At run time, the Data Integration Service creates separate directories for each value in the FileName port and adds the target files within the directories.
The file rollover process closes the current file to which data is being written to and creates a new file based on the configured rollover value.
you can configure the following optional execution rollover parameters at the design-time based on time and size:
Stream Rollover Time in Hours. Specify the rollover time in hours for a target file when a certain period of time has elapsed.
Stream Rollover Size in GB. Specify the size in GB for a target file when the target file reaches a certain size.
When a target file reaches the configured rollover value, the target file is rolled over and moved to the specified HDFS Complex File target location.
The Spark engine creates sub-directories in the specified HDFS comples file target location for each value in the FileName port.
The Spark engine moves the rolled over target files to the sub-directories created for each value in the FileName port in the specified Complex file target location.