Work with Docker* Container

When working with the application inside a Docker* container, you might need to:

Refer to the sections below to see instructions for each scenario.

Note

Some commands differ among the detached and interactive modes of a container. To learn about the modes, see Advanced Configurations.

Pause and Resume Docker Container

To pause a container with the DL Workbench while keeping all your data in it, stop the container and then restart it later to the previous state with the commands below:

  1. Stop the container:

        * When the container was run in *interactive* mode:
        **Ctrl + C**
    
    
    2. Restart the container:
        * Using `docker start` in *detached* mode:
        ```docker start <container-name>
    • Using docker start in interactive mode:

      docker start -ai <container-name>

Run New Docker Container with DL Workbench

To start a new Docker container with the DL Workbench, stop and remove the current container, then run a new one.

1. Pause Container

docker stop <container-name>
  • In the interactive mode:

    Ctrl + C

2. Remove Container

docker rm <container-name>

Note

The rm command clears all loaded models, datasets, experiments, and profiling data.

3. Run New Container

To run a new container on your OS, use the command from installation form. To run the DL Workbench with different parameters, see Advanced Configurations.

Update the DL Workbench

To get the highest version of the application in your Docker container, pause the container, pull the latest version, and run the container.

Note

Updating a container means building a new container on a newer image and not having your previous data in it. To keep your data in a new container, consider upgrading the DL Workbench.

1. Pause Container

docker stop <container-name>
  • When the container was run in interactive mode:

    Ctrl + C

2. Pull the Highest Version of the DL Workbench

docker pull openvino/workbench:2021.4.2

3. Start New Container

openvino-workbench --image openvino/workbench:2021.4.2 --assets-directory ~/.workbench

For full instructions on running a container and description of the arguments in the command above, see the Advanced Configurations page.

Once the command executes, open the link https://127.0.0.1:5665 in your browser, and the DL Workbench Start Page appears:

_images/start_page_crop.png

Upgrade the DL Workbench Inside a Docker Container

To get the highest version of the application in your Docker container, pause the container, pull the 2021.4.2 version, and run the container with the same folder or volume that you mounted to your previous Docker container.

1. Stop Container

docker stop <container-name>
  • When the container was run in interactive mode:

    Ctrl + C

2. Pull the Highest Version of the DL Workbench

docker pull openvino/workbench:2021.4.2

3. Start New Container

Mount the same folder or volume that you mounted to your previous Docker container and run the new container. You can specify the name of the new container using the --container-name argument, for example, workbench_2021.4.

openvino-workbench --image openvino/workbench:2021.4.2 --assets-directory ~/.workbench

For full instructions on running a container and description of the arguments in the command above, see the Advanced Configurations page.

Once the command executes, open the link https://127.0.0.1:5665 in your browser, and the DL Workbench Start Page appears:

_images/start_page_crop.png

Enter Docker Container with DL Workbench

Note

For this step, the container must be running.

1. Enter Container

To enter the container, run the command:

docker exec -it workbench /bin/bash

This command creates a new instance of a shell in the running Docker container and gives you access to a bash console as an OpenVINO user.

2. See Available Files

The container directory displayed in the terminal is /opt/intel/openvino_2021/deployment_tools/tools/workbench/.

To see a list of files available inside the container, run ls.

Note

The /opt/intel/openvino/deployment_tools/tools/workbench/ directory inside the container includes a service folder wb/data. Make sure you do not apply any changes to it.

3. Inspect Entry Point

cat docker-entrypoint.sh

4. Exit Container

To exit the container, run exit inside the container.