This script pings google to see if the WAN is alive... if the network is down it restarts my pfSense VM running on proxmox.

restart_pfsense.sh

#!/bin/bash
# file: restart_pfsense.sh

# This is a solution to a problem I have with my pfSense network. When downloading big files often pfsense will stop
# and needs to be restarted. Or, sometimes pfSense will stay running but the WAN goes down for some reason. Rather than
# investigating further I've found that by restarting the pfSense VM (in Proxmox), then the WAN goes back up. Since it takes a minute
# or two to get the VM back up with WAN connection, I have a 10 minute delay to skip the restart if WAN isn't back up. Otherwise you'd
# keep restarting the VM before it had a chance to initialize. Seems to work great!

# Set Crontab to run this every minute: 
# * * * * * /bin/bash /home/justin/scripts/restart_pfsense.sh > /dev/null 2>&1

domain=google.com

vmid=100

# The amount of time to wait between VM restarts - in minutes
time_pause=10

# Output file to write the timestamp to
output_txt_file=restart_pfsense_temp_time.txt

# Output path for the timestamp text file (/var/example.txt)
output_txt_path='/tmp'

# To find out the path of "qm", type "which qm"
qmToUse=/usr/sbin/qm

#----------------------------

ping -c 1 -4 -B -q $domain

RESULT=$?

if [ $RESULT -eq 0 ]; then

    logger -i "++++++++++ ${domain} is UP."

else

    logger -i "++++++++++ ${domain} is Down."

    status=$(${qmToUse} status ${vmid})

    if [[ $status == *"running"* ]]; then

    	logger -i "++++++++++ ${domain} is down, but pfSense is still running, lets restart it."

    	last_time=$(<$output_txt_path/$output_txt_file)

    	if [ $? == 1 ]; then

    		# If getting the last time stamp errs then maybe the file doesn't exist...
    		date +%s > $output_txt_path/$output_txt_file

    		last_time=$(<$output_txt_path/$output_txt_file)

    	fi

		now_time=$(date +%s);

		time_difference_seconds=$((now_time-last_time))

		time_since_last_in_minutes=$((time_difference_seconds/60))

		logger -i "++++++++++ " $time_since_last_in_minutes "minutes has passed since the last automated restart of pfSense."

		if (( $time_since_last_in_minutes >= $time_pause )); then

	   		logger -i "++++++++++ Stopping VM ${vmid}."

	   		$qmToUse stop $vmid

	   		sleep 15

	   		logger -i "++++++++++ Starting VM ${vmid}."

	   		$qmToUse start $vmid

	   		# To void running into a start / stop loop I am saving the Unix time in seconds after the last qm start
	   		date +%s > $output_txt_path/$output_txt_file

	   	else

	   		logger -i "++++++++++ Waiting to restart pfSense."

	   	fi

	else

		logger -i "++++++++++ Apparently VM ${vmid} has stopped, restarting."

		$qmToUse start $vmid

	fi

fi

exit 0

Added temporary file for storing a timestamp of the last VM restart to check against so the VM isn't constantly starting / stopping if the WAN network is down for some other reason. It will still loop & stop / start VM every 10 minutes if WAN is down...