Just another view of this discussion: It doesn't matter where the data
sets are in relation to one another. The probability of a page fault to
be resolved in the first cylinder of PLPA is very low assuming that the
PLPA is optimally packed.
The real purpose of the recommendation was originally to minimize
seeking; it quickly became a simple way to avoid the need to properly
size the PLPA data set.
Combining the common area page data sets works, doesn't hurt, has nil
performance impact, and eliminates a sizing exercise. If only the
message text of ILR005E were not so ominous and the messages manual
explanation were not so obtuse. (The sizing recommendations in the
Initialization and Tuning Guide should be revisited as well.)
Maybe a doc change would make the issue go away for good.
FWIW,
Steve Samson
Ted MacNEIL wrote:
..
Your disk technology makes the single pack
PLPA/Common unnecessary nearly 10 years ago, so why continue to do it.
..
Channel pathing?
Why not?
If it ain't broke?
I have better things to do than to 'split up' things that have been together
for years.
My point was (all along): who cares?
Don't worry!
Be happy.
And, as for the original poster, why did/does management care?
It's like certain assembler coding constructs that are in general use.
They're understood: they work: they don't hurt.
Even if custom volume are all the rage, I would do it as a volume to fit
PLPA @ 1 cylinder, add COMMON.
-teD
In God we Trust!
All others bring data!
-- W. Edwards Deming
----------------------------------------------------------------------
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
