On Wed, 2010-03-17 at 11:10 -0500, Neil Aggarwal wrote: > John: > > > Have you thought of separating the databases? One for the > > reads and one > > for the write on different raids? Despite what some may believe this > > can be done. > > Our goal is to create redundancy. We want either system to be able > to work if the other is not available. Designating one database > as a write db and the other as a read defeats that. > No it does not. You can have what ever you want: and it will work with brilliance. Master/Writer > Replicate/GEO/Writer Master/Reader > Replicate/GEO/Reader
The Master writer can be on server1 the master reader server 2 then do your replication. You just have to take account for this in the design of you code connection and query wise. Ohh a heads up most python database modules can support this way of connecting. IE They are not connect aware to more than one connection. Meaning you can mix inserts and selects from the write db and then from the read db at once. Nates comment in the latter mail is correct also in doing it. I do this all the time. Just not with your truly (mysql) John _______________________________________________ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
