Creating and using MDC tables and block-based indexes

MDC provides automatic and continuous data clustering on more than one dimension, simultaneously. This method results in significant performance enhancement of queries based on dimensions defined for a table.

Getting ready

We'll have to examine the database configuration for our table and determine the best way to cluster the data. We also have to ensure we have enough space for this table.

How to do it...

1. Determine dimension candidates:

   At the design stage, you might have to determine the dimensions with a data architect, to determine if an MDC is preferable.

   Good candidates for dimensions are columns with low cardinality, that is, the percentage of different values relative to the number of records. The columns should also have as much as possible a static value, and be used frequently in queries. For an existing table, you can do some research, using the RUNSTATS utility, as discussed further, in the Using runstats to determine dimension...