The Dockerfile in this subfolder builds on the MATLAB Container Image on Docker Hub by installing MATLAB® toolboxes and support packages using MATLAB Package Manager (mpm).
Use the Dockerfile as an example of how to build a custom image that contains the features of the MATLAB image on Docker® Hub. These features include accessing the dockerised MATLAB through a browser, batch mode, or an interactive command prompt. For details of the features in that image, see MATLAB Container Image on Docker Hub.
- Docker
Access this Dockerfile either by directly downloading this repository from GitHub®, or by cloning this repository and then navigating to the appropriate subfolder.
git clone https://github.com/mathworks-ref-arch/matlab-dockerfile.git
cd matlab-dockerfile/alternates/building-on-matlab-docker-imageBuild a container with a name and tag.
docker build -t matlab_with_add_ons:R2024b .You can then run the container with the "batch" option. Test the container by running an example MATLAB command such as ver to display the installed toolboxes.
docker run --init --rm -e MLM_LICENSE_FILE=27000@MyServerName matlab_with_add_ons:R2024b -batch verYou can check the installed support packages using the MATLAB command matlabshared.supportpkg.getInstalled.
You can also run the container with the "browser" option to access MATLAB in a browser.
docker run --init --rm -it -p 8888:8888 matlab_with_add_ons:R2024b -browserFor more information, see Run the Container.
This Dockerfile installs any specified products into the MATLAB installation on the MATLAB Docker Hub image.
To customize the build, either pass a list of products into the ADDITIONAL_PRODUCTS
argument when building the Docker image, or edit the default value of that argument in the Dockerfile.
The ADDITIONAL_PRODUCTS argument must be a space separated list surrounded by quotes.
By default, ADDITIONAL_PRODUCTS includes example products, which you can replace.
For example, to build an image containing MATLAB and the Deep Learning Toolbox™:
docker build --build-arg ADDITIONAL_PRODUCTS="Deep_Learning_Toolbox" -t matlab_with_add_ons:R2024b .For a successful build, include at least one product.
mpm automatically installs any toolboxes and support packages
required by the products specified in ADDITIONAL_PRODUCTS.
For more information, see MATLAB Package Manager (mpm).
You can modify the products argument of mpm, but not the destination folder default value, as this is
set to match that of the MATLAB image on Docker Hub. If this default value for the --destination argument is modified, the build may fail.
The Dockerfile supports the following Docker build-time variables:
| Argument Name | Default value | Effect |
|---|---|---|
| MATLAB_RELEASE | R2024b | The MATLAB release to install, for example, R2023b. |
| ADDITIONAL_PRODUCTS | "Symbolic_Math_Toolbox Deep_Learning_Toolbox_Model_for_ResNet-50_Network" | A space separated list of toolboxes and support packages to install. For more details, see MATLAB Package Manager |
| LICENSE_SERVER | unset | The port and hostname of a machine that is running a Network License Manager, using the port@hostname syntax. For example: 27000@MyServerName. To use this build argument, the corresponding lines must be uncommented in the Dockerfile. |
Use these arguments with the docker build command to customize the image.
Alternatively, the default values for these arguments can be changed
directly in the Dockerfile.
For example, to build an image for MATLAB R2022b, use the following command.
docker build --build-arg MATLAB_RELEASE=R2022b -t matlab_with_add_ons:R2022b .To build an image for MATLAB R2022b with Deep Learning Toolbox and Parallel Computing Toolbox™, use the following command.
docker build --build-arg MATLAB_RELEASE=R2022b --build-arg ADDITIONAL_PRODUCTS="Deep_Learning_Toolbox Parallel_Computing_Toolbox" -t matlab_with_add_ons:R2022b .For supported releases see MATLAB Container Image on Docker Hub.
Including the license server information with the docker build command avoids having to pass it when running the container.
To use this build argument, uncomment the corresponding lines in the Dockerfile.
If those lines are uncommented, $LICENSE_SERVER must be a valid license
server or browser mode will not start successfully.
Build container with the License Server.
docker build -t matlab_with_add_ons:R2024b --build-arg LICENSE_SERVER=27000@MyServerName .Run the container, without needing to pass license information.
docker run --init matlab_with_add_ons:R2024b -batch verThe Docker container you build using this Dockerfile inherits run options from its base image. See the documentation for the base image, MATLAB Container Image on Docker Hub (hosted on Docker Hub) for instructions on how to use those features, which include interacting with MATLAB using a web browser, batch mode, or an interactive command prompt, as well as how to provide license information when running the container. Run the commands described there with the name of the Docker image that you build using this Dockerfile.
For more MATLAB Docker resources, see More MATLAB Docker Resources.
You can help improve MATLAB by providing user experience information on how you use MathWorks products. Your participation ensures that you are represented and helps us design better products. To opt out of this service, delete the following line in the Dockerfile:
ENV MW_DDUX_FORCE_ENABLE=true MW_CONTEXT_TAGS=$MW_CONTEXT_TAGS,MATLAB:TOOLBOXES:DOCKERFILE:V1To learn more, see the documentation: Help Make MATLAB Even Better - Frequently Asked Questions.
We encourage you to try this repository with your environment and provide feedback. If you encounter a technical issue or have an enhancement request, create an issue here.
Copyright 2023-2024 The MathWorks, Inc.