For Z/OS system, I can create a physical file compressed with SMS.
The physical file store the SAS database. If i compress the file with Zedc feature (Zip compression code) i save more than 88% of disk space and lot of CPU because i remove SAS compression.
My problem is the SAS database on this compressed physical file is not a recognized SAS structure file.
To use this file, i have to code LIBNAME xxxx V9SEQ engin, so i have to change all SAS programs who use this SAS database.
I believe, it is a good idea if you determine that the file is compressed with Zedc (very easy) you use automaticaly the good SAS Engine.
A primary objective/benefit here is reduced CPU-consumption on the IBM z System general CP...consider a VOTE-UP!!
SAS software on IBM z/OS has the ability to invoke zEDC data-compression as an API-request, not only with SMS-managed output file though (i.e., invoked internally). The size-reduction observed with other types of zEDC-compressed data algorithm is anywhere from 60-80% which is substantial, while offloading the CPU-consumption.
And consider the CPU-consumption for this type of alternative data-compression is executed as an external card/feature and not as a general CP function, thereby not contributing to the overall 4-hour rolling-average MSU consumption, used by IBM and ISVs to charge clients.
For these observations, SAS INSTITUTE is strongly encouraged to explore IBM z System (mainframe) zEDC data-compression feature as a cost-effective alternative to the current-standard CONFIG/OPTIONS COMPRESS= functionality.
Good idea to reduce space and MSU consumption (zEDC feature instead of GCP).
Benefits for Sas Institute and its mainframe customers
For transparency, when the SAS compression option is chosen on a z/OS platform, SAS should check for the existence of and/or enablement to use zEDC and if able to use zEDC compression transparently.
zEDC compression of data will enable SAS to run even faster with more work getting done per I/O. It is essential that SAS move to automatically exploit zEDC whenever possible.
This would improve the usage of storage and CPU resources, but to the benefit of SAS's customers. The reduction in CPU usage via offloading compression to the specialty engine may lead to a reduced 4 hour rolling average, which in turn could lead to smaller software invoices from IBM and other software vendors.
I/S Management Strategies, LTD
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.