sshfs/nfs cause server lockup

Chris chrcoluk at gmail.com
Sun Nov 26 05:18:04 PST 2006


On 23/11/06, Kris Kennaway <kris at obsecurity.org> wrote:
> On Thu, Nov 23, 2006 at 04:38:48PM +0100, s.c.sprong at student.utwente.nl wrote:
> > Chris <chrcoluk at gmail.com> wrote:
> > >kris a development on this, someone else posted about a nfs problem
> > >and reading his post some starkling point he made about network cards,
> > >he stated he only gets the bug on sis rl and fxp.
> >
> > Sorry for the misunderstanding, but I think that the 'NFS via TCP' thread
> > covers other bugs, ie the inability to mount NFS v3 over TCP.
> >
> > I've tested the cards above, and the person I replied to encountered the
> > same bug with a bge card. My solution was to remove custom nfs settings
> > in sysctl.conf. I don't know which one was the culprit because I don't
> > have the time to look into it further.
> >
> > My poking uncovered a set of crashing bugs and potentially a livelock.
> > I would agree that NFS is very fragile in RELENG_6.
> > So far <fingers crossed> I've not run into an NFS server deadlock you
> > described.
>
> Are you sure these are NFS problems and not ethernet driver problems?
>
> Kris
>
>
>
The consistency is the servers that lock up are fine when not using
any nfs or sshfs mounts.  But the 2 servers that are fine use dc and
re network adaptors.

I am not able to setup my local bsd box until a new hd arrives next
monday and will be a day or so before I get it setup to mirror the
production conditions of the servers and I will then mount some nfs on
it to repeat the problem.

Chris


More information about the freebsd-stable mailing list