 UTC [21] LOG:  invalid resource manager ID in primary checkpoint record
2022-02-24 20:19:44.079 UTC [21] PANIC:  could not locate a valid checkpoint record
2022-02-24 20:19:44.530 UTC [1] LOG:  startup process (PID 21) was terminated by signal 6: Aborted
2022-02-24 20:19:44.530 UTC [1] LOG:  aborting startup due to startup process failure
2022-02-24 20:19:44.566 UTC [1] LOG:  database system is shut down

很确定 WAL 文件已损坏，但我不知道如何修复此问题。

Answer 1

警告- 可能会丢失数据

这是一个测试系统，所以我不关心保留最新的交易，也没有备份。

首先，我重写了容器命令以保持容器运行，但不尝试启动 postgres。

...
    spec:
      containers:
        - name: airbyte-db-container
          image: airbyte/db
          command: ["sh"]
          args: ["-c", "while true; do echo $(date -u) >> /tmp/run.log; sleep 5; done"]
...

并在豆荚上产生了一个壳 -

kubectl exec -it -n airbyte airbyte-db-xxxx -- sh

跑步pg_reset_wal

# dry-run first
pg_resetwal --dry-run /var/lib/postgresql/data/pgdata

成功！

pg_resetwal /var/lib/postgresql/data/pgdata
Write-ahead log reset

然后去掉容器中的temp命令，postgres就正确启动了！