Some images do not include a shell in order to save space, however you might need one to debug a misbihaving container. This article explains how to add a shell into an existing image.
- Access to docker
Adding a shell
You will need a script called
add_shell.sh, which you can download from our GitHub repository.
Now you need to start the script and provide it with the name of the image you are using. Optionally you can also specify a UID of the image user if it is not root (this is required for correct permissions of the shell). Start the script like this:
./add_shell.sh $IMAG_NAME (optional)$UID
The script will now create an image with the same name as the provided one, but with tag
:with-shell. You can now run this image and
exec into it using the added shell at
However you will only be able to use this image locally on your computer. If you wish to use this image in Kubernetes you will have to publish it into a docker repository that you have write access to.