Skip to main content

Persistent admin console (Beta)

This topic describes the persistent admin console for Embedded Cluster 3.x — an optional, always-on web UI for browser-driven upgrades and operations.

Overview​

The persistent admin console is an opt-in web service that runs on a controller node and exposes a dashboard at https://<controller-hostname>:30080/console. From the dashboard, end users can view the currently installed application version, open configured application links, see any newer versions available on the channel, view release notes, and trigger an upgrade with a single click — without SSH access to the controller.

The persistent admin console is not enabled by default.

The initial application install and console setup require command-line access to a controller node. Once the console is running, end users can perform upgrades and configuration changes from the browser without command-line access.

note

For automated deployments (such as cloud marketplace offerings using cloud-init or user-data scripts), include console install in your provisioning script after the initial install completes. This gives end users browser-based access from first boot without any manual setup.

Install the console​

Once the application is installed and running, install the persistent admin console on a controller node. The console serves its web UI from that controller's hostname.

  1. On a controller node, run the install command:

    sudo ./<app-slug> console install

    Replace <app-slug> with your application slug.

  2. (Optional) Use the --port flag to override the default port (30080):

    sudo ./<app-slug> console install --port 30443

  3. Verify the service is running:

    sudo systemctl is-active <app-slug>-console-web.service

    The output should be active.

  4. Open the console in your browser at https://<controller-hostname>:30080/console.

For the full set of flags, see console.

Choose a controller​

The console can be installed on any controller node in the cluster. Replicated recommends running it on the primary controller — the first controller node installed in the cluster — because that node is typically the one users already access for cluster operations.

If your release includes an app.k8s.io/v1beta1 Application resource with spec.descriptor.links, those links appear automatically on the console dashboard under Open <app-name>. No additional configuration is needed.

note

If you are migrating from Embedded Cluster v2, any existing Application links carry over automatically.

Recover or move the console​

The console is hosted by a single controller. If the controller hosting the console is lost or the <app-slug>-console-web.service becomes unhealthy, rerun the install command on any controller node to bring it back up:

sudo ./<app-slug> console install

If the original controller is still healthy and you are recovering an unhealthy service, run the command on the original controller. If the original controller is lost, run the command on any other controller node. The console is stateless — the dashboard reads live data from the daemon, so moving it to a different controller is safe at any time.

Uninstall the console​

To uninstall the console from a controller:

sudo ./<app-slug> console uninstall

This stops and removes the <app-slug>-console-web.service systemd unit and frees the port. The cluster is not otherwise affected — the application continues running, and users can still upgrade with the upgrade command from the CLI. See Perform updates in embedded clusters.

Limitations​

The persistent admin console has the following limitations:

  • Air gap installations are not supported.

  • The console can only be installed on a single controller node at a time. Reinstall on a different controller to move it.

  • The console must be installed on a controller node; it cannot run on a worker node.

  • The cluster must have TLS configured at install time. The console reuses the cluster's TLS certificate to serve HTTPS on its port.

On this page