I am running an NFS server and several FreeBSD NFS clients. The NFS clients
are web servers running Apache. I want to be sure that if my NFS server dies
that I don't get odd problems (e.g., the default hard NFS option will cause
the process to hang, and I can't kill it). I realize that the hard option is
meant to support an NFS environment where a NFS server comes back up, and
everything just pops back to normal, but that hasn't been my experience.

I am willing to accept that if an NFS server dies that my processes will get
errors from the IO operation.

I am currently using these options in /etc/fstab: rw,bg,soft

bg is essential for us since a server will hang on boot-up if an NFS server
is down. It's amazing how if one server is down then chances are a lot of
them are. Murphy's Law.

What options do you feel the most comfortable with?

We are willing to lose a new write to a file if an NFS server dies. It's
more important that I can quickly bring the server back up using our backup
NFS server for example. I can't do that if my process becomes completely
unresponsive.

Just asking for general feelings on this issue.

_______________________________________________
[EMAIL PROTECTED] mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-questions
To unsubscribe, send any mail to "[EMAIL PROTECTED]"

Reply via email to