Darek M wrote:
> Hey guys, new member here, go easy on me.
>
> I have two questions to which I didn't really find an answer on the
> page nor the archived lists.
>
> 1. What is clamd, what is its function relating to ClamAV and is it
> required for ClamAV to run? I have ClamAV running on one machine
> without clamd and all seems well. On another, qmail-scanner gives me
> an error as soon as clamd dies. Both use qmail-scanner and both
> scripts use clamdscanner.
>
> 2. clamd dies on me on signal 11 (core dump). Is this a common issue?
> If so, is there a fix? Regardless of the last question, does anyone
> have a solid script that looks for clamd and restarts it if it is down?
> Tried doing my own but it seems flakey. A system call in perl to
> /usr/local/sbin/clamd doesn't bring it back up.
clamd from clamav-0.60 dies ocassionally on me too. I run FreeBSD
4.8-RELEASE on a Dell PowerEdge 4300. I have the core files if anyone
wants to see them (two of them).
I "solved" the problem by running clamd under DJB's daemontools package.
daemontools includes a program called "supervise", which will monitor
the script and restart it if it dies.
I have attached a plain text step-by-step HOWTO guide, roughly after the
style of Life With Qmail. It assumes that you already have clamd installed
and running properly.
Let me know if you like it.
--
Jesse Guardiani, Systems Administrator
WingNET Internet Services,
P.O. Box 2605 // Cleveland, TN 37320-2605
423-559-LINK (v) 423-559-5145 (f)
http://www.wingnet.net
# -----------------------------------------------------------------
# 32.) do the Clam AV post-install configuration
# -----------------------------------------------------------------
vim /usr/local/etc/clamav.conf
# --
# Do the following to the clamav.conf file:
# -----------------------------------------
# 1.) Comment out "Example" line.
# 2.) Uncomment "LogSyslog"
# 3.) Uncomment "StreamSaveToDisk"
# 4.) Uncomment "MaxThreads" and change value to "30"
# 5.) Uncomment "User" and change value to "qmailq"
# 6.) Uncomment "Foreground"
# 7.) Uncomment "ScanMail"
# --
# --
# Create the clamav directory
# --
mkdir -p /usr/local/clamav/bin
# --
# Create the clamdctl script
# --
vim /usr/local/clamav/bin/clamdctl
# *****************************************************************
# NOTE: Place this script in /usr/local/clamav/bin/clamdctl
# *****************************************************************
#!/bin/sh
# For Red Hat chkconfig
# chkconfig: - 80 30
# description: the ClamAV clamd daemon
PATH=/usr/local/clamav/bin:/bin:/usr/bin:/usr/local/bin:/usr/local/sbin
export PATH
case "$1" in
start)
echo "Starting clamd"
if svok /service/clamd ; then
svc -u /service/clamd
else
echo clamd supervise not running
fi
if [ -d /var/lock/subsys ]; then
touch /var/lock/subsys/clamd
fi
;;
stop)
echo "Stopping clamd..."
echo " clamd"
svc -d /service/clamd
if [ -f /var/lock/subsys/clamd ]; then
rm /var/lock/subsys/clamd
fi
;;
stat)
svstat /service/clamd
svstat /service/clamd/log
;;
restart)
echo "Restarting clamd:"
echo "* Stopping clamd."
svc -d /service/clamd
echo "* Sending clamd SIGTERM and restarting."
svc -t /service/clamd
echo "* Restarting clamd."
svc -u /service/clamd
;;
hup)
echo "Sending HUP signal to clamd."
svc -h /service/clamd
;;
help)
cat <<HELP
stop -- stops clamd service (smtp connections refused, nothing goes out)
start -- starts clamd service (smtp connection accepted, mail can go out)
stat -- displays status of clamd service
restart -- stops and restarts the clamd service
hup -- same as reload
HELP
;;
*)
echo "Usage: $0 {start|stop|stat|restart|hup|help}"
exit 1
;;
esac
exit 0
# --
# Create your boot symlink to clamdctl:
# --
ln -s /usr/local/clamav/bin/clamdctl /usr/local/etc/rc.d/clamdctl.sh
# --
# Make clamdctl an executable and link to path:
# --
chmod 755 /usr/local/clamav/bin/clamdctl
chown clamav /usr/local/clamav/bin/clamdctl
ln -s /usr/local/clamav/bin/clamdctl /usr/local/bin
# --
# Create the supervise directories for the clamd service:
# --
mkdir -p /usr/local/clamav/supervise/clamd/log
vim /usr/local/clamav/supervise/clamd/run
# --
# Create the /usr/local/clamav/supervise/clamd/run file:
# --
#!/bin/sh
#
# --------------------------------------------------
# run
#
# Purpose - Start the clamd daemon/service.
#
# Author - Jesse D. Guardiani
# Created - 09/10/03
# Modified - 09/10/03
# --------------------------------------------------
# This script is designed to be run under DJB's
# daemontools package.
#
# ChangeLog
# ---------
#
# 09/10/03 - JDG
# --------------
# - Created
# --------------------------------------------------
# Copyright (C) 2003 WingNET Internet Services
# Contact: Jesse D. Guardiani <[EMAIL PROTECTED]>
# --------------------------------------------------
lockfile="/tmp/clamd" # Location of clamd lock file
path_to_clamd="/usr/local/sbin/clamd"
# Location of the clamd binary
BAD_EXIT_CODE=1 # The exit code we use to announce that something bad has
happened
# The following pipeline is designed to return the pid of each
# clamd process currently running.
get_clam_pids_pipeline=`ps -ax | grep -E "${path_to_clamd}\$" | grep -v grep | awk
'{print $1}'`
# --------------------------------------------------
# Generic helper functions
# --------------------------------------------------
# Basic return code error message function
die_rcode() {
EXIT_CODE=$1
ERROR_MSG=$2
if [ $EXIT_CODE -ne '0' ]; then
echo "$ERROR_MSG" 1>&2
echo "Exiting!" 1>&2
exit "$BAD_EXIT_CODE"
fi
}
# --------------------------------------------------
# Main
# --------------------------------------------------
ps_clamd=""
ps_clamd="$get_clam_pids_pipeline"
if [ -n "$ps_clamd" ]; then
pid_count="0"
for pid in $ps_clamd
do
pid_count=`expr $pid_count + 1`
done
die_rcode $BAD_EXIT_CODE "Error: $pid_count clamd process(es) already running!"
fi
if [ -e "$lockfile" ]; then
rm "$lockfile"
exit_code="$?"
die_rcode $exit_code "Error: 'rm $lockfile' call failed."
fi
exec /usr/local/bin/setuidgid qmailq $path_to_clamd
vim /usr/local/clamav/supervise/clamd/log/run
# --
# Create the /usr/local/clamav/supervise/clamd/log/run file:
# --
#!/bin/sh
exec /usr/local/bin/setuidgid qmailq /usr/local/bin/multilog t s1000000 n20
/var/log/clamd
# Make the run files executable:
chmod 755 /usr/local/clamav/supervise/clamd/run
chmod 755 /usr/local/clamav/supervise/clamd/log/run
# Then set up the log directories:
mkdir -p /var/log/clamd
chown qmailq /var/log/clamd
# Finally, link the supervise directory into /service:
ln -s /usr/local/clamav/supervise/clamd /service
# *****************************************************************
# Note: The clamd script will start automatically shortly
# after these links are created. If you don't want it running
# yet, do the following:
# *****************************************************************
clamdctl stop
