-->-----Original Message-------> --> Has anyone done this with two (or more, if possible!) machines? Is it -->possible to do at the present time?
Your going to need to have a chain master setup. This is what I propose. 1) Application(s) must know about all the master / slaves (or hide it behind a VIP 2) Configure the Master and Sub-Master to log it's slave queries. 3) If the Master goes down then the sub master will maintain all the writes && vice versa. 4) Have the binary logs located on a shared drive or mirror it across the network. This will allow slave to connect to new masters in the event the original master goes down. 5) All sub masters must use the same binary log file and respect the rotation. This is related to 4. 6) software needs to be written to manage this all Problems with this. 1) Primary key violations on Master to Sub-Master 2) Lots of gaps and assumptions. This is generally the approach I have taken. One day I will write a white paper up on it and publish my ugly code that manages failover etc. --> --> -->-- -->MySQL General Mailing List -->For list archives: http://lists.mysql.com/mysql -->To unsubscribe: -->http://lists.mysql.com/[EMAIL PROTECTED] -- MySQL General Mailing List For list archives: http://lists.mysql.com/mysql To unsubscribe: http://lists.mysql.com/[EMAIL PROTECTED]
