When you map the NEXTVAL or CURRVAL output fields, ensure that the data type of the mapped field is appropriate.

When you run the mapping in the Mapping Designer, the current value is not saved so each time you run the mapping, it begins with the initial value.

When you run the task in the

mapping

task wizard, you can edit the current value to start the sequence with a specified value.

Before the Sequence transformation runs,

Data Integration

checks for sufficient space between the starting value and the end value. If you set an end value that's too close to the starting value, the mapping fails with a sequence overflow error. This error can occur even if the starting value plus the number of rows processed would be less than the end value. If you encounter a sequence overflow error, increase the end value.

You can't use a Sequence transformation in a mapplet.

To use the Sequence transformation in advanced mode, the Secure Agent must run on a virtual machine.

A self-service cluster might intermittently fail to connect to the agent, which can cause the mapping to fail. Error messages related to this appear in the session logs.

In advanced mode, generated values might not increase monotonically. Values are generated based on the order that the Spark engine processes the data.

In mappings in SQL ELT mode, you can't use shared sequences.

In mappings in SQL ELT mode, a Sequence transformation needs to be connected directly to a downstream Expression transformation or target.