Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!swrinde!ucsd!ucbvax!SMITHKLINE.COM!dixons%phvax.dnet
From: dixons%phvax.dnet@SMITHKLINE.COM
Newsgroups: comp.sys.sgi
Subject: problems with automount
Message-ID: <9101231344.AA15399@smithkline.com>
Date: 23 Jan 91 13:44:29 GMT
Sender: daemon@ucbvax.BERKELEY.EDU
Organization: The Internet
Lines: 39

I have been having problems with automount.  The details are a little
involved so first my question:
Has anyone been using automount with Irix 3.3.1 on multiprocessor machines?
If so I would like to hear of your experiences, successful or unsuccessful.

Now for the gory details:
I have been having problems which seem to be related to automount although
the hotline thus far claims that they haven't had any other reports of
problems.  
First observed problems usually are with df.  It prints out the hard (fstab)
mounted disks but hangs up before printing info about automounted disks (my
first clue that something was wrong with automount).  Then gradually (over about
an hour or so) other disk related things (like ls and pwd) begin to hang.  When
a process hangs in this case, you can't interrupt it or kill it although you
can ^Z it and leave it in the background.  Eventually you can no longer log
in to the system (although compute bound jobs seem to continue to run) and
the only thing to do is push the reset button.  Trying to reboot before it
is impossible to log in to the console doesn't seem to help since system
shutdown hangs up in unmounting disks and you still have to reset the system.
This behaviour seems to have nothing to do with what jobs are running on the
system since it occurs when they are running multiple jobs or when they are
empty.  Sometimes it takes a few days to happen and sometimes it happens within
5 minutes of turning on automounter.  If you nfs mount the same set of disks
and servers via fstab then the machine seems to run fine for weeks.  Change
over to automount and boom.  Up until recently this has occured only on a
240GTX.  We have been running a couple of PIs with the same or similar
map files without problems.  Recently we installed five 380 servers and within
a week of firing them up, three have hung up in a similar way.  They are all
running Irix 3.3.1.
The only common thread I can see is that the machines that fail are all
multiprocessor machines while our single processor PIs chug along fine.
There are a number of reasons why using automounter would make administration
of all of these machines a lot easier so I would like to get to the bottom
of this.  In the meantime, I guess I will have to go back to fstab mounts
of the various disks.
If anyone else is seeing similar problems, or is running automount fine
on MP machines, I would like to hear about it.

Scott Dixon (dixons@smithkline.com)