docker run -d --name unifi-controller -p 8081:8081 -p 8080:8080 -p 8443:8443 -p 8880:8880 -p 8843:8843 -v /opt/docker-data/unifi/unifi-var:/opt/unifi-var -v /dev/urandom:/dev/random iamjond/unifi-controller
The image is vesioned, along with tag aliases for latest, as well as the Unifi Controller version:
| Image Version | Tag | Description |
|---|---|---|
| 1.0 | latest |
Gets the latest image |
| 1.0 | unifi-5.6.22 |
Get the image by Unifi Controller version |
Best practise would normally dictate that a container should contain only a single process. The Unifi Controller start script does start a MongoDB process, and I've left this as-is to simplify the deployment, and maintain all the data and configuration in a central location.
My prefererred mode of operation is to manage all my containers behind a single reverse-proxy. The reverse proxy handles virtual host mappings as well as SSL termination.
In my own setup I am using nginx with LetsEncrypt certificates. This works well with a roll-your-own setup, as well as Synology NAS.