Our server had a power supply fail, taking with it many sectors of the 320GB SATA hard drive. The bad sectors contained a lot of the mysql database.
Luckily we had a week old backup of the database and it restored ok. Unfortunately other databases did not fare so well and we are still forced to run under a limping mode innodb_force_recovery = 2 inside of /etc/my.cnf Now we try to identify bad tables and remove them to get the database server to run again.
In the meantime we are learning far more about databases then we would like to, yet not enough to easily identify and fix the problem.