Both servers are currently running version 3.2 of the Solaris Cluster software. The intent here is to upgrade these to Solaris Cluster 3.3 (SC3.3) which has some new features and added benefits, as well as other general software patches and fixes. The servers will be upgraded using the "live upgrade" method, which essentially builds a copy of the disk image and installs the SC3.3 software on top of that. After all nodes are upgraded, each will need to be shut down. This will require downtime of the services running on the cluster, which includes AFS and mail. The downtime here is estimated to be far less than two hours, assuming no additional hiccups are encountered along the way.