Execution Modes

This section describes in greater detail the three Extension Technology Execution modes: default, parallel execution, and unload.

Default Mode

Use the default execution mode when the source data store meets all the schema requirements described in the section Schema Requirements of chapter Requirements of the Readme .

Parallel Mode

Use the parallel mode when you want to execute more than one method simultaneously.

Warning: Use of the parallel mode where multiple methods access the same tables within the target data store(s) is not currently supported.

In order to run a parallel execution, the files created after a successful export from Data Masking or Data Subset Extraction must be stored in a different data directory for each method.

  1. Create a new directory for a method.
  2. Within the newly created directory, create the subdirectory config.
  3. Copy the contents of the default config directory into the newly created config directory.

    For more information about the default config directory, see the section Extension Technology Directories in this guide.

  4. Export or manually move the files generated by the Distributed Exporter into the new config directory.
    Note: Bear in mind that the machine where the ODBC Extension is executed may not be the same the same machine where the Windows client software is running.
  5. Use the dxestart command with the -d option to specify the new location.

    See Invoking Extension Technology for more information.

Note: You can execute the first method in default mode, which utilizes your default data directory, but you must execute all subsequent methods in parallel execution mode.

For more information about default directories, see the section Data in the chapter Extension Technology Directories.

Unload Mode

In unload mode, the same schema requirements apply (as described in the section Schema Requirements of chapter Requirements of Readme ) , but with a difference: Data Express does not use privileges against the source data store. Instead, information is unloaded from the source data store to separate staging data store where these privileges are can be authorized.

You can use unload mode for the following two scenarios:

Source Data Store Staging Data Store Target Data Store

Data Store A

Data Store B

Data Store C

Data Store A

Data Store C

Data Store C

These examples show that the staging data store may or may not be the actual target data store.

For example, to use unload mode with Data Masking:

  1. In the Distributed Loader within Data Builder, specify your source data store as the Source database (as normal), and upload tables.

    For more information about the Distributed Loader, see the section Using Distributed Loader of Part 2. Using Distributed Data Stores with Data Express.

  2. In Data Masking, enable masking of all the tables catalogued.

    For more information, see Enable Masking within the section Data Masking of Part 4: Data Express for Distributed Data Stores Tutorials.

  3. In the Distributed Exporter within Data Masking, specify your staging data store as your Target database, and then export files.
  4. Copy the exported files into the config directory.
    Note: For more information about default directories, see the section Data in the chapter Extension Technology Directories.
  5. Execute the Extension Technology using unload mode.

    For more information, see the section Invoking Extension Technology.

    The data from the source data store is extracted and inserted into the staging data store after the required tables were created.

  6. From this point forward, you can process your data normally provided that you specify the staging data store as your source and the target data store as your target. This means that you load from staging and export into target.
Note: A similar approach can be used to unload with Data Subset Extraction; after using the Distributed Loader to upload your tables, you can mark all the tables as "register tables". Then, it is very simple to create a subsetting method of your register tables from Data Subset Extraction. Use the Distributed Exporter to copy the exported files into the config directory, and execute the Extension Technology using unload mode.