WireGuard un-borker

wireguard_healthcheck.sh

#!/bin/bash

STATUS_FILE='/tmp/.wg_fixer_state_unhealthy'
REBOOT_MINIMUM_UPTIME='300' # 5 minutes

SYSTEM_UPTIME=`cat /proc/uptime | cut -d ' ' -f1 | cut -d '.' -f1`

systemctl status wg-quick@wg0 2>&1 >/dev/null
WG_QUICK_STATUS=$?
WG_PEERS=`wg | wc -l`

if [ $WG_PEERS -ne 0 ] && [ $WG_QUICK_STATUS -eq 0 ]
then
	echo "[`date`] WireGuard status appears healthy! Cleaning any status files..."
	rm -f $STATUS_FILE
	exit 0
fi

if [ $WG_PEERS -eq 0 ] || [ $WG_QUICK_STATUS -eq 1 ]
then

	echo "[`date`] Something is wrong with WireGuard :(" 1>&2

	if [ -f $STATUS_FILE ] && [ $SYSTEM_UPTIME -ge $REBOOT_MINIMUM_UPTIME ]
	then
		echo "[`date`] WireGuard service unhealthy state has not recovered, and exceeded uptime threshold. Rebooting..." 1>&2
		rm -f $STATUS_FILE
		/sbin/shutdown -r +0
		exit 1
	fi

	echo "[`date`] Restarting wg-quick systemd units..." 1>&2

	systemctl daemon-reload
	systemctl stop wg-quick@wg0
	systemctl start wg-quick@wg0

	if [ $? -eq 0 ]
	then
		echo "[`date`] Note: wg-quick unit returned 0, service may have recovered ..?" 1>&2
	fi

	touch $STATUS_FILE
fi