Task — Source Component — Entities and Referential Data

The Business Entities & referential data option enables the extraction of Business Entities from a source environment. This is the default option. It also enables the user to add selected tables to the Business Entities.

The following information needs to be set for the Entities & referential data option:

source-entities

Business Entity

This field defines the task’s Business Entity (BE). Select a BE from the drop-down list, which includes all TDM BEs.

The Advanced option is optional and allows you to:

  • Select only specific systems and/or Logical Units (LUs) for the task.
  • Define the affinity and maximum number of workers per LU.
  • Override the task’s default execution mode

When you click Advanced, a pop-up window opens:

advanced be1

Advanced BE - Systems & Logical units Tab

This tab is available for both Source and Target components and displays the systems and LUs associated with the selected BE.

Selected Systems and/or LUs

  • If the selected source environment does not contain all systems or LUs defined in the BE, the Task window automatically removes the missing LUs and displays a warning.
  • You can manually remove an entire system or specific LUs from the task.

If the selected environment includes disabled systems and the Policy for Fetching Data extracts data from the source environment, those disabled systems are automatically excluded from the task.

System & Logical Units Tab - Affinity and Max Number of Workers

Starting with TDM 9.5, each LU in the Advanced BE view supports configuration of:

  • Affinity
  • Maximum number of workers

You can expand each LU and configure these values after selecting an environment.

By default, each LU inherits its values from:

  1. The environment’s system settings (if defined), or
  2. The TDM and Fabric configuration values

You can override these values per LU to control and optimize task execution behavior:

affinity

When both Source and Target components are defined for the task (Load, Extract & Load, Generate & Load):

  • Affinity
    • The task execution process runs using the target affinity
    • The entity extraction runs using the source affinity
  • Max number of workers
    • The task execution process uses the target maximum number of workers

Click here for detailed information on how task execution determines the effective affinity and maximum number of workers.

Advanced BE - Execution Mode Tab

This tab enables changing the task execution mode:

advanced be1

By default, the task execution mode is taken from the task's Business Entity (BE). However, you can set the task's execution mode to be independent to the BE's execution mode. The following options are available:

  • Vertical execution - execution of the entire LU hierarchy for each root entity before moving on to the next root entity. Note that this mode is not available for a task that generates entity clones or synthetic entities generation.
  • Horizontal execution - execution of the task system by system (LU by LU) where all entities are processed in one LU before moving on to the next system in the hierarchy.

The Vertical execution mode can be beneficial when running TDM tasks on a large scale of entities as it ensures better cross-systems data consistency and data alignment.

Source Environment

Select one TDM environment from the drop-down list. The drop-down list displays the list of available source environments for the user. Only environments that contain systems with the select task's BE are displayed. If the source environment is defined as containing sensitive data in the Environment window, the TDM task window displays a message stating that the sensitive data is masked.

Policy for Fetching Data

This setting defines whether the data needs to be extracted from the source environment or whether it can be retrieved from the Test Data Store (Fabric). The following options are available:

  1. Available data from the Test data store, new data from [source environment name] - this is the default option. When selected, new entities will be synced from the source environment. Entities that are already stored in the Test Data Store will be taken from the Test Data Store, unless the sync policy - defined in the LU implementation - states that they need to be synced from the source environment.

  2. All data from [source environment name] - always sync the data from the source environment.

  3. Available [source environment name] data in the Test data store - get the data from the Test Data Store, if exists (extracted from the source environment by previous task executions).

  4. Selected snapshot (version) - get a selected data snapshot (version) created in the Test Data Store by previous task executions.

    Notes:

    • Options 1 and 2 are not available if the source environment is set with the Do not Sync option.
    • Options 1 and 2 are not available if the tester user does not have a Read permission set on the source environment.
    • Option 2 (always sync) is available only for permitted users.
    • The task execution does not access the source environment directly when options 3 or 4 are selected.

​ Click here for more information about the task's data sync modes.

Referential Tables

Check this checkbox to add related tables to the task's entities:

task related tables

​ Note that the list of available tables must be defined in the TDM implementation.

Previous

Task — Source Component — Entities and Referential Data

The Business Entities & referential data option enables the extraction of Business Entities from a source environment. This is the default option. It also enables the user to add selected tables to the Business Entities.

The following information needs to be set for the Entities & referential data option:

source-entities

Business Entity

This field defines the task’s Business Entity (BE). Select a BE from the drop-down list, which includes all TDM BEs.

The Advanced option is optional and allows you to:

  • Select only specific systems and/or Logical Units (LUs) for the task.
  • Define the affinity and maximum number of workers per LU.
  • Override the task’s default execution mode

When you click Advanced, a pop-up window opens:

advanced be1

Advanced BE - Systems & Logical units Tab

This tab is available for both Source and Target components and displays the systems and LUs associated with the selected BE.

Selected Systems and/or LUs

  • If the selected source environment does not contain all systems or LUs defined in the BE, the Task window automatically removes the missing LUs and displays a warning.
  • You can manually remove an entire system or specific LUs from the task.

If the selected environment includes disabled systems and the Policy for Fetching Data extracts data from the source environment, those disabled systems are automatically excluded from the task.

System & Logical Units Tab - Affinity and Max Number of Workers

Starting with TDM 9.5, each LU in the Advanced BE view supports configuration of:

  • Affinity
  • Maximum number of workers

You can expand each LU and configure these values after selecting an environment.

By default, each LU inherits its values from:

  1. The environment’s system settings (if defined), or
  2. The TDM and Fabric configuration values

You can override these values per LU to control and optimize task execution behavior:

affinity

When both Source and Target components are defined for the task (Load, Extract & Load, Generate & Load):

  • Affinity
    • The task execution process runs using the target affinity
    • The entity extraction runs using the source affinity
  • Max number of workers
    • The task execution process uses the target maximum number of workers

Click here for detailed information on how task execution determines the effective affinity and maximum number of workers.

Advanced BE - Execution Mode Tab

This tab enables changing the task execution mode:

advanced be1

By default, the task execution mode is taken from the task's Business Entity (BE). However, you can set the task's execution mode to be independent to the BE's execution mode. The following options are available:

  • Vertical execution - execution of the entire LU hierarchy for each root entity before moving on to the next root entity. Note that this mode is not available for a task that generates entity clones or synthetic entities generation.
  • Horizontal execution - execution of the task system by system (LU by LU) where all entities are processed in one LU before moving on to the next system in the hierarchy.

The Vertical execution mode can be beneficial when running TDM tasks on a large scale of entities as it ensures better cross-systems data consistency and data alignment.

Source Environment

Select one TDM environment from the drop-down list. The drop-down list displays the list of available source environments for the user. Only environments that contain systems with the select task's BE are displayed. If the source environment is defined as containing sensitive data in the Environment window, the TDM task window displays a message stating that the sensitive data is masked.

Policy for Fetching Data

This setting defines whether the data needs to be extracted from the source environment or whether it can be retrieved from the Test Data Store (Fabric). The following options are available:

  1. Available data from the Test data store, new data from [source environment name] - this is the default option. When selected, new entities will be synced from the source environment. Entities that are already stored in the Test Data Store will be taken from the Test Data Store, unless the sync policy - defined in the LU implementation - states that they need to be synced from the source environment.

  2. All data from [source environment name] - always sync the data from the source environment.

  3. Available [source environment name] data in the Test data store - get the data from the Test Data Store, if exists (extracted from the source environment by previous task executions).

  4. Selected snapshot (version) - get a selected data snapshot (version) created in the Test Data Store by previous task executions.

    Notes:

    • Options 1 and 2 are not available if the source environment is set with the Do not Sync option.
    • Options 1 and 2 are not available if the tester user does not have a Read permission set on the source environment.
    • Option 2 (always sync) is available only for permitted users.
    • The task execution does not access the source environment directly when options 3 or 4 are selected.

​ Click here for more information about the task's data sync modes.

Referential Tables

Check this checkbox to add related tables to the task's entities:

task related tables

​ Note that the list of available tables must be defined in the TDM implementation.

Previous