I have built an active-active failover solution for NFS with shared
scsi. Everything about the failover seems to work fine, but I run into
some occasional problems on nfs service. I do not believe these to be
related to clustering, as I can experience these problems when there are
no kind of failover scenario occuring..
the problem is this:
exportfs hangs sometimes, both when exporting, and unexporting (exportfs
-u). As a note, I do all of my exportfsing on an export basis. I only
have 1 export listed in /etc/exports, just to ensure that nfs runs at
bootup. Aside from that, each export directory is manually exportfs'd by
a script.
Problem 2 is that the nfs server stops responding to clients. Clients
are solaris 8,6, and irix 6.5. It only seems to be one export that stops
responding. There are a total of 3 exports on the machine. Only the ONE
problem export stops functioning. However, the other exports continue to
work. exportfs -a does nothing.
In both problem scenarios, no log messages (/var/log/messages) appear.
I'm baffled. Hopefully, someone has an idea.
I am using Redhat 8.0, linux-2.4.18-27.8.0smp kernel (from redhat)
nfs-utils-1.0.1-2
Thanks for your help,
--
Matt Schillinger
[email protected]
-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs