thanks, i will try to do that... even though i already did the cluster using sarge as loadbalancers and etch as web servers...
but is there a way of repplicating data without using rsync mirror?? For example, if server1 goes down and i make changes on server2 files, then server1 comes back again, how to replicate the data from server2 to server1, because rsync only replicates data from server1 to server2(mirror), and is there a way to do this without having to install a HA-NFS server??
Well in the other hand, about the loadbalanced MySQL Cluster
In your howto says that when i create a database table that i want it to be clustered, i have to use ENGINE=NDBCLUSTER, and then create the same table in the other node without using ENGINE=NDBCLUSTER in order to have data replicated, right? but.... what would happen if for example i was running a hosting server with apache clustered and mysql clustered and for example a hosting-user creates a database table without using that engine and doing it thru phpmyadmin??
1. would the data be replicated??
2. wouldn't be easier to just use mysql database repplication
and use the ultramonkey load servers to simulate the cluster like it is done with apache high availavility cluster???
3. or should i use mysql database replication
additionally to mysql cluster configuration, having both configurations at the same time???
4. Is there a way to install a high-availability load balanced MySQL cluster using DRDB (because most of the articles i've read use MySQL and DRDB no as a loadbalanced cluster, they use it as a backup server like your howto of HA-NFS server)
thank you for your help