SUMMARY: automount not responding..

From: Frank Sargent (fsargent@rri.on.ca)
Date: Tue May 12 1998 - 13:11:39 CDT


amazing what oyu can fix by applying the recommended patches to a system...

( still amazes me it worked fine for so long.. )

Thanks to:

Ronald Loftin <reloftin@mailbox.syr.edu>
Colin_Melville@mastercard.com
bismark@alta.Jpl.Nasa.Gov (Bismark Espinoza)
"Marco Greene" <cmgreene@netcom.ca>

original question:
-----Original Message-----
From: Frank Sargent <fsargent@rri.on.ca>
To: sun-managers@ra.mcs.anl.gov <sun-managers@ra.mcs.anl.gov>
Date: Thursday, May 07, 1998 9:00 PM
Subject: automount not responding..

I have a real problem on my hands and I am starting to run in circles..

-mix of machines running 4.1.1 through to 2.5.1 on every different arch.
-NIS server 2.5.1 with no patches yet.. and is the file system server as
well
-user home area automounted via direct map

we will be fine for several hrs - user logs in and their area is
automounted under them - they log out and it times out and gets released...

then after several hrs. the file server will stop answering requests from
the automounter on the requesting system.

a reboot cleans things up again - yippy....

when I run automountd -T on a system I see the mount request go out - then
the
server returns a status=2 when things are not working - a status=0
( and the mount ) when things are good...

I can do a manual mount of the file system fine at any time... and hard
mounting the user areas solves the users problem as well - but not what I
am looking for as I have a large mix of machines IRIX 6.3 to 5.3 - Sun
4.1.1 to 2.5.1 - Linux 2.0.32 to 2.1.98.. all have the same problem...

rpcinfo show nfs and mountd registered.. not a running twice issue I
think...

rpcinfo -u filerserver nfs 2 ( and 3 ) both return
program 100003 version2(3) ready and waiting

when things are good

- but I get an RPC timeout program 100003 not available when things are
bad to the same request above...

the fileserver in question is also a DNS server running bind 8.1.1

I am searching the archives and going through the O'Reilly books... but I
have found nothing that has helped so far...

nothing in the logs indicates anything about why this refusal to mount
would happen - the nfs processes are still there 'runnning' if you list
them, they don't die and dissappear...

I may find the answer in the archives - and I am continuing to search -
just thought I would send this out to get a jump on it.....

----------
Frank Sargent - Manager, Systems and Networks
John P Robarts Reseach Institute
100 Perth Dr.
London, ON
N6A 5K8
Voicemail (519) 663-5777 x4029
Fax (519) 663-3900
http://www.rri.on.ca/~fsargent



This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:12:40 CDT