Solaris

In our lab we are in the process of reevaluating the overall setup of our Sun workstations. I'd like to get the opinions of other system administrators on some general strategic questions:

1. If the goal is to try to guarantee identical functioning of all workstations within the lab, which parts of the Sun should be served from one central machine to the others?

2. Anybody have any thoughts on what might be the optimal backup/restore strategy, that is, optimal in the sense of being able to get a crashed machine back up and running in the minimal time?

Thanks very much,
Eric

Simplest soln:

Use NFS for shared apps and home area, and a standard flash jumpstart for provisioning, ~15-20 minutes for recovery of a workstation, unless the central machines dies you'll have nothing locally to backup.