On Tue, 4 Mar 2008 12:31:01 +0100, Vernooy, C.P. - SPLXM wrote: <----- snip -----> >Yes, as I understood too, it is fully managed by the hardware, not by >WLM. So this means that ASM may have to compete for exposures. >Possibly/hopefully ASM's I/O priority will give it the needed advantage. >Otherwise Shane's remark about SECP may be not the joke it seemed at >first glance. <----- snip ----->
Last information I have found about status of ASM and HyperPAV was in the thread "PLPA and COMMON PAGESPACE Size", mail from 13 Sep 2007: <----- snip -----> > But note that ASM only supports starting 2 I/Os to a page data set > when there are dynamic PAVs (as opposed to static PAVs). Unless > something has changed since z/OS 1.3 when this support was introduced. That is a very good point, especially because I was going to say "Dynamic PAV and of course HyperPAV", but thought I had better check the code first. It turns out that the ASM code which decides whether to create two sets of I/O control blocks for a page data set was not updated for HyperPAV (an oversight on our part). It checks to see if you specified WLM PAV when you defined the device in HCD. Since WLM doesn't need to manage PAVs when the control unit has been told to use HyperPAV mode, our intention was that the specification of WLM PAV in HCD would be irrelevant for HyperPAV devices. But for now, it looks like you would want to specify WLM PAV for HyperPAV paging devices, until we can update the ASM code (probably via an APAR) to automatically recognize HyperPAV paging devices. <----- snip -----> Zaromil ---------------------------------------------------------------------- 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
