Thread: pg_proc_proname_args_nsp_index error

I've got a standard setup on Debian woody, with postgreSQL backports
from http://people.debian.org/~elphick
Last night, I started getting mails from the cron daemon that this command:
/usr/bin/test -x /usr/lib/postgresql/bin/do.maintenance &&
/usr/lib/postgresql/bin/do.maintenance -a -F

was spitting out this errormessage:

ERROR:  could not read block 6 of relation
"pg_proc_proname_args_nsp_index": Input/output error


I'm not really sure what's going on here, and how serious this is.
Anyone got idea what this is, and how it can be fixed? Alternatively,
links/urls to somewhere online where I can read up on this?

Tommy

Tommy Gildseth <gildseth@start.no> writes:
> ERROR:  could not read block 6 of relation
> "pg_proc_proname_args_nsp_index": Input/output error

> I'm not really sure what's going on here, and how serious this is.

It's not good --- it means your disk has developed an unreadable block.
You can get out of this particular problem by REINDEXing pg_proc, but
the larger question is whether your disk is on the brink of worse
failures.  Might be time to buy a new one.

            regards, tom lane