After setting up 329423 services and 823423 containers, you might want to manage your environment in case some service fails. This can be automated restarting, getting notified about failures or a similar thing. Say no more, I’ve created a Dockerfile and image for Monit for this.

Monit is extremely configurable and allows a maximum of flexibility when it comes to monitoring.

Initially, a configuration file is required to get started. To get one, call

docker run \
   -d \
   --name=monit \
   --restart=always \
   --expose 2812 \
   -p 127.0.0.1:2812:2812 \
   --entrypoint monit ps1337/monit-docker -I && \
docker exec -it monit bash -c "cat /etc/monitrc" && \
docker rm -f monit

and save the output to monitrc.

To finally monitor services on a machine, you can use this command:

docker run \
   -d \
   --name=monit \
   --restart=always \
   -v $PWD/monitrc:/etc/monit/monitrc \
   --expose 2812 \
   -p 127.0.0.1:2812:2812 \
   ps1337/monit:latest

This runs Monit on localhost for security reasons - change this if required.

After doing the initial setup, one can start monitoring for example

• Remote services
• Local containers using --link <ContainerName>
• System load

You can then refer to the Monit manual for information on adding hosts to monitor.

One helpful option to use is exec, which executes shell scripts in case a service fails:

check host yolo with address yolo.party
  if (failed port 80 protocol http with timeout 15 s) retry 5 then exec "/opt/scripts/telegramNotification.sh 'FAIL'"