I created a shell script to register ClamAV in Cron, and when I ran it for confirmation, an error occurred during virus definition pattern file updates.

By the way, the shell script content is as follows:

/usr/local/scripts/clamav.sh

#!/bin/bash

logfile="/var/log/clamav/clamav_$(date +%Y%m%d%H%M).log"
update_logfile="/var/log/clamav/clamav_update.log"
mail_to="notice@localhost"
mail_from="[email protected]"

/usr/bin/freshclam --quiet -u root -l ${update_logfile} > /dev/null 2>&1
/usr/bin/clamscan -r -i --quiet -l ${logfile} \\
--exclude-dir="/sys" \\
/ > /dev/null 2>&1

if [ $? != 0 ]
then
  cat ${logfile} | mail -s "[ClamAV report] $(hostname)" ${mail_to} -- -f $mail_from
fi

Error Content

WARNING: Clamd was NOT notified: Can't connect to clamd through /var/run/clamav/clamd.sock

A warning appeared, so I checked /var/run/clamav/clamd.sock but it didn’t exist.

That makes sense because I haven’t started clamd as a daemon, so freshclam can’t notify clamd after updating the virus definition pattern files.

You can fix this by commenting out the NotifyClamd line in the freshclam.conf file.

/etc/freshclam.conf

NotifyClamd /etc/clamd.conf

↓（Comment out）

#NotifyClamd /etc/clamd.conf

Reference Information

If /etc/freshclam.conf has NotifyClamd /etc/clamd.conf configured, freshclam will try to notify clamd after updating virus definition pattern files However, if clamd is not running, it cannot notify and the above error appears

If you don’t start the clamd daemon, disable the above setting #NotifyClamd /etc/clamd.conf

If the clamd daemon is running and an error appears, check if /tmp/clamd.sock exists If clamd is running and the sock file doesn’t exist

Check if /etc/clamd.conf has LocalSocket /tmp/clamd.sock #TCPSocket 3310

・Source Question Board - Building Home Server with CentOS

That’s all from the Gemba.