> Steven N. Hirsch wrote:
> > Some gathered information on my ongoing NFS problems.
> > [... many tests]
> > Conclusions:
> >
> > By elimination, I would conclude that something is broken in the
> > 2.1.55 kernel-based NFS server.
> >
> > For the umpteenth time:
> >
> > This stuff is 100% repeatable. If there are any debugging hooks you would
> > like me to put in, or any patches to try, please say the word!
>
> Hi Steven,
>
> After studying the code of the nfs server a while back I posted messages
> expressing doubts that the dentry-as-filehandle could possibly work, and
> the symptoms you've found (stale filehandles, random failures, possible
> trashing of files) are what I would expect.
>
> I have some ideas for fixing this, but in the shorter term would rather
> focus on getting nfs client working reliably.
>
> What is your assessment of the current state of nfs client under 2.1.xx?
>From everything I've seen in my latest set of tests, I'll conditionally
give the client a clean bill-of-health. I've rebuilt the system on both
boxes to use the knfs client with the older user-based server. My
intention is to beat the daylights out of it in both directions over the
next few days.
I'll get back to you if there are any problems that can reasonably be
blamed on the client.
Thanks very much for your attention to the problem, Bill. I'm absolutely
amazed that no one else is reporting it, though it's possible that only
the low transaction-latency of the 100Base connection triggers it. Not
too many folks are using 100Base (compared to 10Base).
Steve