NetNews Usenet Archive 1992 #16/NN_1992_16.iso/spool/comp/protocol/appletal/2863

Internet Message Format | 1992-07-21 | 3.7 KB

Path: sparky!uunet!munnari.oz.au!yoyo.aarnet.edu.au!sirius.ucs.adelaide.edu.au!escher.arch.adelaide.edu.au!coppins
From: [e-mail] (Simon Coppins)
Newsgroups: comp.protocols.appletalk
Subject: AUFS servers slowly die
Keywords: CAP, aufs
Message-ID: <[e-mail]>
Date: 22 Jul 92 07:41:50 GMT
Sender: [e-mail]
Organization: University of Adelaide
Lines: 63
Nntp-Posting-Host: odin.arch.adelaide.edu.au

I have been seeing aufs servers closing down from apparent boredom. I run two
aufs servers on a Sun SPARCstation 1+ running SunOS 4.1.1 (very patched). One
of the aufs seems happy to stay up and run many sessions each day with no
problem. The other aufs (serving a different filesystem) can only keep client
sessions up sometimes for as little as a minuite. These two processes are
running from the same binary, very strange. The output from the log file of the
sick aufs looks typically like this:

13:35:17 07/22/92  Waiting for session 6 to activate
13:35:17 07/22/92  Login requested for bootapps (we are running as root)
13:35:18 07/22/92  Login: user bootapps, home directory /usr/tmp
13:35:18 07/22/92  Using protocol 'AFPVersion 2.0'
14:06:47 07/22/92  Server timeout on session 2 pid 5956, not talking to remote a
nymore
14:06:47 07/22/92  Server timeout on session 5 pid 6230, not talking to remote a
nymore
14:06:47 07/22/92  Superior told us to shutdown - probably tickle timeout
14:06:47 07/22/92  Server timeout on session 4 pid 6017, not talking to remote a
nymore
14:06:47 07/22/92  Recorded terminated inferior Aufs PID 6230
14:06:47 07/22/92  Server timeout on session 1 pid 11781, not talking to remote
anymore
14:06:47 07/22/92  Superior told us to shutdown - probably tickle timeout
14:06:47 07/22/92  Server timeout on session 3 pid 11702, not talking to remote
anymore
14:06:47 07/22/92  Server timeout on session 0 pid 5538, not talking to remote a
nymore
14:06:47 07/22/92  Superior told us to shutdown - probably tickle timeout
14:06:47 07/22/92  process 6230, session 5 was terminated due to a timeout
14:06:47 07/22/92  Superior told us to shutdown - probably tickle timeout
14:06:47 07/22/92  22 messages out, 22 in, CPU 59.22 user 172.06 system
14:06:47 07/22/92  Process 6230 terminated
14:06:47 07/22/92  Recorded terminated inferior Aufs PID 11781
14:06:47 07/22/92  process 11781, session 1 was terminated due to a timeout
14:06:47 07/22/92  6 messages out, 0 in, CPU 0.51 user 1.70 system
14:06:47 07/22/92  Process 11781 terminated
14:06:47 07/22/92  Recorded terminated inferior Aufs PID 5538
14:06:47 07/22/92  process 5538, session 0 was terminated due to a timeout
14:06:47 07/22/92  20 messages out, 8 in, CPU 24.24 user 90.57 system
14:06:47 07/22/92  Process 5538 terminated
14:06:47 07/22/92  Superior told us to shutdown - probably tickle timeout
14:06:47 07/22/92  Superior told us to shutdown - probably tickle timeout
14:06:47 07/22/92  Recorded terminated inferior Aufs PID 5956
14:06:52 07/22/92  process 5956, session 2 was terminated due to a timeout
14:06:52 07/22/92  32 messages out, 57 in, CPU 78.58 user 224.94 system
14:06:52 07/22/92  Process 5956 terminated

Lots of timeouts.

I've been running CAP6.0pl100 for months without any problems. I won't go into
any detail about our network as the problem seems quite independant of MAc type
or EtherTalk/LocalTalk connection. I have not carefully checked the client
lifetimes to see if there is any pattern.

The question: Has _anyone_ seen anything like this?


Simon.
--
|Simon Coppins                  Phone: [phone]
|Computer Systems Administrator Fax:   [phone]
|Department of Architecture     Email: [e-mail]
|University of Adelaide         Postal:GPO box 498,Adelaide SA 5001,Australia