Thanks all. I think I am either 100% safe or about 99.99% safe without changing the JCL. If it turns out to be .01% unsafe I will cross that bridge when I come to it.
Barbara, "my application" is named the IBM Enterprise COBOL compiler so I have no knowledge whether it does an EOV, but I would assume not. Would not make much sense to document "must be one volume" and then do an EOV. (But then, moving DD overrides between versions does not make much sense.) Lizette, thanks. Would just as soon not have to tell customers "you have to define this special SMS class" (and fight with their SMS naming convention Nazis?) Charles -----Original Message----- From: IBM Mainframe Discussion List [mailto:[email protected]] On Behalf Of Ron Hawkins Sent: Wednesday, October 16, 2013 10:37 PM To: [email protected] Subject: Re: Unit-count default with UNIT=group-name? Charles, Yes you have implicitly requested one volume. Assuming the DATACLAS has 1 or blank for the unit count field you have effectively coded a unit count of 1 in this JCL. You would have to code UNIT=(SYSDA,p) where "p" is >1, or multiple volsers to increase the unit count. Ron > -----Original Message----- > From: IBM Mainframe Discussion List [mailto:[email protected]] > On Behalf Of Charles Mills > Sent: Wednesday, October 16, 2013 3:32 PM > To: [email protected] > Subject: [IBM-MAIN] Unit-count default with UNIT=group-name? > > I'm looking at the Enterprise COBOL V5.1 U/G which says that SYSUT5, > 6, 7 and 15 must all be single-volume allocations. > > If I code SPACE=(CYL,(2,2,0)),UNIT=SYSDA is that implicitly a request > for only > one volume, or might z/OS give me one cylinder each on two volumes or > a secondary allocation on a second device (assuming, of course, that > SYSDA were multiple volumes, as it usually would be)? The example in > the U/G shows only UNIT=SYSDA,SPACE=(CYL,(1,1)). The JCL Reference > says that UNIT=device-address is a request for only a single volume > but appears to be > silent on UNIT=group-name. > > It's machine-generated JCL so it is not so simple as "why don't you > just code a > unit count of 1 and quit worrying about it?" A unit count will require > a program change. ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
