Hello Toasters.

This isn't really a NetApp specific problem, but what better group to tap for NFS info than this. I've got several NetApp Filers serving, among other things, a pair of VALinux boxes running Linux (kernel 2.2.13 and 2.2.14).

The odd thing is that one of the boxes seems to have no problems using the filers. However the other box (2.2.14-5.0.14csmp) keeps throwing fits that look like this:

Mar 7 14:06:37 willy kernel: nfs: server 10.100.2.203 not responding, still trying Mar 7 14:06:37 willy kernel: nfs: server 10.100.2.203 not responding, still trying Mar 7 14:06:39 willy kernel: nfs: task 4788 can't get a request slot Mar 7 14:06:39 willy kernel: nfs: task 4789 can't get a request slot Mar 7 14:06:39 willy kernel: nfs: task 4790 can't get a request slot Mar 7 14:13:07 willy kernel: nfs: server 10.100.2.203 OK Mar 7 14:13:13 willy last message repeated 4 times

The strange thing here is that the filer does not go offline, the other box is using it at the same time with not problems. Also these errors (in this cycle) start when as user starts a transfer and end a few minutes later.

Any ideas? I don't belive this to be a networking issue at this time.

Ben Rockwood brockwood@homestead-inc.com

I am running RedHat Linux 7.1 and had problems with NFS v3 over UDP. I was getting "NFS server not responding" and then "NFS server OK" messages every now and then. No other NFS clients had any problem, including Solaris, IRIX, and AIX. It was particulary annoying when using vi on a NFS mounted file. I think vi auto saves periodically and this would sometimes trigger a NFS hiccup, and vi would freeze for 10 sec or so until NFS recovered. This was very frustrating because vi would hang for 10 sec or so every two or three minutes. After switching to TCP all has been well.

On the filer:

options nfs.tcp.enable on (in case it's off)

On Linux in /etc/fstab:

filer:/vol/volname /dir nfs rw,hard,intr,tcp,bg 0 0

Steve Losen scl@virginia.edu phone: 434-924-0640

University of Virginia ITC Unix Support