Server monitoring · Docker

Alert when a Docker container stops

Containers fail quietly: one exits, the site keeps half-working, and nobody notices until a customer does. Tick the containers that matter and you are told when one goes down.

How it works

  • Docker is detected automatically when the agent is installed. Nothing to configure.
  • The server's page lists every container with its state, restart count, CPU and memory.
  • Tick "Alert me" on the containers you care about. You get one message when one stops, keeps restarting or disappears, and one when it is running again.
  • Ten containers failing together arrive as one message.

Details that avoid false alarms

  • A container that was already stopped when you ticked it does not alert. It has to be seen running first.
  • Containers recreated on every deploy (Docker Swarm, Dokploy) are recognised by their service name, so a redeploy is not a disappearance.
  • Unticking a container ends its alert silently.

Limits

  • Alerts for 5 containers per server on Free, 15 on Pro and Business.
  • CPU and memory per container are read for up to 40 running containers.
  • It reads from Docker only. It cannot start, stop or restart a container.

Leaving Docker out

Reading the container list needs membership of the docker group. If you prefer not to grant that, install the agent with --no-docker: everything else works without it.

Questions

Frequently asked questions

The honest answers, before you sign up.

1

Does it work with Docker Compose and Swarm?

Yes. It reads the container list from the Docker socket, whatever started the containers.

2

Can it restart a failed container?

No. The agent only reads. Use a restart policy in Docker for that, and let the alert tell you when a container is stuck in a restart loop.

3

Does it monitor Kubernetes?

Not as such. It sees the containers on the node it is installed on, but it knows nothing about pods or deployments.