Creating database partition groups

In a multipartitioned environment, we have a separation mechanism named partition groups. Inside partition groups, we can isolate tables, based on transaction type, or we can group tables that are frequently used in joins together. Data in partition groups are by default spread on all partitions on which they are defined, using a round-robin distribution method, unless you decide to spread data by your own rules using customdistribution maps. However remember that, not every table is a good candidate for spreading on all partitions, especially small tables or huge tables on which we are constantly involved in ad hoc queries. In these cases, the inter-partition communication mechanism used in multipartitioned environments could act as a performance bottleneck. Any complex environment comes with its advantages and disadvantages. Therefore, you have to take into consideration these aspects and be careful with the application design.

Getting ready

In the following...