Edward E Jaffe wrote: > That is -- and AFAIK has always been -- the VIO trade-off. CPU time for > clock time.
Yes, that's right. And VIO should be compared and contrasted with other DIM techniques for Batch. I've a feeling I wrote a blog post about it a few months ago. I should've because I used to present on VIO - especially the DFSMS management thereof. The "DIM Coffee Table" book came about as the result of a number of studies Development did - in the late 1980s. I DIDN'T participate. But it led onto residencies I DID participate in. Anyhow the book documented many performance tests that mapped out the performance characteristics (including CPU) of many DIM techniques. And, as noted, the reason there's the CPU cost is the emulation of disk devices - rather than the "block mode" operation of eg VSAM LSR. I also note that, like some other techniques, VIO is usually seen as a more or less "write once then read once" use of memory. I.e. not dense in terms of I/Os avoided per MB of memory deployed. I wonder if the economics of today make that much less of an issue. I think the answer would have to be that (UK) :-) English word "somewhat".* Martin Martin Packer Performance Consultant IBM United Kingdom Ltd +44-20-8832-5167 +44-7802-245-584 [EMAIL PROTECTED] * "Somewhat" is the younger sibling of "it depends" and the (less evil) twin of "quite". :-) Unless stated otherwise above: IBM United Kingdom Limited - Registered in England and Wales with number 741598. Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU ---------------------------------------------------------------------- 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
