Way back when IBM first started putting dummy VTOCs on devices used by VM and ACP, I pointed out that relying on a Format 5 DSCB that said there was no space was not the correct way to do it. It needs an F1 DSCB that allocates all available space instead. In those days, relying on the F5 alone could have led to real trouble. Whenever the O/S DASD allocation routines, looked at a VTOC, they turned on the DOS VTOC bit, known also as the Damaged VTOC bit, which was used to indicate that the allocation routines were working on the volume. If, for some reason, that bit were already on when the routines started, they first built a new F5 by starting with everything free and processing the F1 and F3 DSCBs to re-allocate the space that was in use. No F1 meant that all space was free.
It looks like nothing has really changed. Regards, Richard Schuh > -----Original Message----- > From: The IBM z/VM Operating System > [mailto:[EMAIL PROTECTED] On Behalf Of Kris Buelens > Sent: Thursday, January 17, 2008 11:35 PM > To: IBMVM@LISTSERV.UARK.EDU > Subject: Re: S213 Abend Backing Up 530RES Using z/OS FDR > > Same here (for a VM resident) > DITTO/ESA for VM DVT - Display VTOC Line 1 of 2 > > Unit 0130 VFRRES 3390 with 3339 cyls, 15 trks/cyl, 58786 bytes/trk > > --- Data Set Name --- Ext Begin-end Reltrk, > 1...5...10...15...20...25...30. seq Cyl-hd Cyl-hd numtrks > *** VTOC EXTENT *** 0 0 0 0 0 0,1 > *** FREE EXTENT *** 0 0 1 3338 14 1,50084 > *** This volume is currently 0 % full with 50084 tracks > available Maybe MVS is looking at some other field to find > out if there still is free space > > 2008/1/18, Alan Altmark <[EMAIL PROTECTED]>: > > On Thursday, 01/17/2008 at 09:42 EST, Kris Buelens > > <[EMAIL PROTECTED]> wrote: > > > VM doesn't use VTOCs to describe disk contents; the CP directory > > > tells > > which > > > cylinders are allocated as minidisks, and the CP allocation map > > > tells CP > > which > > > cylinders to use for PAGE, SPOOL, etc. > > > > After a CPFMTXA FORMAT, I used DITTO to look at the VTOC > that *is* on > > the > > volume: > > > > --- Data Set Name --- sorted by NAME ----- Ext > Begin-end Reltrk, > > > > 1...5...10...15...20...25...30...35...40.... seq Cyl-hd > Cyl-hd numtrks > > *** VTOC EXTENT *** 0 0 0 > 0 0 0,1 > > > > *** FREE EXTENT *** 0 0 1 > 4 14 1,74 > > *** This volume is currently 1 percent full with 74 tracks > available > > > > I had always thought that a "full" VTOC was written by DSF > so that MVS > > would not attempt to allocate space on a CP-owned volume. > It is a 5 > > cyl 3390. > > > > Alan Altmark > > z/VM Development > > IBM Endicott > > > > > > -- > Kris Buelens, > IBM Belgium, VM customer support >
