The way I think about this is that there are two separate views of one data source.
The "hardware view" is extracted from the IODF data set and loaded into the machine when you load the IOCDS into the hardware. You use the HCD dialog to load the appropriate "hardware view" into one of the IOCDS "slots" in the desired machine. The IOCDS is then read by the machine when you perform the Power On Reset function. When you first install a machine, you must perform the Power On Reset function to load the first IOCDS. After that initial Power On Reset, it is possible to perform many types of changes dynamically via the MVS ACTIVATE command. The "software view" is extracted from the IODF data set and loaded into memory when you IPL and your LOADxx specifies an eight character Operating System Configuration on the IODF statement. After IPL, it is possible to perform many types of changes dynamically via the MVS ACTIVATE command. At IPL time, if the "hardware view" and the "software view" are sufficiently similar that MVS is able to find its necessary devices, then the IPL will be successful. If the two views are identical - that is, the two views were extracted from the same data source (the same IODF), then the operating system will permit the MVS ACTIVATE command to make dynamic changes. Interestingly (at least to me) there really is no MVS or hardware arbitrator function which guarantees or requires that, at IPL time, the "hardware view" and the "software view" are identical. In many ways this seems to me to be a "typical IBM solution" - they took the old MVSCP and IOCP process, and simply tacked on a shiny new interface (HCD / HCM). Yes, you get cool new functions (dynamic ACTIVATE) when you use the new tooling as IBM intended, but there is NO requirement that the IOCDS and the Operating System configuration come from the same IODF data set. They just have to match sufficiently to IPL. Now, I think it is an EXCELLENT IDEA to ALWAYS make sure your IOCDS and Operating System configuration come from the same IODF, but it is not required to IPL an MVS operating system. One more thing.... As far as I know, the SNA name is ONLY used when you use the HCD function to load an IOCDS into one of the IOCDS data set "slots" in the machine. Brian On Wed, 21 Dec 2005 23:20:56 +0100, R.S. <[EMAIL PROTECTED]> wrote: >Scenario: >IODF file contains definition of all (actually three) CPCs and all the >DASD & other I/O equipment. >When IOCDS is generated, one has to specify CPC for which the IOCDS is >generated. However IODF can be the same (*) on every LPAR on every CPC. >Let's make another assumption, the machines are same models, i.e. >2064-1C10. >Question: How system recognize relevant CPC definition ? By "CPC name" >field in SNA address ? >What happens when that field remain empty ? > >(*) AFAIK one can create subsets of IODF using option 6.7 >-- >Radoslaw Skorupka >Lodz, Poland ---------------------------------------------------------------------- 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
