Partitioning indexes

You can partition primary and secondary indexes into multiple KSDS data sets. The index partitioning is done by an index key or partition selection routine.

In many large databases, the secondary index, and in some cases even the primary index, can be as large as the target database. Also, large KSDS data sets take longer to recover and rebuild after a reorganization. Partitioning indexes increase the size of the indexes while allowing utilities to maintain index KSDS data sets more efficiently.

The number of partitions in the secondary index can be different than the number of partitions in the target database. For primary indexes, the number of index partitions must be the same as the number of partitions in the target database. Additionally, the key ranges of the primary index and target partitions must be the same when high keys are used.

Secondary indexes are supported with both HDAM and HIDAM partitioned databases. Shared secondary indexes, where multiple secondary indexes are in a single KSDS, and non-unique secondary indexes are not supported.

You can partition large secondary indexes to increase the size of both the index and the target database.