The Data Redistribution utility uses a partitioning map to do the data redistribution. It can create its own target partitioning map, or you can provide one for the utility to use. If you create one, the entry or entries determine the type of nodegroup that results from the data redistribution:
If the target partitioning map has more than one database partition, all tables in the nodegroup must have a partitioning key defined.
The target partitioning map can only contain databases which are defined in the SYSCAT.NODEGROUPDEF catalog table, excluding those with an IN_USE value of 'T'. ('T' means that the partition is not in the target partitioning map.) For all partitions that have an IN_USE value of 'D' (meaning to drop), and which also do not appear in the target partitioning map, these databases are dropped when the redistribution operation has completed successfully.