I have been reading and researching ways to create a failover system for
our MySQL databases that require as little intervention as possible.
However I am having trouble coming up with a way to get the system back
into a stable state after a failover has occurred and the main master has
been fixed.
The idea is a system along the lines of...
Master (A) -> Standby-Master (B) -> { Slave 1 , Slave 2 , ... , Slave N }
I have defined 2 possible failures, SOFT and HARD. If the master (A)
becomes unresponsive or slow and gets failed out, it would constitue as a
SOFT failure and it would ideally automatically reset the system to its
initial failover capable state when the master (A) has recouperated. A
HARD failure would be anytime the database (A) has crashed and the data
needs to be recreated.
In both cases/failures, the problem I run into is what to do when bringing
the system back into the optimal state without interruption or data
corruption. In degraded mode, writes go to (B) and needs to be switched
back to (A) while keeping replication alive. It can be done with circular
replication but data corruption will happen because of auto increment
fields.
I have found quite a few discussions on the topic of failover setups and
circular replication but haven't found anything that satisfies my needs
yet. Any help/pointers would be greatly appreciated.
Thanks!
Atle
-
Flying Crocodile Inc, Unix Systems Administrator
--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe: http://lists.mysql.com/[EMAIL PROTECTED]
