Skip to content


Switch branches/tags

Name already in use

A tag already exists with the provided branch name. Many Git commands accept both tag and branch names, so creating this branch may cause unexpected behavior. Are you sure you want to create this branch?

Latest commit


Git stats


Failed to load latest commit information.
Latest commit message
Commit time



This repo is no longer being maintained. Users are welcome to fork it, but we make no warranty of its functionality.


Generates a nice graph showing the hierarchy of Docker images in your local image cache.

Looks at all of the Docker image layers cached on the local system and generates a PNG image showing the relationship of the various layers.


The Ruby image-graph script is itself packaged as a Docker image so it can easily be executed with the Docker run command.

Since the script interacts with the Docker API in order to inspect your local image cache it needs access to the Docker API socket. When starting the container, the -v flag needs to be used to make the Docker socket available inside the container.

By default, when the container is started it will generate a PNG that is streamed to stdout:

docker run --rm -v /var/run/docker.sock:/var/run/docker.sock \
  centurylink/image-graph > docker_images.png

You'll need to redirect the container's output to a file in order to save/view the generated image. The name of the output file does not matter, but it is recommended that you use a .png extension so that your image viewer will properly recognize the format of the file.

If you supply an environment variabled named PORT using the -e flag the container will spin-up a web server on the designated port that hosts a web-based version of the image graph:

docker run -d -v /var/run/docker.sock:/var/run/docker.sock \
  -e PORT=3000 -p 3000:3000 centurylink/image-graph 

In the example above, the web server will be started on port 3000. When using this option you'll also want to use the -p flag to map the container port to a port on the host so that you can access the server from outside the container. The container port you specify with the -p flag should match the value you specified for the PORT environment variable.

When a value for the PORT envrionment variable is provided, the PNG version of the image is not streamed to stdout.


Here's an example graph generated by this utility:

Sample Image

And the web-based graph:

Sample Image


How is this different than running docker images --viz | dot -Tpng?

Actually, it's not different at all. The Ruby script executed by the image essentially generates the same output that the --viz flag does and then pipes it to the Graphviz dot tool to generate the graph.

So, why do we need to execute a container when the same thing can be achieved with a one-line Docker command?

While this will work for some versions of Docker, the --viz flag has been marked as deprecated and will likely be removed in some future release.

Additionally, my primary Docker environment is CoreOS which does not have Graphviz installed and provides no way for me to install it myself -- in an environment like this, the only option is to run Graphviz inside a container.


No description, website, or topics provided.






No releases published


No packages published