On Mon, 2008-07-14 at 16:44 -0400, Todd Denniston wrote: > I can only agree with this comment if the client machines are soft > mounting or > they are working in a network that is incredibly busy. soft mounts > will cause > data loss in a busy network before hard mounts, but if you can ping > flood -s > 1400 the server and not get any loss, then hard,intr UDP NFS should be > still > be _reasonably_ reliable (as NFS goes) and TCP NFS should be fine. +1, but be prepared for client hangs if the *server* goes down for a while. With NFS you can have safety or interactivity, not both. Since client software mostly assumes that when it writes to a file the write will either succeed or fail, it doesn't handle the case of "no reply yet" and just leaves the user in the air. poc -- fedora-list mailing list fedora-list@xxxxxxxxxx To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list
