In a recent note, Bruce Black said: > Date: Thu, 25 May 2006 10:10:26 -0400 > > > Wouldn't it blow you out of the water when you tried writing new > > records in that space? AFAIK the VSAM PUT logic expects the space to > > be CI formatted. > If allocation wrote the EOF into the VSAM cluster, then VSAM OPEN or EOV > would immediately format the CA with CIs as soon as it is needed, > disregarding any previous contents, including the EOF. So the EOF > would just be unnecessary work for a VSAM allocation. > Finally, some good sense in this thread. I was dismayed by the presumption of some readers that the design would put the operations in the wrong order -- first CI format, then write the EOF; naturally it works better the other way around.
My guiding rationale here is that when a new extent is allocated, the component (DADSM?) pays little attention to the former use or content of the space allocated. If, by happenstance, the first track in the extent contains an EOF, I'd expect other components to be prepared to deal with it. Empirically, this appears so -- I've never heard of a software failure because an unitialized extent contained an unexpected EOF. -- gil -- StorageTek INFORMATION made POWERFUL ---------------------------------------------------------------------- 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
