Containers
Manage Docker containers from Arcane.
Use the Containers page to create, inspect, edit, and remove individual containers on the host. To manage Compose services as a group, use Projects.
Browse containers
Section titled “Browse containers”Open Containers in the sidebar. The table shows name, ID, image, and status. Long port lists collapse behind a +N expander, and the view options menu can hide exposed-only ports to show only published host mappings. Select a row checkbox, then Shift-select another to select the rows between them on the current page.
Filter by Docker label
Section titled “Filter by Docker label”In the Labels filter, enter a key such as com.example.team (matches any value) or a key and value such as com.example.team=media.
To keep a container out of the default list and dashboard counts, add the com.getarcaneapp.arcane.hidden=true label or set x-arcane.hidden in Compose. See Hide containers for examples, project-wide defaults, and unhiding.
Create a container
Section titled “Create a container”- Click Create Container.
- Fill in name and image. Ports, volumes, environment variables, restart policy, and other fields are optional.
- Click Create.
Start, stop, restart, pause, kill
Section titled “Start, stop, restart, pause, kill”- Start / Stop / Restart — change the running state.
- Pause / Unpause — suspend and resume all processes in the container. Requires
containers:pause. - Kill — send a signal to the container’s main process. Requires
containers:kill. - Redeploy — pull the latest image and recreate the container with the same name, mounts, labels, networks, and restart policy, updating it in place.
On Podman, Quadlet containers show a Quadlet badge. Start, stop, and restart go to their systemd unit; Arcane won’t edit, redeploy, or remove them. See Quadlets.
Edit a container
Section titled “Edit a container”Choose Edit from a container’s detail page or row menu to change its image, ports, mounts, variables, restart policy, networks, resource limits, or Linux capabilities. Network options include static IPv4 addresses and aliases. Editing requires containers:edit; network connections also require networks:connect / networks:disconnect.
Applying changes recreates the container after confirmation, giving it a new container ID. Existing mount options are preserved, and the new image is pulled only if missing locally. If recreation fails, Arcane restores the original, except for auto-remove containers.
Convert a container to a Compose project
Section titled “Convert a container to a Compose project”This feature is experimental. Click the version number at the bottom of the sidebar and switch on Experimental Features (requires settings:write). Standalone containers then offer Convert to Compose on their detail page, row menu, or in bulk (requires projects:create).
It opens the new-project page with a generated Compose file. Check bind mounts, networks, and environment values, edit as needed, then click Create Project.
Original containers keep running; stop them before deploying to avoid name and port conflicts. With containers:delete, select Remove original container(s) after creation to delete them at project creation, before deployment. Deletion can’t be undone.
Commit a container to an image
Section titled “Commit a container to an image”On a container detail page, click Commit to create an image from its current filesystem (requires images:commit). You can set repository and tag, comment, author, and whether Docker skips its default pause. The image then appears on Images.
Inspect a container
Section titled “Inspect a container”Click a container’s name or Inspect to open the detail view, with tabs for configuration, network settings, mounts, and logs. Overview volume and network counts open their tabs. Network details show assigned IPv4 and IPv6 addresses.
Compose tab
Section titled “Compose tab”Containers in an Arcane-managed Compose project get a Compose tab with the source file: the root compose file, or the included file when the service comes from a Compose include. It’s read-only for Git-synced projects.
Updates
Section titled “Updates”The Overview tab’s Auto Update toggle turns automatic installation on or off; checks continue either way. For the labels, and stopping checks too, see Per-container labels.
When a newer image exists, the header shows an Update available badge and an Update action (requires containers:autoupdate). Project containers also link to their project.
Open a shell
Section titled “Open a shell”With containers:exec, a running container’s detail view has a Shell tab. Pick sh, bash, ash, zsh, or a custom path. Your browser remembers it per Compose service (or container name), so it survives recreation. Otherwise the environment’s Default Shell (Environments → the environment → Docker) is used, /bin/sh by default.
View logs
Section titled “View logs”Open the container’s Logs tab. The viewer renders JSON and logfmt logs as structured rows, keeps multiline messages such as stack traces together, and shows small CPU and memory monitors.
Download logs
Section titled “Download logs”Click Download to save a plain-text .log file with timestamps and all stdout and stderr Docker still retains, ignoring the viewer’s filters and line limit. Logs removed by Docker’s rotation aren’t included.
Remove a container
Section titled “Remove a container”- Click the trash icon on the container row.
- Confirm.