How to estimate compression savings achieved with option REORG

If you run DSN1COMP with the REORG option on small data sets, the resulting estimates might vary greatly from the estimates that are produced without the default REORG option. Alternatively, if you run DSN1COMP and specify a small number (n) for ROWLIMIT, the estimates might vary greatly from the estimates that are produced without REORG.

DSN1COMP does not try to convert data to the latest version before it compresses rows and derives a savings estimate.

Without the REORG option, DSN1COMP uses the first n rows to fill the compression dictionary. DSN1COMP processes the remaining rows to provide the compression estimate. If the number of rows that are used to build the dictionary is a significant percentage of the data set rows, little savings result. With the REORG option, DSN1COMP processes all the rows, including those that are used to build the dictionary, which results in greater compression.

The DSN1COMP utility determines possible saving estimates at the data set level for a unique partition only. Therefore, if DSN1COMP is run against an image copy data set that contains several partitions or against a single partition of partition-by-growth table spaces (PBGs), the results will be different from what the REORG utility would produce.