On Thu, 28 Aug 2008 01:54:51 -0400, Alan Altmark <[EMAIL PROTECTED] > wrote:
... > >Look in the archives of this listserver and you will find z/VM Developme nt >stating that the volume label is safe from CP, even if cyl zero is >allocated as page or spool. > >From a system management point of view, however, and for your sanity's >sake, however, I suggest that cylinder zero SHOULD always be allocated a s >PERM. Adding one additional cylinder will not help the system in any >meaningful way and only serves to create controversy and risk. > >So, "someone" is wrong from a technology perspective, but right in terms >of Best Practice. > >BTW, CP will happily hand out cylinder 0 to a guest if it is marked as >T-disk. Dumb, perhaps, but true. > >Alan Altmark >z/VM Development >IBM Endicott >======================== ========================= ======================== For what it's worth, ICKDSF (which CPFMTXA uses) was changed several year s ago and no longer permits allocation cylinder 0 as tdisk, though the prob lem could still occur with volumes formatted before that change (or by some other utility). Page and Spool usage of cylinder 0 has worked properly and been supported by the VM/XA -> VM/ESA -> z/VM lineage the entire time I've been with VM (th at is, since 1986). In that entire time, I'm unaware of even one problem involving page or spool usage of cylinder 0 (and I've been the team leade r / subsystem owner of the most affected area for 18+ of those years, so I wo uld have heard). Still, as long as there's any risk (with tdisk allocations) and confusion and misunderstanding, I won't argue with the recommended "best practices" of avoiding cylinder 0 simply to avoid having to have these discussions ever y few months (or at least, to keep them short). - Bill Holder, z/VM Development, IBM