How to update a Docker container without losing the configuration using Dockge

I'm sure it's happened to you: you have your entire ecosystem of services set up in Docker, maybe with Home Assistant or some home automation tool, and then you get the cervical fear of losing the entire configuration when trying to upgrade to the latest version. Dockge It is the ideal resource for dealing with ephemeral containers, in which any changes made within them vanish as soon as you delete them to put in a new one.

The good news is that, if things are done right, updating is a piece of cake and you don't have to configure everything from scratch each time. The key lies in separating the application (the image) from the data (the configuration), allowing the new container to simply "hook" onto the information you already had stored on your hard drive.

The secret of persistence: Volumes and Bind Mounts

To avoid any surprises, the first thing to understand is that data only survives if it's outside the container's lifecycle. If, when doing a docker inspect If you see that the "Mounts" section of your service is empty, stop immediately; if you delete that container, You will lose absolutely everything..

There are two main ways to save your skin: volumes (managed by Docker) and the bind mounts (specific folders on your PC or NAS). As long as the container is pointing to an external folder, you can delete the image a thousand times and, when you create the new one, it will instantly recognize your settings and databases.

What is Dockge?

Dockge is an open-source tool for manage Docker applications based on Docker Compose It's accessible through a simple web interface. It's designed to simplify container management without requiring manual editing of docker-compose.yml files in the terminal.

These are the actions that Dockge allows us to perform:

• Update images and deploy changes.
• Create and edit Docker Compose stacks from the browser.
• Manage multiple applications from a single web panel.
• Import existing Docker Compose projects.
• Start, stop, and restart containers.
• View real-time logs.

To update a Docker container without losing the configuration using Dockge, simply edit the stack If necessary, press Pull to download the latest imageNext, you need to select Redeploy to recreate the container. The configuration and data are preserved as long as they are stored on volumes or folders mounted outside the container.

Other methods for upgrading: From manual labor to automation

Not convinced by Dockge? If you're someone who prefers to have total control and leave nothing to chance, then Dockge is for you. via manual This is your path. It involves stopping the service, removing the old container, and downloading the latest version with a Docker pull and relaunch it. However, keep in mind that you'll need to replicate the exact same ports and environment variables you had before for everything to work the first time.

For those who have already made the leap to Docker ComposeLife is much simpler. Just execute docker-compose pull to download the new releases and then docker-compose up -dThe system will detect that there is a new image and recreate the container while maintaining the structure of the .yml file, thus ensuring that the environment is fully reproducible and stable.

If managing dozens of containers by hand seems like a nightmare, there are tools that do the dirty work. Watchtower It's the most popular option; it installs like any other container and monitors Docker Hub every 24 hours to update everything automatically. For those who are more cautious, What's up Docker (WUD) It's ideal because it doesn't update on its own, but rather it updates you. sends a notification notifying you that there is a new version, leaving the final decision to you.

Visual management with Portainer and advanced strategies

If you don't get along with the terminal or Dockge doesn't quite convince you, Portainer It offers a very intuitive graphical interface. To update here, simply go to the images section, download the latest version, and then use the option to Duplicate/Edit in the current container to assign the new image and deploy it again.

Sometimes the question arises whether it's better to update the entire image or run a dnf update o apt-get update inside the powered-on container. The short answer is: don't do it in production. Updating internal packages is useful for a quick temporary patchBut it breaks reproducibility. If you make internal changes and the container dies, those changes die with it, unless you use docker commit to create a personalized image.

For a professional strategy, the ideal is the mixed modelLet Watchtower manage non-critical services that won't break anything, use WUD to monitor important ones, and perform manual processes for services where a version error could leave you offline or delete sensitive data. A good practice is set version labels specific ones (like v3.3.0) instead of using the :latest tag in the most sensitive services.

Having a clear persistence strategy using volumes and choosing the right update method, whether through the efficiency of Dockge, the simplicity of Docker Compose, the visual management of Portainer, or the automation of Watchtower, allows you to keep services up to date without risk of data loss and ensuring that each deployment is stable and reversible.