Hello fine NFS folks,
I'm running an OpenSolaris server exporting ~3000 ZFS shares (yes, it
takes /forever/ to start...)
Until today, it's been running mostly without a hitch on snv_101b. I
decided to pkg image_update and reboot, and now all I can get is stuff
like this:
[ Jun 4 19:53:08 Executing start method ("/lib/svc/method/nfs-server
start"). ]
[ Jun 4 19:56:59 Method "start" exited with status 0. ]
[ Jun 4 19:57:03 Stopping because process dumped core. ]
[ Jun 4 19:57:03 Executing stop method ("/lib/svc/method/nfs-server
stop 103"). ]
... continues bouncing up and down infinitely
I thought to myself, "Well no problem - I have a snapshot from
before, I'll just roll back!" Of course, if that had worked, I
wouldn't be here... it now exhibits the same problem. Crashes after
about 4 seconds of running.
I'm wondering how I can go about debugging this. On a BSD/Linux
system, I would run nfsd or whatnot with a -d or a -v flag or
something, to try to capture debug output. However, nfsd(1M) does not
mention any such useful flags to get more output from it. Dmesg says:
nfssrv: [ID 939466 kern.warning] WARNING: nfsauth: mountd has not
established door
which may or may not be important... Google says nothing on the
subject, unfortunately, and searching for something as generic as
"nfsd core dumps" is not very helpful.
Any way I can gather more information, or perhaps someone out there
can pinpoint the problem?
Thanks for any help you can provide!
Steven Schlansker
