Data filtering
Why data filtering?
Actually there are a number of tools a user could use to filter data e.g. DFSORT, Syncsort, CA-Sort, QPAC, Easytrieve, SAS etc.
However the biggest problem arrises when trying to use any of these products in conjunction with e.g. SMF processing or subsystems to trap output.
The DFSORT is unable to process SMF SYS1.Man1 and of no use e.g. to trap data within a subsystem. The filtering does not already happen e.g. at the E15 EXIT level but rather on the E35 level. DFSORT as well does not understand relocatable sections for mysteries reasons. Other products have shortcomings when trying to filer data e.g. IRRADU00 etc.
Hence a couple of interfaces were introduced to overcome such problems. IRRADU00 poses one of the worst problems when trying to obtain only those decoded SMF records you really need. Already today there are over 8700 fields a user ought to be able to select. However the user has no such possibilities, he/she must offload ALL or NOTHING to a flat file, which makes no sense.
- Have you ever tried to run IBM's IRRADU00 program with >20 Mio/daily RACF SMF records and even tried some of the supplied ICETOOLS? You mostly run out of luck due to the crazy disk space required as you have no filter options. Only our subsystem and or our SMF exits and or RA/7 can help you. It saves you tens of thousands on I/O's and disk space!
Filtering records
Using symbolic names
SMF records
IRRADU00 records
IRRDBU00 records
DCOLLECT records
