postgresql 错误 PANIC : could not locate a valid checkpoint record

Question

当我加载 postgres 服务器 (v9.0.1) 时，我遇到了一个阻止它启动的 panic :

PANIC: could not locate a valid checkpoint record

我该如何解决这个问题？

Answer 1

它正在事务日志中寻找可能不存在或已损坏的检查点记录。您可以通过运行来确定是否属于这种情况:

# Postgres >= 10
pg_resetwal DATADIR

# Postgres < 10
pg_resetxlog DATADIR

如果事务日志损坏，您会看到如下消息:

The database server was not shut down cleanly.  
Resetting the transaction log might cause data to be lost.  
If you want to proceed anyway, use `-f` to force reset.

然后您可以按照说明并使用 -f 运行以强制更新:

# Postgres >= 10
pg_resetwal -f DATADIR

# Postgres < 10
pg_resetxlog -f DATADIR 

这应该会重置事务日志。但是，它可能会使您的数据库处于不确定状态，如 PostgreSQL documentation on pg_resetwal 中所述。 :

If pg_resetwal complains that it cannot determine valid data forpg_control, you can force it to proceed anyway by specifying the-f (force) option. In this case plausible values will be substitutedfor the missing data. Most of the fields can be expected to match, butmanual assistance might be needed for the next OID, next transactionID and epoch, next multitransaction ID and offset, and WAL startinglocation fields. These fields can be set using the options discussedbelow. If you are not able to determine correct values for all thesefields, -f can still be used, but the recovered database must betreated with even more suspicion than usual: an immediate dump andreload is imperative. Do not execute any data-modifying operations inthe database before you dump, as any such action is likely to make thecorruption worse.