Type 30 records are frequently "split" when there are more DD segments
that can fit in a physical SMF record, known as MULTIDD='Y'
to MXG users, and there most defintely can be other SMF records
between the multiple phyical 30 records, as it takes finite
time to create each of the MULTIDD records.
You can detect this is the first, middle, or last of a
series of split records by the count fields (for DD
segments, SMF30EON/EOR/EOS).
It's also possible to have an SMF 30 record split for
reasons other than too many DD's: an abundance of OMVS
segments and/or MULC, APPC, or ARM segments can also
split a logical SMF 30 into multiple, separately written,
physical SMF 30 records.
Barry Merrill
P.S. Of course, the holding of a partial record from today's SMF file
until tomorrow's processing, after detection of an incomplete
SMF 30 logical record, requires additional logic, provided in
MXG's case by its "SPIN" logic to hold incomplete jobs.
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html
