This content has been marked as final. Show 2 replies
In this scenario how are you going to start a second node UCM server when the first is already running ? In your case since this is a cluster set up all the nodes would share the same config.cfg and have same IntradocServerPort . So at any 1 point in time only 1 node would be up and running and the whole purpose of clustering (if it is for HA / Performance ) is defeated .
I think you should first do one step and explain what you want to achieve.
a) it is possible to install two instances of UCM on a single physical server - if you use two domains. If I remember, you will need 2 admin servers, though.
This scenario makes sense if you want to use the same HW e.g. for testing/training/ development purposes. I'd discourage you, however, to pack/unpack domains to migrate environments (maybe this initial copy might be OK, but then environments should use Content Server's migration utilities for config/data/etc.) This is also not a config for a production environment.
b) if your aim is more disaster recovery, I think you won't need two domains, but rather should look into clustering. By the means of Weblogic server you should be able to create either cold-failover (with a single node, capable of self-recovery), or hot-failover (two or more nodes in a cluster, where a failing node can be replaced by others without losing your sessions). Note that hot-failover on a single physical server won't prevent cases when the whole server goes down.