We have a production issue where MySQL has started spitting InnoDB corrupt page errors and falling over.
The error message seem to relate to indexes so I assume the data itself is ok and when I restarted the server with innodb_force_recovery=1 I was able to dump the lot using mysqldump.
However, re-importing hasn't fixed the error, dropping/recreating the database didn't help either and I've even gone as far as moving the innodb files out the way and letting MySQL recreate them before re-importing.
No joy, corruptions keep re-occuring within a few minutes and the server dies again.
The weird thing is the database is a mix of ISAM and InnoDB tables and I had a couple of errors in the ISAM tables as well on some attempts.
Is this likely to be something data-related, something corrupt deeper in MySQL that isn't cleared by the above or signs of a hardware fault underneath it all?
EDIT: This is Mysql 5.0.51 running on Ubuntu 8.04