sushibox/autoheal repository overview

⁠Docker Autoheal 2

⁠A better Autoheal

Docker Autoheal is a tool that monitors and restarts unhealthy Docker containers. It ensures that your containers are always running in a healthy state by detecting any failures and automatically restarting them.

The script in this version of Autoheal has been improved to incorporate changes such as:

⁠Logging

Added a logging mechanism to append log messages to the specified LOG_FILE for future reference.

⁠Notification

Implemented email notifications using the mail command and the NOTIFICATION_EMAIL environment variable. Also added Discord webhook notifications using the DISCORD_WEBHOOK_URL environment variable.

⁠Container Exclusion

Added support for excluding specific containers from autohealing using the EXCLUDED_CONTAINERS environment variable.

⁠Scheduling

Introduced time window configuration with the SCHEDULE_START and SCHEDULE_END environment variables for enabling or disabling autohealing based on a predefined schedule.

⁠How to Use

⁠UNIX Socket Passthrough

    docker run -d \
        --name autoheal \
        --restart=always \
        -e AUTOHEAL_CONTAINER_LABEL=all \
        -v /var/run/docker.sock:/var/run/docker.sock \
        -v /etc/localtime:/etc/localtime:ro \
        docker.io/sushibox/docker-autoheal:latest

⁠TCP Socket

    docker run -d \
        --name autoheal \
        --restart=always \
        -e AUTOHEAL_CONTAINER_LABEL=all \
        -e DOCKER_SOCKET=tcp://HOST:PORT \
        -v /etc/localtime:/etc/localtime:ro \
        docker.io/sushibox/docker-autoheal:latest

⁠Options

Docker Autoheal provides several options for customization:

a) Apply the label autoheal=true to the containers you want to monitor and restart.

b) Set the environment variable CONTAINER_LABEL=all to monitor all running containers.

c) Set the environment variable CONTAINER_LABEL to the name of an existing label that has the value true.

Note: Before using Docker Autoheal, you must apply a HEALTHCHECK to your Docker images. Refer to the Docker documentation⁠ for more details. Additionally, you can configure TCP with mTLS by following the instructions in the Docker documentation⁠.

⁠Environment Variable Defaults

Docker Autoheal includes the following environment variables with their default values:

CONTAINER_LABEL=autoheal
INTERVAL=5   # Check every 5 seconds
START_PERIOD=0   # Wait 0 seconds before the first health check
DEFAULT_STOP_TIMEOUT=10   # Docker waits a maximum of 10 seconds for a container to stop before killing it during restarts (container-specific timeout can be set via label, as explained below)
DOCKER_SOCKET=/var/run/docker.sock   # Unix socket for curl requests to Docker API
CURL_TIMEOUT=30     # Maximum timeout in seconds for curl requests to Docker API
LOG_FILE=/var/log/autoheal.log   # Path to the log file
[email protected]   # Email address for notifications
DISCORD_WEBHOOK_URL=   # Webhook URL for sending Discord notifications (leave empty if not using Discord notifications)

Feel free to modify these environment variables according to your needs.