VPE

VPE does the following:

Eliminates redundant I/Os through buffering.
Uses memory above the 16 Mb line for any program. It is possible to have files completely loaded into ESA's large virtual storage areas.
Allows multiple files to share a single buffer pool (Local Shared Resources (LSR)).
Optimizes sequential buffers using Read Look-Aside and Deferred Write (DFW) techniques for batch jobs.
Uses Sequential Insert Strategy (SIS) and Deferred Write (DFW) techniques for VSAM files used by CICS.
Gives batch programs the advantage of VSAM Local Shared Resources (LSR) using a proprietary technique.

It is not necessary to modify your application programs to use VPE. VPE invocation is controlled in 2 ways, both external to the application:

Control statements in the VPE Rules Table, and/or
Special DD statements added to the application JCL.

VPE transparently gains control each time a VSAM data set is opened through an interface program, which is installed and removed by a batch job. An IPL is not required.

VPE provides reporting to give visibility to its benefits and to aid in tuning.