Docker has a built-in health check mechanism that, if defined for an image, probes a defined command regularly to determine the health status of the container. Because containerization is all about automation of processes, health checks are not only important for monitoring purposes but even more important for triggering events in an automation chain in your cluster. Many tools depend on the status information the container daemon provides. The Traefik reverse proxy uses in our Docker development setup for example automatically removes a routing to a backend, when the backends health status is not healthy.

If you for example run

docker ps --format '{{.Names}} - {{.Status}}'

you will get a status report for all running containers like this:

elastic-worker - Up 15 hours (healthy)
user-changes - Up 15 hours (healthy)
cae-preview - Up 15 hours (healthy)
cae-live - Up 15 hours (healthy)
site-manager - Up 15 hours
studio-server - Up 15 hours (healthy)
overview - Up 15 hours
content-feeder - Up 15 hours (healthy)
cae-feeder-live - Up 15 hours (healthy)
commerce-adapter-wcs - Up 15 hours (healthy)
headless-server-preview - Up 15 hours (healthy)
master-live-server - Up 15 hours (healthy)
cae-feeder-preview - Up 15 hours (healthy)
headless-server-live - Up 15 hours (healthy)
studio-client - Up 15 hours (healthy)
workflow-server - Up 15 hours (healthy)
content-management-server - Up 15 hours (healthy)

The health check command is defined in the Dockerfile of the application, that is

HEALTHCHECK --start-period=30s --interval=30s --timeout=3s \
 CMD curl -Lf http://localhost:8081/actuator/health || exit 1