That seems to imply that the x-link part of CSE can't be on FCP attached drives... I thought I've read that z/VM can be installed entirely on FCP attached drives. But I didn't read that much in detail if there were functions that wouldn't work in that world. Perhaps a future release would address that issue?
Tom Duerbusch THD Consulting >>> Kris Buelens <kris.buel...@gmail.com> 3/9/2009 1:15 PM >>> Shall I plead a tiny bit for CKD? - VSAM has the ability to store (replicate?) the lowest level index set in front of the track where the data reside. This to avoid seek, the idea was that the higher level index blocks would be buffered in storage. But, who cares/who has time nowadays to get that last bit of performance improvement. - our own beloved VM has one function not working on FBA... Please guess.... The x-link part of CSE needs CKD. The locking mechanism doesn't need Reserve/Release. With one, complex, I/O the XLINK SW is able to test if a cylinder is free for e.g. a R/W LINK and if so, mark it as being used. Kind of TestAndSwap. Impossible on FBA I've been told. 2009/3/9 David L. Craig <d...@radix.net> > > On Sun, Mar 08, 2009 at 02:10:24PM -0400, Richard Troth wrote: > > > > ECKD (today) is an IBM hardware solution for a software problem. > > Originally, CKD exposed not only counts and keys (thus the acronym) > > but more significantly tracks and records. NO ONE but MVS (and TPF) > > has a firm requirement for that. What I mean is that CP and VSE can > > at least tolerate a lack of tracks and records. (They can run from > > IPL to shutdown on things like SAN.) More significantly, CMS, Linux, > > and Solaris (or UTS or AIX or USS) explicitly THROW AWAY the track > > and record semantics that our precious storage subsystems worked so > > hard to present on the channel. They can't use it! They just care > > about the data on the disk, not its geometry. > > Are you implying VSE VSAM KSDS files do not utilize CKD > architecture and/or there is no net performance advantage > to that in this day and age? Maybe I need to look at the > effort to migrate to 3370... > > -- > > May the LORD God bless you exceedingly abundantly! > > Dave Craig > > - - - - - - - - - - - - - - - - - - - - > "'So the universe is not quite as you thought it was. > You'd better rearrange your beliefs, then. > Because you certainly can't rearrange the universe.'" > > --from _Nightfall_ by Asimov/Silverberg -- Kris Buelens, IBM Belgium, VM customer support
