Column-organized tables in a Partitioned database environment

Using column-organized tables in a massively parallel processing (MPP) environment may improve performance or reduce cost, depending on the nature of your workload.

There are no general guidelines to determine whether a particular workload should use column-organized tables. This should be examined on a case by case basis. In general, the same guidelines for performance in non-massively parallel processing (MPP) environments apply in MPP environments. In certain scenarios, vector processing of column data in a partitioned database environment provides significant improvements to storage, query performance, and ease of use through simplified design and tuning.

If you already have column-organized tables, but not in a massively parallel processing (MPP) environment, consider moving to an MPP environment. The following are some of the benefits of column-organized tables in an MPP environment:

• Column-organized table storage capacity can be increased. The maximum column-organized table size in a non-MPP environment is 64TB of compressed data . A partitioned database environment distributes the table across additional database members. Each new member allows the maximum table size to increase.
• The columnar workload's resources such as processors, memory and disks are increased by having multiple servers in parallel. Consider this factor of MPP only when the maximum resources of your non-MPP environment have already been applied. Consider whether or not the target workload will benefit from MPP parallelism, what other options exist to achieve performance objectives, and the cost and operational considerations of each option.

You must decide whether or not column-organized tables in an MPP environment will meet the performance goals of your target workload and be cost efficient. For example, MPP environments may increase operational costs since there will be multiple database partitions to monitor and maintain.