pg_dump error on hot standy db (msg or pg_toast missing a chunk number).

[lazyPGDBA]

Hi all,
relatively new to Postgres. I'm on 9.0 database instance on CentOS. I was doing a pg_dump of my replicated (hot standby) database (b/c doing it on the master was killing other processes running against it).

My pg_dump failed. I was getting this error:

pg_dump: SQL command failed
pg_dump: Error message from server: ERROR: missing chunk number 0 for toast value 2950316 in pg_toast_1733250
pg_dump: The command was: COPY schema.tablename (col1,col2,...col10) TO stdout;


doing a select * also gives me an error of:
ERROR: missing chunk number 0 for toast value 2950316 in pg_toast_1733250

Doing a search i see that a possible reindex of the table:
REINDEX index pg_toast.pg_toast_1733250;
or REINDEX table pg_toast.pg_toast_1733250;
would help ... but being this is a hot standby database, i'm getting the message:
ERROR: cannot execute REINDEX during recovery

Now on the master database the table is ok. I did an analyze and it was ok. I'm not sure how I can fix this issue, unless i kill the replication or redo the replication. But by doing that, it'll take a very long time.
Any help here would be greatly appreciated as to what my next steps would be.
Thanks.