Chris Sarginson schrieb: > Hi guys, > > I'm looking for some assistance setting up heartbeat in a version 2 > style, I've used it in version 1 style before, but am just having a > bit of a trouble getting my head around this whole set up. > > I have 3 servers behing a loadbalancer, each running web and MySQL > services. Server 1 is the master server and 2 and 3 replicate the > MySQL data off that server. > > I am looking to achieve a set up where if server 1 goes down (ping or > MySQL stops responding) then a VIP (initially on server 1) will fail > over to server 2 AND httpd should be stopped on server 3, BUT if > server 2 is also having issues then server 3 should take on the VIP. > I will be disabling autofailback in order to prevent my data getting > entirely messed up. > > If server 2 or 3 go down at any other point this is not as much of an > issue as they are only slave machines. > > Is what I am looking to do even possible? I've had a look through the > website, read some of the tutorials and watched the video/read the > slides from linuxconf 06, but I am just struggling a little at the > minute. Any assistance would be greatly appreciated. > > Chris > Hi,
no problem to accieve this setup 1) Loadbalancer: Why? Throw it out. Or use LVS with ldirectord as heartbeat resource. 2) Decide if you want to have (a) MySQL and Webserver running on the same machine of (b) some kind of active/active stetup. a) Make both into a group together with the IP address of the cluster. If you want to you can pin the whole group to one node with a location constraint. But this is not nescessary. b) Make a group with MySQL and IP address 1 and an other group with the webserver an the IP address of the cluster. Make a colocation constraint that forbids both groups running on the same node. The webserver will access the database under the first IP address. In both cases you have to think of o good database replication. DRBD is your friend! Have monitoring installed in all ressources and pingd in all nodes. Make constraints that make the ressoruces running on the "healthiest" nodes. Please mail me or the list if you have further questions. Michael Schwartzkopff _______________________________________________ Linux-HA mailing list [email protected] http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
