Peter Hunkeler wrote:

>And there is the (maybe) unexpected and (IMHO) unlucky relationship that 
>REGION=0K or 0M implies MEMLIMIT=NOLIMIT.

True for the word 'implies' [1] :

"If no MEMLIMIT parameter is specified, the default is the value defined to 
SMF, except when REGION=0K/0M is specified, in which case the default is 
NOLIMIT."

But there is another catch:

"Unlike the REGION parameter, MEMLIMIT=0M (or equivalent in G, T, or P) means 
that the step can not use virtual storage above the bar."

and 

"A specification of REGION=0K/0M will result in a MEMLIMIT value being set to 
NOLIMIT, when a MEMLIMIT value has not been specified on either the JOB or EXEC 
statements, and IEFUSI has not been used to set the MEMLIMIT."

MEMLIMIT=NOLIMIT (specified or not) is really *no* limit, but MEMLIMIT=0<scale> 
is simply zero, nil, nada, nothing memory.

I don't have limitless memory to handle this mixup... ;-D

You can use SMF30SFL, SMF30RGN, SMF30MEM, SMF30MES and friends to conduct 
JOB/TSO/STC autopsies.

Groete / Greetings
Elardus Engelbrecht

[1] - For me the word 'implies' means in this context: 'fall back to default, 
because nothing was specified at all.'

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

Reply via email to