I have a slony cluster with 2 machines. Our infrastructure detects failure on the master via a heartbeat mechanism, and initiates failover processing on the slave.

During failover processing on the slave, the failover logic first checks to see if it can connect to the master db, and if so does a MOVE SET. Otherwise it does a FAILOVER.

My problem is that the failure could have been from somebody kicking a power cable loose or something, and the master database might be perfectly OK, but because the failover has to happen automagically, without human intervention, I don't have any option but to do a slonik FAILOVER, which means I have to completely throw away a perfectly good master database and re-build it from scratch to get my replication cluster back. This is my understanding of the situation, but I would love to be corrected!

Given these two databases that have been broken apart by a FAILOVER, is there any way to get the old master re-subscribed without loosing all of the data? Assuming there is no existing slonik functionality for this, is there some combination of sl_* table fiddling and slon re-sync that I could implement to make slony accept the old master as a new subscriber without going through the subscribe initial truncation?

- DAP
----------------------------------------------------------------------------------
David Parker    Tazz Networks    (401) 709-5130
?