Using the zmmt command to create migration definitions on z/OS

Before you migrate a WebSphere® Application Server Version 7.0 or later node to Version 9.0, you must create Job Control Language (JCL) jobs (CNTL and DATA datasets) that you will run during the actual migration. You can use the zmmt command to create the appropriate jobs and instructions. Learn how you can use the zmmt command in the migration process.

Before you begin

Supported configurations:

This topic is about profile configuration migration. To migrate your applications to the latest version, use the WebSphere Application Server Migration Toolkit.

Install WebSphere Application Server Version 9.0 on a z/OS® system containing a Version 7.0 or later configuration.

Note: The zmmt command is intended for use by a systems programmer or administrator who is familiar with the z/OS target system on which the migrated Version 9.0 nodes run.

About this task

  • You can use the zmmt command to create migration jobs for migrating WebSphere Application Server Version 7.0 or later nodes to Version 9.0. The zmmt command itself does not perform the migration; it creates batch jobs and instructions that you can then use to perform the actual migration tasks.

    The zmmt command is an alternative to the z/OS Migration Management Tool. You might want to use this command rather than the tool to create migration jobs if you have no workstations with the WebSphere Customization Toolbox available or need to automate the generation of your migration jobs.

    You run the zmmt command using a migration response file that you create manually.

    Tip: Migrating a WebSphere Application Server application-serving environment requires a fair amount of planning and coordination. If you have not previously migrated an application server, you should first migrate a "practice" standalone application server using a sample response file.
    Note: In WebSphere Application Server Version 9.0, you can use the z/OS Migration Management Tool or the zmmt command and the jobs that they generate to migrate from Version 7.0 or later to Version 9.0 nodes. After you actually migrate a node, however, you use the WebSphere Application Server Version 9.0 administrative console or scripting to administer it.
  • The zmmt.sh shell script is located in the install_root/bin directory.

    The shell script uses the following parameters:

    -responseFile
    This parameter is used to specify the path to your response file.

    This cannot be used with the -profilePath parameter.

    -profilePath
    This parameter is used to specify a fully qualified path name to an existing set of generated jobs.

    This cannot be used with the -responseFile parameter.

    -workspace
    This required parameter is used to specify the Eclipse workspace directory.

    This must match the path specified in the response file before the .metadata directory for the profilePath and templatePath variables.

    -transfer
    This optional parameter is used to copy generated jobs from a UNIX System Services (USS) file system to a pair of partitioned datasets.

    The zmmt command first writes the migration jobs to a USS file system.

    -allocate
    This optional parameter is used to attempt to allocate the target datasets.

    This parameter cannot be used without the -transfer parameter.

    Datasets are determined by appending the values ".CNTL" and ".DATA" to the zTargetHLQ value for the profile containing the jobs that are being copied. This operation overwrites existing files of the same name in those datasets.

Procedure

  1. Prepare the type of response file that you need.
  2. On your target z/OS system, run the zmmt.sh shell script using your prepared response file.

    The command creates the CNTL and DATA files needed to run the required jobs. The response file needs to be located in the UNIX (USS) file system.

Results

After you run the zmmt.sh script, you should see something similar to the following output:
install_root/bin:> zmmt.sh  /zmmt/workspace
-responseFile  /ZMigDmgr01.responseFile  -allocate  -transfer<

osgi> Customization definition successfully written to 
  /zmt/workspace/.metadata/.plugins/com.ibm.ws390.mmt.config/profiles/ZMigDmgr01

Attempting to allocate dataset: ZMMT.CNTL
Allocation successful.
Attempting to allocate dataset: ZMMT.DATA
Allocation successful.
Copying CNTL files to ZMMT.CNTL...
Copy successful.
Copying DATA files to ZMMT.DATA...
Copy successful.

The migration jobs have been copied to the CNTL and DATA datasets. Follow the generated instructions provided in the CNTL dataset to perform the actual migration.

Example

The following examples describe typical command lines with attributes for the zmmt command. In these examples, /xxx can be any directory to which the user invoking the zmmt.sh script has read and write access.

  • zmmt.sh -workspace /xxx -responseFile /xxx/ZAppSrvcmd.responseFile

    This example generates the migration jobs in the location specified by the profilePath option in the response file.

  • zmmt.sh -workspace /xxx -responseFile /xxx/ZAppSrvcmd.responseFile -transfer -allocate
    This example performs the following actions:
    • Generates the migration jobs in the location specified by the profilePath option in the response file
    • Transfers the jobs from the file system to the CNTL and DATA datasets
    • Allocates the target CNTL and DATA datasets using the high-level qualifier specified by the targetHLQ option in the response file
  • zmmt.sh -workspace /xxx -profilePath /xxx/ZAppSrvcmd -transfer -allocate
    This example performs the following actions:
    • Transfers the generated jobs at the location specified by the -profilePath parameter to those datasets
    • Allocates the target CNTL and DATA datasets using the high-level qualifier specified by the targetHLQ option in the response file
    Note: This usage assumes that migration jobs have already been generated with a previous invocation of the zmmt.sh script.
  • zmmt.sh -workspace /xxx -transfer -responseFile /xxx/ZAppSrvcmd.responseFile

    This example transfers the generated jobs from the location specified by the profilePath option in the response file to the generated CNTL and DATA datasets.

    Note: This usage assumes that the jobs have already been generated with a previous invocation of the zmmt.sh script and that the target CNTL and DATA datasets have already been allocated.