Script to output 404 request summary per virtualhost.

404stats.sh

#!/bin/dash

# Parse access log from a virtual host and output 404 requests.
# Logfile is $LOGDIR/host1.domain.tld/access.log for example.
# If you want this report by mail, set a crontab:
#
# 0 22 * * * 404stats.sh | mail -s "404 report" [email protected]
#

#apache log directory
LOGDIR="/var/log/apache2"

#virtual host domains log directories
SITENAMES="\
host1.domain.tld
host2.domain.tld
"

for www in $SITENAMES; do
	output=$(awk '{if ($9 == 404) {print $7}}' $LOGDIR/$www/access.log \
                | sort -rn | uniq -c | sort -rn)

	if [ -n "$output" ]; then
		echo " * 404 pages for http://$www"
		echo ""
		echo "$output"
	else
		echo " * no 404 request for http://$www"
	fi
	echo ""
done

echo "-- "
echo "$(basename $0)"