ADSM-L

Re: ANS401E Stale NFS handle problem

2000-03-03 16:04:59
Subject: Re: ANS401E Stale NFS handle problem
From: Tom Tann{s <tom.tannas AT USIT.UIO DOT NO>
Date: Fri, 3 Mar 2000 22:04:59 +0100
Sorry, my last message was corrupted..

This sounds like a V2-client-problem?

We have a lot of nfs-mounts on our clients, and before moving to V3 this
was a real pain.

You could try to include a virtualmount /xxx in your dsm.sys. This will
cause the client not to try to access the directory at all.
With the exclude.dir, the client will still try to access the directory.

With V3 these problems are more or less solved.

And with the 3.1.0.8-client and I guess the first tsm-client, the
nfs-mounts will not longer be backed up unless you tell the client to do
so...
Haven't tried the .8-client yet. Some clients are scheduled for testing
next week.



Has Anyone else tried/confirmed that the .8 -client really fixes these
nfs-poblems?

On Thu, 2 Mar 2000, Gary Grider wrote:

> I have a Solaris machine that is trying to back up the / filesystem
> which is a local file system.  It has a directory /xxx and
> there was an NFS mounted file system mounted on /xxx, but taht
> mount has become stale.  I am not backing up the /xxx nfs mounted
> file system, but I do want to back up / .
> When I do an incremental on / I get the ANS401E Stale NFS handle problem
> error when it goes by /xxx.  I have tried to exclude /xxx in an exclude.dir
> statement, but it still fails with that error.  Is there any way
> to get the client to go by this directory short of booting the client
> to clean up the stale nfs mount.
>
> Thanks
> Gary
>
<Prev in Thread] Current Thread [Next in Thread>