> We're looking to "clean up" our Stand-alone Dump process, and get
> "up to snuff" for z/OS 1.13 and beyond. Toward that end we are
> studying the latest Tech Doc from IBM:
>
> http://www-03.ibm.com/support/techdocs/atsmastr.nsf/WebIndex/TD103286
>
> But one question (so far) remains: What are the minimum ASIDs we
> should specify nowadays for a SAD? What should also be in the "nice
> to have" list?
There is no need to worry about the MINASID=ALL vs. MINASID=PHYSIN
option. As of z/OS 1.8, there is no longer any physical swapping,
so ALL and PHYSIN have the same effect.
As for the DUMP= options, there really are no good
recommendations. Anything you do is a tradeoff between
dump time and size vs. first failure data capture.
To dump all paged out storage:
DUMP=('DSP OF ASID(ALL) ALSO PAGETABLES OF DATASPACES X
ALSO SP(ALL) IN ASID(ALL) X
ALSO HIGH VIRTUAL IN ASID(ALL)'), X
Some of our test systems use:
DUMP=('DSP OF ASID(1,'TCP*','CTTX','APPC','GRS','JES*', X
'IRLM*','DB?1IRLM','WLM','RRS','IXGLOGR','ZFS*','VTAM*',X
'OMVS') ALSO SP(ALL) IN ASID(1,'DUMPSRV','JES*','VTAM*',X
'XCFAS','GRS','RRS','IXGLOGR','ZFS*','OMVS')'), X
Which is not to say that this is a "good" choice. Usually
what happens is that someone complains about some stuff
that wasn't in some dump, and that results in it getting
added to this DUMP= list. And I see that as of yet, we
don't specify HIGH VIRTUAL for any these spaces.
Jim Mulder z/OS System Test IBM Corp. Poughkeepsie, NY
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
