Leveraging Cron job to monitor network connection
Scenario:
There are a handful of ways to run some script periodically on a linux machine. Cron jobs, watch command. Cron jobs are heavily used in server infrastructure for task like regular disk clean up, network monitoring, hard disk storage monitoring, setting up alerts, etc.
I have been running Ubuntu 18.04 and one common issue that I have observed is, in certain wireless connections, after a certain time network-manager stops working. After verifying that I had correct and the updated network drivers installed (lshw -class network
), I had no other option rather than resorting to just re-starting the network-manager. As it started happening frequently I thought of just writing a crontab entry for this task and in the process I learned quite a few things.
First, detecting network status can be quite straight forward. You can just ping an IP or DNS IP (8.8.8.8 Google’s DNS IP in my case).
ping -q -c 1 -W 1 8.8.8.8
If the ping is not succesful we can then restart the network-manager
The final script is:
set -x
if ping -q -c 1 -W 1 8.8.8.8 >/dev/null; then
echo "$(date '+%d/%m/%Y %H:%M:%S') ::network is up"
else
echo "$(date '+%d/%m/%Y %H:%M:%S') ::restarting network-manager"
/usr/sbin/service network-manager restart
sleep 5
ssid=$(nmcli -t -f active,ssid dev wifi | egrep '^yes' | cut -d\' -f2)
if [ "$ssid" = "yes:company_wifi_ssid" ]; then
echo "connected to company wifi.. changing namserver to 10.10.4.17"
echo "nameserver 10.10.4.17">>/etc/resolv.conf
fi
set -x
is there so that I can see all the executables output. I am echoing timestamp and some information to be logged in crontab entry for inspection.
Instead of using service
, I had to use /usr/sbin/service which is where service bin in actually located, (whereis service
). I had to find out the hard way that using fully executable path is preferred while writing jobs.. Since my workplace uses it’s own DNS resolver I am also changing nameserver (10.10.4.17 for example). To accomplish this I am comparing ssid of the connected wifi (after waiting for 5 seconds which turns to be just about right)
Adding crontab entry:
Since starting network-manager requires root privileges, I added it in root’s crontab entry using: sudo crontab -e
* * * * * sh /usr/local/bin/check_network.sh >> /var/log/myjob.log
I appended the output log to a log file inside /var/log, just to monitor and find out the time when network was down and was restarted.
A small section of this log file as run in my system looks something like:
19/09/2018 22:23:01 ::network is up
19/09/2018 22:24:01 ::network is up
19/09/2018 22:25:01 ::network is up
19/09/2018 22:26:01 ::network is up
19/09/2018 22:27:01 ::restarting network-manager
connected to cleartrip wifi.. changing namserver to 10.10.4.17
19/09/2018 22:28:01 ::network is up
19/09/2018 22:29:01 ::network is up
19/09/2018 22:30:01 ::network is up
19/09/2018 22:31:01 ::network is up