Data Services All Products
The Model repository requires a minimum values for the maximum heap size setting. Set maximum heap size to the recommended value of 1 GB, and the maxPermGen size to 512 MB. Lower settings generate an error.
When you select a parent group in the Create Group wizard, the new group appears in the Native folder but is not nested under the parent group.
Workaround: After you create the group, clickand select a parent group.
The DTM process does not create DTM log files for mappings included in workflow Mapping tasks when the following conditions are true:
When you remove a node with the compute role from a Data Integration Service grid or disable the compute role on a node in the grid, the node is no longer listed in the Compute view for the service. However, the service retains the previously configured compute values for the node. If you add the node back to the grid or enable the compute role again, the node is configured with the previous values instead of the default values.
Workaround: Verify that the properties on the Compute view use the correct values after adding the node back to the grid or after enabling the compute role again.
The Email Service does not support environment variables for processes. If you configure an environment variable for the Email Service process, then it will not affect the process.
When you run the infacmd sch updateschedule command, the schedule end date changes to No End Date. The end date changes regardless of whether you specify a value for the -ed option.
Workaround: Use the
Schedulesview in Informatica Administrator to update schedules.
When you schedule a job to run every 23 hours, the Scheduler Service might run the job at the wrong time.
If you run multiple concurrent mappings on a Data Integration Service grid configured to run jobs in separate remote processes and the Model repository is not configured to store run-time statistics, some of the mappings might fail to run with the following error:
Workflow recovery fails when the workflow database is a Microsoft SQL Server database that uses a non-default schema.
Mappings are not equally distributed across compute nodes in the grid when the following conditions are true:
In a domain that uses Kerberos authentication, some views in the Administrator tool display the following message to users who are assigned the Operator role:
This occurs even when the Model repository is configured.
Model Repository is not configured. Please contact the Administrator.
Workaround: Assign the Operator users and group the Administrator role for the Model Repository Service that is configured for monitoring.
In a mapping with a flat file data source that includes a column with double data type, the Data Integration Service erroneously reads data that should be rejected because it contains non-numeric characters. For example, the row should be rejected when it contains a value such as
12345678901234567890123456ab, but the Data Integration Service fails to reject the row. Instead, it reads the numeric characters and ignores the non-numeric characters.
The consolidated log file for a mapping might contain the incorrect DTM log file when the following conditions are true:
Workaround: Configure the Mapping task to save the Mapping task log file by timestamp.
Mappings run on a Data Integration Service grid might hang indefinitely when the following conditions are true:
Workaround: Enable the Resource Manager Service to continue running the mappings.
In the Monitoring tool in a domain that uses Kerberos authentication, the Log Out menu does not log users out of the Monitoring tool.
Workaround: To log out of the Monitoring tool, close the browser window.
In a domain that uses Kerberos authentication, when you log in to the Administrator tool after a session expires, the
Monitortabs might display a login page.
Workaround: Log out from the Administrator tool, and then log in again.
When you update the compute role on a node assigned to a Data Integration Service grid and then recycle the Data Integration Service, you might encounter inconsistent behavior across the Informatica client tools. For example, mappings might fail to run in the infacmd command line program but succeed in the Developer tool.
Workaround: Restart the domain.
After you join a node to the domain, the Administrator tool takes 10 to 15 seconds to display the properties of the node.
If you force a running workflow application to stop and you redeploy the workflow application, the application fails to deploy. You cannot run another workflow instance until you drop and recreate the workflow database contents. The issue arises when you abort the workflow while the Data Integration Service attempts to create a large number of tasks.
You cannot use the "wait" option [-w] when you run the infacmd wfs abortWorkflow command or the infacmd wfs cancelWorkflow command.
In a Kerberos domain, mappings fail to run on a Data Integration Service grid configured to run jobs in separate remote processes.
Workaround: Configure the Data Integration Service to run jobs in separate local processes.
If you import a relational data object and you use a control file to specify a connection for the object, the operation might disregard the connection that you specify. The issue arises when the following conditions are true:
When the conditions are true, the import operation assigns the relational data object to the database connection with the original name. If the object is a reference table, the import operation fails.
Workaround: Remove the original database connection from the domain and import the data object again. Or, update the database connection in the Developer tool.
A Data Integration Service grid configured to run jobs in separate remote processes does not use a secure connection to communicate with remote DTM processes even though the domain is enabled for secure communication.
When you run a mapping on a Data Integration Service grid configured to run jobs in separate remote processes, the Monitor tab of the Administrator tool might indefinitely list the mapping state as Running even though the infacmd command line program and the mapping log indicate that the mapping failed.
If you enable the Create or replace target at run time option, and the mapping target connection is JDBC, the Data Integration Service converts the incoming datatypes to ANSI datatypes which might not be supported by the target database. You might see an error that contains a message like:
After you configure the Log Collection Directory property for a node, you cannot clear the Log Collection Directory property.
If you run web service requests on a Data Integration Service grid and you incorrectly configure the external HTTP load balancer to use nodes with the service role only, the Data Integration Service does not redirect requests to nodes with both the service and compute roles. Some web service requests dispatched to the node with the service role only might fail.
Workaround: Configure the external HTTP load balancer to use nodes with both the service and compute roles.
Users can use the IMFCryptographer file to read network packets that contain information about Model repository objects that they do not have permission to see. To prevent this access, use Kerberos authentication to prevent transmission of passwords between client and server.