You need to create filters for your large data sets. The volume mount analyzer default for large is 600 MB, the amount of data that can be contained on a single cartridge system tape, compacted three-fold with Improved data recording capability (IDRC). You can supply your own value for LARGE. Look for programs or HLQs that tend to generate primarily large data sets, and exclude those entire filters. Consider as LARGE those data sets that are large enough to make efficient use of available tape capacity. Use the LARGE value used by the volume mount analyzer as an indicator rather than an absolute cutoff value. In the sample, most of the tape data sets are candidates for the DASD buffer, but some of the very large data sets should continue to go directly to tape.

For the sample, program name identifies some large data sets that account for many mounts. Typically, program name can be used as a filter to identify mounts for large data sets that can be excluded with ACS filtering techniques. Often, one program is responsible for creating multiple data sets stacked on tape. Intercepting mounts for multi-file data sets in this way simplifies your design problems linked with breaking up these stacked data sets.

Use the Top Report to identify efficient filters for capturing many large data sets. The >LARGE and AVG SIZE columns give a good indication of the tendency of particular programs or applications to generate LARGE data.