* "Automatic" is not possible.
* Need to monitor for failure, then be able to double check that it is not a false alarm.
* DNS for redirecting traffic --> can take many minutes (TTL + caching)
* Port (3306) forwarding -- a significant possibility
* Software or hardware load balancing, such that the passive master appears to be "not
available" -- a significant possibility. Change the health check after establishing it
as the read-write master.
* Different port or different logical host name -- helpful in segregating readonly
connects versus read-write connects.
No, I can't juge which is the best starting point.
MySQL Geeks - Consulting & Review
> -----Original Message-----
> From: Carl Marcinik [mailto:cmonthenet@stripped]
> Sent: Wednesday, May 13, 2009 12:39 PM
> To: replication@stripped
> Subject: advice on master-master replication failover support
> We've been using MySQL for a few years but have only done
> master-slave replication. For several reasons, for an
> upcoming project we want to use master-master replication in
> active/passive mode (active master can read/write, "passive"
> can only read). We need automatic failover support, but it
> has to work for 1) multiple app servers sharing the same
> MySQL master-master backend and 2) redundant networks between
> the app servers and the DB servers (all servers have two
> network ports connected to different subnets).
> I have been looking at MMM Replication Manager and Heartbeat,
> but neither seem to have all that I need. The question is: Is
> either one of these a better starting point for what I want
> to do? Is there a better way (taking into consideration, I
> need to fail over the write role when the active master fails)?
> I would appreciate any sage advice!