Pardon me but if you write only on one node the it's not multi master :-) Btw, never tried with dbmail but, in general, I have at least two cluster instances of multi master multi writers running since almost 2 years without any particular issues. In one case the replicated cluster is distributed geographically, with two node in a datacenter for wider use and one node in the office for local use. It mainly serves web applications written in php and the load is quite high (and the db is pretty big as well, almost 80gb), although it's clearly different from dbmail's db usage pattern.
I tried the same thing with dbmail (just for the sake of clarity the mta part was running just one node, the external one - the internal one was planned to "proxy/cache" imap connections), and seemed to work quite correctly but I didn't finally implement it because we moved to google services for domain on that particular customer. There are things you have to watch out, like auto increment offset and such, but it's not rocket science... (and have an automated monitoring script that halts everything if replication goes out sync!!) Andrea Brancatelli Il giorno 24/feb/2011, alle ore 10:16, Paul J Stevens <[email protected]> ha scritto: > On 02/23/2011 05:29 PM, [email protected] wrote: > >> Both mysql 5.1.37 databases are replicated on master-master mysql >> replication. about half of the user read their mail on the first >> server and the second half on the other server. The problems happens >> on both. > > > Aaii. Yes, I know some people have reported using this setup without > problems, but really: this is the number one no-go. > > /// > Do not use multi-master setups while writing on more than one node! > /// > > Master-master replication is great, as long as you only do writes on > *one* of the participating nodes. The problem you suffer from is > perfectly explained by collisions and race-conditions. > > try reading this thread for starters: > > http://www.mail-archive.com/[email protected]/msg14937.html > > In your specific case: make sure the daemons on both locations connect > to the same mysql instance. > > I've used mysql-mmm with great success for managing the highly available > IP addresses, if both cluster-nodes are in the same segment. > > > -- > ________________________________________________________________ > Paul Stevens paul at nfg.nl > NET FACILITIES GROUP GPG/PGP: 1024D/11F8CD31 > The Netherlands________________________________http://www.nfg.nl > _______________________________________________ > DBmail mailing list > [email protected] > http://mailman.fastxs.nl/cgi-bin/mailman/listinfo/dbmail _______________________________________________ DBmail mailing list [email protected] http://mailman.fastxs.nl/cgi-bin/mailman/listinfo/dbmail
