Using CERMM

Containers

On Linux, the agent reads Docker (or Podman's Docker-compatible socket) with each report. The device's page shows every container.

What you see

A device's Containers section, grouped by compose project. Each container has its name, image, state and status, health, compose project and service, published ports, CPU (worked out between two reports) and memory (without page cache).

What you can do

• Click a container: Logs, Restart, Stop, Start.
• A project's ⋯ menu: restart, stop or start the whole project (compose -p).

These are signed jobs through the Run drawer, using docker or podman by the container's short ID, so they need Run commands.

Containers as devices

An agent running inside an LXC, Docker or Podman container is a device of its own, under the Containers pill with its own icon. It still counts as a server for alerts and for the Servers schedule target.

Not yet

• Windows and macOS (Docker Desktop runs in a VM the agent doesn't see into).
• Image updates and container alerts.