Container Management lists every container on every server. Search by name, filter by status, image or server, group the list by server, application, owner or environment, and select several containers to start, stop, restart or remove them together.
Create a container
Click Create container, choose a server, and fill in the image and name. Optional settings include the command, environment variables, port mappings, named volumes, a restart policy, and resource limits for CPU, memory and number of processes. If a host port is already in use, the form tells you before you create the container.
Container details
Open a container to see its tabs:
- Overview — image, ports, networks, volumes, restart policy, resource limits, health-check results and configuration.
- Resources — live CPU, memory, network, disk I/O and process count, with an Insights panel that spots spikes and possible memory leaks and recommends better limits after about 30 minutes of history.
- Logs — live and historical logs with search and AI summaries.
- Terminal — a shell inside the running container.
- Events — recent events such as starts, stops and health changes.
Actions
- Start, Stop, Restart, Pause and Resume.
- Rename, or Clone its configuration into a new container that is not started yet. Data in volumes is not copied.
- Change the restart policy (On failure, Always, Unless stopped) or Resize its CPU and memory limits without recreating it.
- Recreate — stop and replace the container with an edited configuration.
- Rollback — recreate it on its previous image, keeping every other setting.
- Remove — permanently delete it. Force-kill stops an unresponsive container immediately without a graceful shutdown.
If a container belongs to a deployed stack, actions on its page affect only that container and do not change the stack's configuration. To change the app for good, edit and redeploy its deployment.
Schedules
Use Schedules on a container to start or stop it automatically: once at a chosen date and time, or on chosen days at a set time each week. Recurring schedules are stored in UTC, and the list shows when each one last ran and runs next.
Example: keep test data when recreating a container
For a test database, map a named volume to the data directory required by its image. When changing ports or environment variables, review the volume mapping in Recreate. The replacement container can use the same volume; cloning container configuration alone does not copy database data.