| Hi,
Thanks for the response so far. Here is some additional information from
the customer.
Any ideas?
I've appended the tail end of our "messages" file
(it's actually in /var/adm/syslog/messages, not /var/adm/messages).
You can see why these messages are a problem when sent to the console,
just by noticing how many there are, and how frequently they're written.
For the record, here's the fstab entry for /var/spool/mail:
# grep mail /etc/fstab
var_domain#mail /var/spool/mail advfs rq,userquota,groupquota
As you can see, quotas are enabled. I'll append the script which turns
quotas on at boot time.
To demonstrate that quotas really *are* enabled, here's an excerpt from
vrepquota:
# vrepquota -u /var/spool/mail | head
Block limits File limits
User used soft hard grace used soft hard grace
root -- 456 0 0 3 0 0
daemon -- 0 500 5000 1 0 0
bin -- 0 500 5000 1 0 0
uucp -- 0 500 5000 2 0 0
auth -- 0 500 5000 1 0 0
cron -- 0 500 5000 1 0 0
lp -- 0 500 5000 1 0 0
tcb -- 0 500 5000 1 0 0
Oops! The above command also resulted in the following error messages:
DECthreads Last Chance handler: thread 1 exiting on status exception
0x177db013
Exception: Broken pipe (dce / thd)
Abort (core dumped)
I don't know if that's another symptom of this problem, or a sign of a
different problem altogether. :-( I deleted the core dump, but it can be
recreated easily enough if you'd find it helpful.
Similarly, here's a quota report for two (randomly chosen :-) users
with non-zero usage (entries for other file systems removed for the sake
of privacy):
# vquota a_bengho
Disk quotas for user a_bengho (uid 16248):
Filesystem blocks quota limit grace files quota limit
grace
/var/spool/mail 10 500 5000 1 0 0
# vquota a_boula
Disk quotas for user a_boula (uid 27395):
Filesystem blocks quota limit grace files quota limit
grace
/var/spool/mail 544* 500 5000 7days 1 0 0
To summarize, yes, we definitely have quotas enabled and turned on, and
there have been no changes to the domain or fileset configuration since
they were last established under 3.2C.
Excerpt from /var/adm/syslog/messages:
Feb 14 11:38:58 alcor vmunix: /var/spool/mail: warning, user disk quota
exceeded
Feb 14 11:43:03 alcor vmunix:
Feb 14 11:43:03 alcor vmunix: /var/spool/mail: warning, user disk quota
exceeded
Feb 14 11:43:06 alcor vmunix:
Feb 14 11:43:06 alcor vmunix: /var/spool/mail: warning, user disk quota
exceeded
Feb 14 11:44:56 alcor vmunix:
Feb 14 11:44:56 alcor vmunix: /var/spool/mail: warning, user disk quota
exceeded
Feb 14 11:47:38 alcor vmunix:
Feb 14 11:47:38 alcor vmunix: /var/spool/mail: warning, user disk quota
exceeded
Feb 14 11:48:46 alcor vmunix:
Feb 14 11:48:46 alcor vmunix: /var/spool/mail: warning, user disk quota
exceeded
Feb 14 11:49:00 alcor vmunix:
Feb 14 11:49:00 alcor vmunix: /var/spool/mail: warning, user disk quota
exceeded
Feb 14 11:49:04 alcor vmunix:
Feb 14 11:49:04 alcor vmunix: /var/spool/mail: warning, user disk quota
exceeded
Feb 14 11:49:06 alcor vmunix:
Feb 14 11:49:06 alcor vmunix: /var/spool/mail: warning, user disk quota
exceeded
Feb 14 11:50:06 alcor vmunix:
Feb 14 11:50:06 alcor vmunix: /var/spool/mail: warning, user disk quota
exceeded
Feb 14 11:50:08 alcor vmunix:
Feb 14 11:50:08 alcor vmunix: /var/spool/mail: warning, user disk quota
exceeded
Feb 14 11:50:13 alcor vmunix:
Feb 14 11:50:13 alcor vmunix: /var/spool/mail: warning, user disk quota
exceeded
Feb 14 11:53:04 alcor vmunix:Feb 14 11:53:04 alcor vmunix: /var/spool/mail:
warning, user disk quota
exceeded
Feb 14 11:53:05 alcor vmunix:
Feb 14 11:53:05 alcor vmunix: /var/spool/mail: warning, user disk quota
exceeded
Feb 14 11:53:07 alcor vmunix:
Feb 14 11:53:07 alcor vmunix: /var/spool/mail: warning, user disk quota
exceeded
Feb 14 11:53:20 alcor vmunix:
Feb 14 11:53:20 alcor vmunix: /var/spool/mail: warning, user disk quota
exceeded
Feb 14 11:54:49 alcor vmunix:
Feb 14 11:54:50 alcor vmunix: /var/spool/mail: warning, user disk quota
exceeded
Feb 14 11:55:02 alcor vmunix:
Feb 14 11:55:03 alcor vmunix: /var/spool/mail: warning, user disk quota
exceeded
Feb 14 12:05:47 alcor vmunix:
Feb 14 12:05:47 alcor vmunix: /var/spool/mail: write failed, user disk quota
exceeded too long
Feb 14 12:05:52 alcor vmunix:
Feb 14 12:05:52 alcor vmunix: /var/spool/mail: write failed, user disk quota
exceeded too longFile "/sbin/rc3.d/S02LOCALvquota"
(this is actually a symlink to /sbin/init.d/LOCAL.vquota):
8<---------------------------- cut here -------------------------->8
#!/bin/sh
#>>----------------------------------------------------------------
#>> Steven Winikoff, 96/07/18
#>>
#>> This is basically a copy of "S01quota", modified to work on
#>> AdvFS quotas as opposed to UFS quotas. This is intended to be
#>> used *in addition to* S01quota, *not* to replace it.
#>>----------------------------------------------------------------
#
# *****************************************************************
# * *
# * Copyright (c) Digital Equipment Corporation, 1991, 1995 *
# * *
# * All Rights Reserved. Unpublished rights reserved under *
# * the copyright laws of the United States. *
# * *
# * The software contained on this media is proprietary to *
# * and embodies the confidential technology of Digital *
# * Equipment Corporation. Possession, use, duplication or *
# * dissemination of the software and media is authorized only *
# * pursuant to a valid written license from Digital Equipment *
# * Corporation. *
# * *
# * RESTRICTED RIGHTS LEGEND Use, duplication, or disclosure *
# * by the U.S. Government is subject to restrictions as set *
# * forth in Subparagraph (c)(1)(ii) of DFARS 252.227-7013, *
# * or in FAR 52.227-19, as applicable. *
# * *
# *****************************************************************
##
# @(#)$RCSfile: quota,v $ $Revision: 1.1.3.2 $ (DEC) $Date: 1992/09/29
12:08:19
$
#
QUOTA_CONFIG=no
if [ -f /etc/rc.config ]; then
. /etc/rc.config
else
echo "$0 ERROR: /etc/rc.config defaults file MISSING"
exit 1
fi
export QUOTA_CONFIG
#
# setup or shutdown disk quotas
#
case $1 in
'start') if [ "$QUOTA_CONFIG" = 'yes' ]; then
if [ -f /usr/sbin/vquotacheck ]; then
echo "Checking AdvFS disk quotas"
/usr/sbin/vquotacheck -a
else
echo "$0: /usr/sbin/vquotacheck does not exist"
exit 1
fi
if [ -f /usr/sbin/vquotaon ]; then
echo "Enabling AdvFS disk quotas"
/usr/sbin/vquotaon -a
else
echo "$0: /usr/sbin/vquotaon does not exist"
exit 1
fi
fi
;;'stop')
if [ "$QUOTA_CONFIG" = 'yes' ]; then
if [ -f /usr/sbin/vquotaoff ]; then
/usr/sbin/vquotaoff -a
echo "Disabling AdvFS disk quotas"
fi
fi
;;
*)
echo "usage: $0 {start|stop}"
exit 1
;;
esac
exit 0
# HISTORY
[Posted by WWW Notes gateway]
|
| Now I understand. There was, indeed, another change in 4.0X. In 3.2X
versions, a "uprintf" call was made to inform the user that he was over
his quota limit. This function call would only print out the message on
the user's terminal. If there was no terminal present (as would probably
be the case for a mail program), no message would be printed anywhere.
In 4.0X, the "uprintf" call was changed to a "ms_uprintf" call which will
first try to do a "uprintf" but if that fails, it will do an "aprintf"
which is an asynchronous write to the console. This is the behavior
difference that your customer is seeing. My question for him would be:
If you routinely have users exceeding this soft limit, why not raise the
soft and/or hard limit(s)? Otherwise, unless a case can be made that this
is a bug, it would be under the category of design change request.
[Posted by WWW Notes gateway]
|