May be I am misunderstanding your goals but it seems that you want to have a backup server for two independant masters (with different databases). Something like:
M1 --> S <-- M2 DB1 DB1 DB2 DB2
If this is the case, you can run two mysqld instances on S, each one replicating from one master: M1 --> S1 DB1 DB1 S2 <-- M2 DB2 DB2
Hope this helps Joseph Bueno
Erik Olsen wrote:
Yes. The idea was to backup 2 masters, 1 that is ours and 1 that is a
costumer. The slave's job is just going to have a synchronised db of
both servers db. The plan was to have it on a different place in case of
fire.
But I must find another solution then.
Erik Olsen wrote:
Is it possible for slave to connect to 2 different masters and have synchronized database from both?
So you would have updates on 2 masters M1 and M2 which would be replicated to the read-only slave S1 ?
The point of MySQL's replication is that after an replication-event there is allways a moment in time where slave == master.
In your model you'd build : M1 --> S <-- M2
An update of M1 would be carried to S but wouldn't be transferred to M2 so S had no chance to get in the state S == M2 anymore.
I suppose that'd break the replication process and it'd stop.
You have to do it in a circle as the manual describes :
... --> M1 --> M2 --> M3 --> M1 ...
I'd rather have it like a star formation but that seams to be impossible, too.
Updates on all machines with replication to a central supermaster SM
M1 <--> SM <--> M2
This way I wouldn't rely on all hosts in the replication-circle to stay up and do their job since one is off site and hanging on a slow dial-up line.
-- MySQL General Mailing List For list archives: http://lists.mysql.com/mysql To unsubscribe: http://lists.mysql.com/[EMAIL PROTECTED]