In a message dated 9/13/2007 9:43:43 A.M. Central Daylight Time, [EMAIL PROTECTED] writes: >I noticed that both format 1 and format 3 DSCBs use 'lower limit' and 'upper limit' to describe the scope of each extent and they're all in CCHH format, not CCHHR as I assumed. >Does that mean all data sets in z/os will consist of complete tracks? Yes. The unit of allocated space is one track in z/OS and all its predecessor variants (MVT, etc.). >We cannot, say, allocate a half of a track to one data set and the left to another one. We cannot use standard system services, such as DADSM, to do that. We could provide our own space suballocation scheme that would do what you said. >I know we can allocate space in other units such as BLKS, KB, RECORDS... but I guess they'll be rounded to tracks or cylinders by system at end) Correct. >But I'm not sure because I once read from the manual about VTOC: A VTOC consists of complete tracks. True. A VTOC contains a Format 4 DSCB (it's always the first one in the VTOC) that contains the lower and upper bounds of the extent containing the VTOC itself. >If that is true for all data sets, why the manual took the trouble to tell me again? I don't know to what you are referring here. >However, if it's not true, It is true for all data sets that are allocated using DADSM in z/OS. >how can you find the start of a data set by CCHH? The data set always begins at the beginning of the first track in the extent descriptor for the first allocated extent in that data set. I keep saying "in z/OS" because VM allows a VTOC to be less than a whole track. The Format 4 DSCB of a VTOC created by VM has a VTOC extent of only one track. The next DSCB on that track is a Format 5 DSCB that describes the free extents on the volume, and it shows that there is no free space on the volume. This is because VM manages the space allocation on such volumes. There are also no more DSCBs on this track, so this kind of VTOC consists entirely of only two DSCBs and not even a whole track. This kind of VTOC allows other operating systems, such as z/OS, that are VTOC-dependent to access the volume without any error indication. You can mount such a volume in z/OS, the system will read the volume label (which has the CCHHR of the VTOC), then read the Format 4 and 5 DSCBs, discover that there is no free space on the volume, and then never attempt to allocate any space on it. At least what I just wrote was the case in the 1980s when I studied a VM-created VTOC. By the way, the volume label does not contain the CCHH of the VTOC. It contains the CCHHR of the Format 4 DSCB, which is always supposed to be the first recorded block in the VTOC. This means that a VTOC could theoretically begin at some record number other than 1, and thus the first track in the VTOC, and only this first track, might not be entirely filled with DSCBs. Theoretically. Bill Fairchild Plainfield, IL
************************************** See what's new at http://www.aol.com ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
