>> Or just use a RAID array (eg software RAID in mirror mode: RAID1). >> > > RAID IS NOT BACKUP > > To the OP: > It would be helpful if you were more descriptive about what you are > trying to accomplish. Are you worried about disk failures? Are you > worried about the whole system failing? What about the case where > invalid data is added to the database? (or the database gets corrupted > in general?) Do you want to have a hot backup of the database > standing by so you can switch to it if the main one goes down? Each > of these have different solutions. > _______________________________________________ > CentOS mailing list > CentOS at centos.org > http://lists.centos.org/mailman/listinfo/centos 1. === RAID IS NO BACKUP! RAID is only to survive hardware failure of the hard disk(s) (and only if you don't use RAID0!). Other people are mentioning the Master/Slave setup. This will do if you just need one up to date backup for the coincidence the complete master server fails but not for the occasion someone / something enters the wrong SQL query and deletes the wrong data! Then the slave server will also execute this SQL query and deletes the data also! If you want to have the opportunity to go back in time, then you have to make dumps through mysqldump. 2. === The bigger the data, the slower the backup will be, the longer the tables will be locked, the greater is the change your users will notice it with using the application. Best regards, -- Joost Waversveld