Building an Image Containing an Application to use with Enterprise Developer Build Tools for Windows

This topic outlines the steps that a Dockerfile must carry out to create an image that contains a COBOL application that can be run under Enterprise Developer Build Tools for Windows.

The process defined in this topic is used by the supplied container demonstrations that create images for applications that you can use with Enterprise Developer Build Tools for Windows. When you need to create an image to contain an application of yours, Micro Focus recommends that you base your Dockerfile on a container demonstration's Dockerfile (rather that write your Dockerfile from scratch) because that Dockerfile already performs the steps described in this topic. For more information on one of those container demonstrations, including information on all the files it contains, see The CICS Container Demonstration.

Note: A separate document, Best Practices for Moving Your COBOL Applications to Containers, is available that describes best practices that Micro Focus recommends you adopt when moving an existing COBOL application to run in a containerized environment. See Micro Focus: Best Practices for Moving Your COBOL Applications to Containers for more information.

Before you follow the steps in this topic you must have built a base image for Enterprise Developer Build Tools for Windows. For information on creating a base image for Enterprise Developer Build Tools for Windows see Building a Base Image Containing Enterprise Developer Build Tools for Windows.

Note: Remember that Enterprise Developer is a development and test environment so not for use in production situations. If you want to run COBOL applications in containers in a production environment you must use containers that contain Enterprise Server.

Before building an image containing an application to use with Enterprise Developer Build Tools for Windows you need to ensure that you have available the following:

To build an image that includes an application to use with Enterprise Developer Build Tools for Windows your Dockerfile needs to perform the following steps:

For users of Enterprise Developer for Eclipse:

  1. Specify a base image to work from. This should be an image containing only Enterprise Developer Build Tools for Windows which was built for 32-bit or 64-bit as required and includes the additional build functionality, typically microfocus/edbuildtools-build:win_7.0_x86 or microfocus/edbuildtools-build:win_7.0_x64.

    See Building a Base Image Containing Enterprise Developer Build Tools for Windows for more information.

  2. Define metadata for your image. This will make it easier to establish significant details of the image when you use the docker inspect command.
  3. Define any variables for filenames and folder locations.
    Note: If you will be using this image to run applications under Enterprise Server, you need to set the MFDS_EXTERNAL_ADDR environment variable to specify a resolvable external address string. This is to enable client browsers to resolve the URLs used by ESMAC and other utilities in Enterprise Server Administration.

    The value that you specify for this environment variable is used to replace the internal container address in the URL.

  4. Copy the source files for your application to a folder on the image's filesystem and set the working directory to that folder.
  5. Perform whatever processing you need to do on your application's source file. This would typically be compiling and building but could be anything that you can do with Enterprise Developer Build Tools for Windows.
  6. Copy your built application files to the required folder on the image's filesystem and set the working directory to that folder.
  7. Wait until Micro Focus Directory Server (MFDS) is running, then configure and start any required enterprise server regions. You should consider putting the commands to perform these operations in a separate script. The CICS container demonstration uses a script called DeployAndWait.ps1.
  8. Perform any required clean-up. This includes tasks such as resetting variables and deleting temporary folders.
  9. Specify the name of the executable to run when the image is run.

For users of Enterprise Developer UNIX Components:

  1. Specify a base image to work from. This should be an image containing only Enterprise Developer Build Tools for Windows. See Building a Base Image Containing Enterprise Developer Build Tools for Windows for more information.
  2. Create a new user under which the application will run.
  3. Create a folder to hold the application files then copy the application files into it. If you are using SUSE Linux you will need to change ownership to specify the new user as the owner of this new folder.
  4. Specify the user name to use when the image is run.
  5. Specify the name of the executable to run when the image is run.
  6. Start Micro Focus Directory Server (MFDS), configure and start any required enterprise server regions, and wait until MFDS is running. Check the Dockerfile for the CICS container demonstration for details of the commands to use for this step and the order in which they need to be run.
Note: You might also want to create an image that you can log in to and execute shell or Enterprise Developer commands. This option is useful if you are not adding any application files to a base image but want to be able to use Enterprise Developer commands from it.

All of the container demonstrations offer the option to create such a login image, and those images are tagged with the suffix "_login". See Running the Container Demonstration for the Enterprise Developer Base Image for information on how to specify that you want to create a _login image. For details on the commands required to build such an image, see the bld.sh script in any of the container demonstrations.

Note: When starting a container that includes an application to be run under Enterprise Developer Build Tools for Windows, to be able to manage the container's enterprise servers using Micro Focus Directory Server (MFDS) and Enterprise Server Monitor and Control (ESMAC) from a machine other than the host machine you must specify some additional parameters on the docker run or podman run command. The additional parameters are:
  • -p to map the container's port to a port on the host computer
  • -e to specify the MFDS_EXTERNAL_ADDR environment variable, which is used to specify the name of the host computer where the container is running.

For example, when running an image for the CICS container demonstration you should ensure the docker run or podman run command contains parameters such as the following:

-p 16001:86/udp -p16001:86/tcp -p 9000-9010:9000-9010 -e MFDS_EXTERNAL_ADDR=host-computer-name