Thank you ALL for your help. I finally got back from a looong road trip including a week at POK and had time for this again.
Thank you for the IEBDG and SPACE=(TRK,(1,1)) suggestions. Yes, perfect, I got three volumes allocated before I got my expected SB37. Now for the bad news. Even deleting the dataset with IEFBR14 DISP=(OLD,DELETE) DFSMS in its wisdom cut three one-volume SMF 17 records, not the hoped-for single three-volume SMF 17 record. (Same for IDCAMS as indicated below.) I did get to see three nice SMF 15 EOV records -- not sure I have noticed any of those before. And thanks to your help, yes, I am definitely getting a three-volume dataset without any interference from SMS. I've about given up ... Charles -----Original Message----- From: IBM Mainframe Discussion List [mailto:[email protected]] On Behalf Of Bill Godfrey Sent: Friday, April 08, 2016 8:01 AM To: [email protected] Subject: Re: How force dataset non-SMS? On Wed, 6 Apr 2016 08:35:10 -0400, Charles Mills <[email protected]> wrote: >Oh guys, the reason is really prosaic: I want an SMF 17 with multiple VOLSERs >in it. > I have noticed that when multi-volume data sets are deleted using IDCAMS or the TSO DELETE command, there are separate type 17 records written for each volume, with one volume in each record. But when they are deleted using unallocation, as in DISP=(OLD,DELETE) or TSO FREE, there is just one type 17 record containing all of the volumes. I have seen this in the current z/OS and in MVS 3.8. It depends on whether the parameters given to the SCRATCH macro include all the volumes or just one at a time. It's something you might want to be aware of once you get past the SMS issue. ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
