z/VM is adept at finding its way through the thicket to discover reachable devices. However, 'reachable' is determined by the hardware, and hardware is driven by the IOCDS in control whether activated dynamically or PORed. We killed a z/VM LPAR a while back by dynamically activating (from z/OS) an IODF that hosed all chpid connections to the z/VM DASD. z/VM IPL failed repeatedly (don't remember messages) until we corrected the IODF and reactivated it. Even z/VM cannot overcome the you-can't-there-from-here syndrome.
. . J.O.Skip Robinson Southern California Edison Company Electric Dragon Team Paddler SHARE MVS Program Co-Manager 626-302-7535 Office 323-715-0595 Mobile [email protected] From: "Norman.Hollander" <[email protected]> To: [email protected], Date: 05/07/2014 03:46 PM Subject: Re: z10 IPL from Utility Tape Sent by: IBM Mainframe Discussion List <[email protected]> z/VM discovers the I/O by itself. No need for RIO gen, or even specifying in the PARM disk. If z/OS is just running under z/VM, it just uses the software IOCDS. As long as the device is Defined, no need for an actual Control Unit or channel. Although I do go through the exercise in case I want to run z/OS natively. -----Original Message----- From: IBM Mainframe Discussion List [mailto:[email protected]] On Behalf Of Skip Robinson Sent: Wednesday, May 7, 2014 3:23 PM To: [email protected] Subject: Re: z10 IPL from Utility Tape Until Bob Shannon recounted his experience with failed DASD (!), I would have looked for a more likely explanation. Here's my first take. IODF gets dynamically activated over a period of time. Depending on how rigorous you are, the IOCDS in the SE may get updated less often. It's possible for the actual system configuration to stray quite far from the IOCDS that was last written out. After a power failure, you necessarily do a POR that reads the IOCDS. If the last saved IOCDS cannot map the actual physical configuration, very strange results can occur during IPL. I have little experience with z/VM, so my guess might be way off base. . . J.O.Skip Robinson Southern California Edison Company Electric Dragon Team Paddler SHARE MVS Program Co-Manager 626-302-7535 Office 323-715-0595 Mobile [email protected] From: Karl Severson <[email protected]> To: [email protected], Date: 05/07/2014 03:07 PM Subject: Re: z10 IPL from Utility Tape Sent by: IBM Mainframe Discussion List <[email protected]> >OOPS, my bad. I should have realized from the VOLSERs that it was z/VM. BTW - I, personally, _never_ IPL via an activation profile. I just use the "LOAD" function on the "Recovery" page. I drag the LPAR icon onto the LOAD icon. This gives me a "pop up" in which I can put the IPL volume and LOAD parameters (it remembers what I put in last time, however, and initializes the popup with those values). I don't know if trying the above would give you a better message than using the activation profile. But, hey, you're down, so it can't really hurt to try, right? < No worries. We have the HMC set up tree style but next time I'm out there I might change it to classic and try it the way you do it. It seems now that the primary SE is saying that there are "one or more objects in an unexceptable state" so I'll have to find out what they are. Could be the root of all our problems right there. Karl ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
