A Docker Healer - Auto Restarting Unhealthy Containers

# How To

d-ealer will check the health of all of your containers, if one of them is unhealty, it will restart it.

d-ealer will only restart containers having the label com.dealer.activate set to 1.

This is an example:

docker run -l com.dealer.activate=1 -d -P helloworld:healthcheck

Please adapt the configuration to your need (d-ealer.conf):

[logging]
logger_level = logging.INFO
handler_level = logging.INFO
log_format = %(asctime)s - %(name)s - %(levelname)s - %(message)s
log_file = d-ealer.log

You can find logs in d-ealer.log file. You can mount your log file to your host, if you are using Docker.

docker run -it --name d-ealer -v /var/run/docker.sock:/var/run/docker.sock  -d eon01/d-ealer

This is an issue in docker-py, for me it was solved by:

pip install urllib3==1.14
export PYTHONPATH=/usr/local/lib/python2.7/dist-packages:/usr/lib/python2.7/dist-packages

On docker-py github issues, other people solved this by:

apt-get install python-openssl

or

sudo pip install --upgrade pip

# ToDo

• Allow users to choose the timeout and retries number ..etc

# Links

• Docker Hub: https://hub.docker.com/r/eon01/d-ealer/
• GitHub : https://github.com/eon01/d-ealer