I'd like to see a clear answer to this question not talk of rabbits, even though the original poster seems to have given up long long ago.
mysql-cluster (multi master) is not the same as mysql in a (master-slave) cluster, the stock mysql offers scale out but a read-only copy of the master, via the binary replication log, and can not tolerate loss of the master, or promote the slave to a master (without a lot of manual pain). The slave cannot tolerate even a short outage to master (with a high database write rate), the slave will stop receiving changes and become a static snapshot in time, and requires manual re-syncing of the correct position of the change log on the master. So it is not robust at all, except for having a backup, maybe.
mysql-cluster is a different product, it offers redundancy, not just scale-out, if a master dies another server can take on the role, and clients can be failed over (via a content switch, or software appliance, to another server. But as the writer above said, it used the clustered database storage engine ndbcluster (whereas on mysql you have InnoDB / MyISAM the storage engines are compatible and only slightly different) which has some significant syntactical differences.
It would be a relatively simple answer, if the mysql supported was the standard mysql (master-slave) or clustered mysql-cluster (master-master).. if this were true there would need to be a specific schema written for it, as it is different to regular mysql. If there is only a single mysql schema, then I'm assuming mysql-cluster not tested, but it might work with a little (maybe a lot?) of tweaking.
