I know this might not be a very good question to be asked in this site, but I really need some helps/hints to be directed to the right path.
I want to have failover strategy for 2 MySQL servers which are in Master-Master Replication mode. I have done the replication, and it's working just fine. Yet, when in comes to the failover strategy, I'm kind of clueless...
I've been reading on the internet over the past couple of days (I know this is a very complex procedure, and I could not master it in a week or so), but the more I read the more confuse I get. Basically, as I understood, I need to have a software to monitor the network and report any failure and another one to switch between two systems. Heartbeat apparently is the best in doing so, but in all tutorials I read, they are using many other software ( like MON for monitoring, or other software for dumping or getting backups) as well. Is it necessary to have them? Can one have fail-over strategy only using MySQL and HA tool like Heartbeat?
Each website is introducing different tool and ways, which they have their own pros and cons, and of course it depends on how I want my "cluster" to work, and how much high availability is important to my system. I am just considering pointing it out to my customers, and it's not going to the production phase any soon, since my customers running small businesses and don't care about loosing data that much, however, I wanted to dig into the concept by setting up a simple master-master replication where in case of failure of any servers, clients automatically commit changes to the standby server.