You can purge the data of your organization. When you purge data, you permanently delete data from the data store without deleting data models and other configuration details.
You can choose to purge all business entity records, source records, and relationships, along with historical data. You can also choose to purge all historical data from business entity records, hierarchies, and relationships or purge records that belong to a specific business entity.
You can purge data in the development, QA, and production environments. When you purge data, a purge job is automatically created and triggered.
Rules and guidelines for purging data
Before purging data, consider the following rules and guidelines to prevent any data issues:
•No users have logged in to the business applications.
•No jobs are running.
•All workflow tasks are complete.
•Don't update the metadata of any business entity while the purge job is running.
•Don't train any machine learning models.
•If you purge deleted source records that contribute to the business IDs of active master records, the business IDs of those master records are updated. The business IDs are updated because deleted source records are first unmerged from their master records and then purged. To track master records whose business IDs are updated, configure publishing events for unmerge actions before you purge.
•If all source records of a master record are deleted, then purging the deleted source records also purges the master record.
Performance considerations for purging data
The performance of data purge depends on the data set, data volume, and POD-level limits.
To optimize performance of data purge, consider the aspects of data, such as volume and complexity, and purging limits.
Complexity of the data set
The performance of purge jobs depends on the size and complexity of your data set, such as the number of relationships, source records associated with master records, number of record pairs generated for matching, and historical data.
Small volume data purge
When you purge up to 1,000 master records or deleted source records, the purge job runs immediately without queuing.
Large volume data purge
You can purge up to 1 million master records at a time. To purge larger volumes, run the purge job multiple times.
You can purge up to 10 million deleted source records. To ensure proper scheduling and resource allocation for purging volumes exceeding 1 million records, contact Informatica Global Customer Support.
For information about viewing the number of master records and source records in your organization, see Monitor MDM SaaS.
Concurrency limits at the POD level
Informatica limits the number of concurrent purge jobs on a POD to three. As a result, your purge job might be queued if multiple jobs run concurrently on a POD.
Setting up data purge
You can purge all data from the data store, including historical data and records that belong to a specific business entity.
1Click Global Settings > Purge Data.
2Click Start.
3In the Purge Data dialog box, select one of the following options:
- Purge all. Permanently deletes all business entity records, source records, relationships, search, report, and hierarchy indexes, and history data.
- Purge the history of all records. Permanently deletes all historical data from business entity records, hierarchies, and relationships.
- Purge records that belong to a specific business entity. Permanently deletes business entity records or deleted source records of a specific business entity along with their dependencies based on the records purge option that you select.
4If you selected Purge records that belong to a specific business entity, perform the following steps:
aSelect a business entity.
bSelect one of the following options for records to purge:
▪ Purge all records. Permanently deletes all the business entity records, relationship records, search, report, and hierarchy indexes, and history data of the specified business entity.
Note: You can't purge invalid relationships where one of the records in the relationship doesn't exist.
▪ Purge deleted source records. Permanently deletes only the deleted source records of the specified business entity along with their dependencies, such as relationships, associated inactive master records, history, and match details.
5Click Next.
6Enter your organization ID, then click Submit.
You can monitor the status of the job on the My Jobs page. For more information about monitoring jobs, see Monitor Jobs.