Configuring Pipelines

Pipelines are used to edit standard chains of data operations.

Note

The configuration of pipelines directly affects the operation of basic system functions. All basic execution threads are included in the official Universe MDM delivery set

Before you start:

• Create an account with access to the "Pipelines", "Operations", "Duplicates", "Workflows", and "Libraries" sections. Access to sections is configured for a role that is subsequently assigned to the account.

Reindexing Relations

1. In the "Pipelines" section of the system administrator interface, create the pipeline of relation reindexing:

   • Select start segment: [RELATION_UPSERT_START].

   • Add two segments of Point type: [RELATION_UPSERT_TIMELINE] and [RELATION_UPSERT_INDEXING].

   • Select finish segment: [RELATION_UPSERT_FINISH].

2. In the "Operations" section, create an operation to reindex data and turn on the "Reindex relations" checkbox.

3. Run the operation and wait for its execution.

See the description of the segments in the main creating/updating relation pipeline example.

Figure 1. Reindexing operation

Data Matching While Saving Record Changes

1. In the "System Parameters" section of the system administrator interface, enable the "Real-time matching" chekbox in the Data matching settings block.

2. In the "Pipelines" section of the system administrator interface, create the pipeline for matching data when saving changes:

   • Select start segment: [RECORD_UPSERT_START].

   • Add a segment of the Connector type: [RECORD_UPSERT_MATCHING_CONNECTOR] .

   • Select finish segment: [RECORD_UPSERT_FINISH].

3. Configure matching mechanisms.

4. View results in the "Duplicates" section.

See the description of the segments in the examples of main data matching pipelines.

Figure 2. System Parameters

Figure 3. Duplicates

Publishing Record Draft According to Workflows

1. In the "Libraries" section, download .jar file with the service task and process listener implementations.

2. In the "Workflows" section, add an event executor using previously downloaded library.

3. Go to the "Workflows" tab and create a new process. In the "Diagram" tab, you will be able to create events that should be processed by service tasks (Service Task type).

4. In the "Pipelines" section of the system administrator interface, create the pipeline of the record draft publishing:

   • Select start segment: [RECORD_DRAFT_PUBLISH_START].

   • Add a segment of the Selector type: [RECORD_DRAFT_PUBLISH_WORKFLOW_SELECTOR].

   • Select finish segment: [RECORD_DRAFT_PUBLISH_FINISH].

For a description of the segments, see the record draft publishing pipeline.

Figure 4. Adding service task to process diagram