On Sep 12, 2006, at 3:49 AM, Joachim Schipper wrote:
On Mon, Sep 11, 2006 at 06:08:22PM -0500, Doug Carter wrote:
I really doubt that this is a system problem; I just can't figure out
what stupid thing I have done.
Using: OpenBSD 3.9 (GENERIC) #617: Thu Mar 2 02:26:48 MST 2006 on a
Dell 1850, RAID 1 (rest of dmesg below).
One entry every day:
Sep 10 02:16:58 tma0 /bsd: nfs server amd:16867: not responding
As far as I know I don't have NFS running...
I've ignored this for a while today I noticed about 100 instances of
sh, /etc/security, mail & find with the latter in state 'nfsrcv'
This appeared to lead to too many files open and hung the impad
(Dovecot). Killing the find, sh & mail processes and restarting
Dovecot; all appears OK now...
Also, I notice that I can issue a sudo find / -name anything and it
will hang in state 'nfsrcv'
Are you really sure you did not mess with NFS at some point (maybe via
an automounter)? You might even want to reboot, as this will at least
clear any mount-ish confusion in the system.
If this message is logged at the same time every day, tcpdump(8) might
at least confirm that the box is, indeed, trying to do NFS.
Ideally, it
could also give at least a little insight into what it's trying to do.
Finally, try showmount(8) and nfsstat(8) - they might help point
you to
the problem.
Joachim
No resolution yet but I have found a file or device in the root
directory named "list" that acts rather oddly.
It shows up with ls / but ls -l / hangs with the same nfsrcv
state. I can't tell if it is a file, directory or device. rm -rf
list and umount -f list both hang with the (sic) usual 'nfsrcv' state.
Any suggestions short of a restart?
