'Lo all- I've got someone on the network who's setup up a small nfs client/server (1 server, 2 clients) using RHL. They mount everything read-only. The trouble is they've been complaining because in general the "clients" in this case are more important and more robust machinesthan the server, and if the server goes down they can't reboot the clients for extremely long periods of time (they freeze at the NFS unmounting stage of the init switch). I realize this is somewhat expected behavior for hard-mounted nfs volumes, and that you can put the intr option in to make NFS respond to interrupts. However, I don't think this'll work when a client is switching runlevels. My first instinct as one who is not an NFS guru was to switch to soft mounts since it's read-only, but lots of people seem to think this is a sin. What does everyone else think of this? Is soft mounting evil? ;-) Is there a better way to handle this situation that I'm not seeing? Thanks! Brian _______________________________________________ techtalk mailing list [EMAIL PROTECTED] http://www.linux.org.uk/mailman/listinfo/techtalk
