Hi,
We have an issue with a database on IP SAN disks (Solaris x64), where
some tables and indexes got corrupted, apparently in correlation with
the use of forcedirectio in the nfs for data and indexes:
ERROR: unexpected data beyond EOF in block ...
HINT: This has been seen to occur with buggy kernels; consider
updating your system.
Now I'm interested in a way to get the list of all corrupted objects.
The only way I know is to check the errors of pg_dump or vacuum.
Is there a faster way, something like "vacuum check_only" ?
Thanks,
Marc Mamin