ISTR that DIAG 0 returns multiple levels of information for each layer until
you get back to the first level system. If so, won't that layer have the
info you want?
On Sun, May 30, 2010 at 1:56 PM, Michael MacIsaac <mike...@us.ibm.com>wrote:
>
> Thanks Kris for a quick answer, and also thanks to another for an off-line
> reply. I do appreciate it.
>
> However, I'm still not sure I have a solution. Let me describe the problem
> a bit better. I'm trying to ascertain the location of resources in the
> System z hierarchy - that is CECs have LPARs, and LPARs have systems which
> may be Linux or z/VM, and z/VM systems have systems, which may also be Linux
> or z/VM (I know there are a lot of other possibilities, but these are the
> most interesting). I'd like to do as much of this as possible on Linux
> rather than from z/VM, thus the request for CP commands, not CMS (so I'm
> even more hesitant to use CPHOST which is a z/VM add-on).
>
> I can get a lot of information from /proc/sysinfo. For example, on a third
> level Linux, the following data are interesting:
>
> # grep "LPAR Name:" /proc/sysinfo | awk '{print $3}'
> LVM1
> # grep "VM01 Name:" /proc/sysinfo | awk '{print $3}'
> VM140
> # grep "VM00 Name:" /proc/sysinfo | awk '{print $3}'
> ZVMMAPLX
>
> I know the system is running on the LPAR named "LVM1". Because there is a
> VM01 entry, I can assume this is a second level z/VM running on the user ID
> named "VM140". Also I know the third level Linux is running on the user ID
> "ZVMMAPLX". I can also determine the CEC's machine type, model number and
> serial number with other data in /proc/sysinfo.
>
> So that tells almost the entire hierarchy *except* for the system
> identifier of the first level z/VM system:
> CECs
> |
> +-mach-type-X-model-num-Y-serial-num-Z
> |
> +-LPARs
> |
> +-LVM1
> |
> +first-level-zVMs
> |
> +-?????????
> |
> +-second-level-z/VMs
> |
> +-VM140
> |
> +-third-level-Linuxes
> |
> +-ZVMMAPLX
>
>
> Neither /proc/sysinfo, DIAG 0, nor the STSIUSE EXEC seem to supply that.
> Any other ideas other than CPHOST? (I could require this value as a
> parameter, but it would be more convenient and reliable if it could be
> ascertained.) Thanks in advance.
>
> "Mike MacIsaac" <mike...@us.ibm.com> (845) 433-7061
>
> P.S. Typing out the hierarchy has shown to me that I may be mixing user IDs
> with system identifiers - perhaps I *do* have the entire hierarchy in that I
> know there is one z/VM running on the LPAR LVM1. However, we usually
> remember z/VM systems by system identifier, not by LPAR name.
>
--
zMan -- "I've got a mainframe and I'm not afraid to use it"
