Sorry Kees, you lost me. If (as recommended) all the page datasets are the same size, and the occupancy is n%, why does the question arise ???. ASM will (presumably) balance the allocation given similar response from the disk. What does it matter what the underlying volume model is - apart from pathing considerations ???. The total slot count should (nominally) be the same - total and occupied.
If you do a one-for-one replacement of (full) mod-3 with 4 Gig allocations on mod-9, the occupancy drops by about half. Which situation are you contemplating ???. Shane ... ----- Original Message ----- From: "Vernooy, C.P. - SPLXM" > Hello, > > The recommandation was (possibly still is) not to fill a pageds for more > than 25%, to allow ASM's contiguous slot algorithme to always find enough > contiguous free slots. Does this recommandation also apply to pageds on > 3390-09 models? > > I suppose the 25% figure statistically gave ASM a good guarantee to always > find nn contiguous slots on a 3390-03. > > A pageds on 3390-09 can be almost twice as large as on 3390-03 (5825 cyls > versus just over 3300 cyls). If ASM needs a maximum of nn slots, I suppose I > can fill a pageds on 3390-09 far more than 25% while still guaranteeing nn > contiguous free slots for ASM. > > Is this true, is there a new recommandtion for a maximum size (4 GB) pageds > from ASM's congiguous slot algorithmes point of view? Performance is of > course taken care of by PAV's. ---------------------------------------------------------------------- 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
