I am testing a cluster that would have the following setup: Server1 - production postgres server Server2 - production postgres server Server3 - production postgres server Server4 - hot standby for any of the 3 production postgres servers So there would be 3 failover domains and Server4 would be in all of them. Then there would be 3 services, one in each failover domain. This may not be a recommended setup but I think the odds of 2 production servers failing at the same time are low enough to make this worth looking at. If they do, the first service to fail over would succeed and the other would fail. I would have issues one way or another if more than one server failed at a time. In setting this up I get a pricmary attribute collision since the shared file system mount point is the same for more than one file system resource (even though they are private resources). Obviously logic is built in to see that Server4 would have a problem mounting to the same mount point more than once. Has anyone tried this or have a suggestion for a workaround? I thought about unique mount points and then adding a script to each service that would set up a sym link, but there might be a more elegant way. Thanks, Brad Crotchett brad@xxxxxxxxxxxxxx http://www.bradandkim.net -- Linux-cluster mailing list Linux-cluster@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/linux-cluster
