Before You Create the Data Integration Service
Before you create the Data Integration Service, complete the prerequisite tasks for the service.
Perform the following tasks before you create the Data Integration Service:
- •Set up the databases that the Data Integration Service connects to.
- •Create connections to the databases.
- •If the domain uses Kerberos authentication and you set the service principal level at the process level, create a keytab file for the Data Integration Service.
- •Create the associated Model Repository Service.
Create Required Databases
The Data Integration Service can connect to multiple relational databases. The databases that the service can connect to depend on the license key generated for your organization. When you create the Data Integration Service, you provide connection information to the databases.
Create the following databases before you create the Data Integration Service:
- Data object cache database
- Stores cached logical data objects and virtual tables. Data object caching enables the Data Integration Service to access pre-built logical data objects and virtual tables. You need a data object cache database to increase performance for mappings, SQL data service queries, and web service requests.
- Profiling warehouse
- Stores profiling information, such as profile results and scorecard results. You need a profiling warehouse to perform profiling and data discovery.
- Workflow database
- Stores all run-time metadata for workflows, including Human task metadata.
For more information about the database requirements, see
Application Service Databases.
The Data Integration Service uses native database drivers to connect to the data object cache database, the profiling warehouse, and source and target databases. To establish native connectivity between the service and a database, install the database client software for the database that you want to access. For more information, see
Configure Native Connectivity on Service Machines.
Create Connections to the Databases
The Data Integration Service uses connections to access the databases. You specify the connection details when you create the service.
When you create the database connection in the Administrator tool, specify the database connection properties and test the connection.
The following table describes the database connections that you must create before you create the Data Integration Service:
Database Connection | Description |
|---|
Data object cache database | To access the data object cache, create the data object cache connection for the Data Integration Service. |
Workflow database | To store run-time metadata for workflows, create the workflow database connection for the Data Integration Service. |
Profiling warehouse database | To create and run profiles and scorecards, create the profiling warehouse database connection for the Data Integration Service. To create and run profiles and scorecards, select this instance of the Data Integration Service when you configure the run-time properties of the Analyst Service. |
Create the Service Principal Name and Keytab File
If the Informatica domain uses Kerberos authentication and you set the service principal level for the domain to process level, the domain requires an SPN and keytab file for each application service that you create in the domain.
Before you enable a service, verify that an SPN and a keytab file are available for the service. Kerberos cannot authenticate the application service if the service does not have a keytab file in the Informatica directory.
For more information about creating the service principal names and keytab files, see the Informatica Security Guide.
Create Associated Services
The Data Integration Service connects to the Model Repository Service to perform jobs such as running mappings, workflows, and profiles.
Create the Model Repository Service before you create the Data Integration Service. When you create the Data Integration Service, you provide the name of the Model Repository Service. You can associate the same Model Repository Service to multiple Data Integration Services.