We started this way, but I felt that this made it harder to isolate, list, or find the LPAR specific minidisks. With them in the same userid, you have the entire list at hand.
On 7/28/08 2:04 PM, "Kris Buelens" <[EMAIL PROTECTED]> wrote: > Now that you wrote this, I remember my setup which is easier in m eyes > (we only had 2 systems sharing the directory). For example > USER VMUTIL > MDISK 1191 .... RR > MDISK 2191 .... RR > SYSAFFIN system1 > LINK * 1191 191 M > SYSAFFIN system2 > LINK * 2191 191 M > Nowadays, one would probably use the RRD on the MDISK records so that > there is even no RR link to the system that doesn't need it. > > I find this easier as the directory entry of the user completely tells > what is what, otherwise having the remember that 03 stands for RSCS > etc is something I couldn't, not even when 15 years younger. > > 2008/7/28 RPN01 <[EMAIL PROTECTED]>: >> In principle, what you've done is correct... But very wrong. :-) >> >> We NEVER put an mdisk statement under a SYSAFFIN statement. All the >> minidisks that need to be "owned" by a specific LPAR are defined in the >> userid DISKOWNR, and the only thing under the SYSAFFIN statements are LINK >> statements back to DISKOWNR. We use a four digit address for the minidisks. >> The first digit is the LPAR that owns the disk (In our case, 1 for POLAR and >> 2 for GRIZZLY). The next two digits define which userid owns the disk, say >> 01 for OPERATOR, 02 for TCPIP, 03 for RSCS, ... Down the list. The last >> digit is the minidisk within the user, such as 0 for the 191, 1 for the 195, >> ... >> >> Defining things this way allows all the minidisks to be seen from either >> system, if necessary, correctly maps the use of the space on all systems, >> allows access to all the disks from either system (hopefully only by >> read-only links), identifies the ownership of the minidisks, and allows >> SYSAFFIN to isolate the main purpose of the minidisks to specific LPARs. >> >> -- >> Robert P. Nix Mayo Foundation .~. >> RO-OE-5-55 200 First Street SW /V\ >> 507-284-0844 Rochester, MN 55905 /( )\ >> ----- ^^-^^ >> "In theory, theory and practice are the same, but >> in practice, theory and practice are different." >> >> >> >> >> On 7/28/08 8:08 AM, "Florian Bilek" <[EMAIL PROTECTED]> wrote: >> >>> Dear all, >>> >>> I am working now, thanks to Kris, with SysAffin in the Directory. In >>> principle this works fine but I have now an issue with the DISKMAP utilit >>> y. >>> The report run on one system shows space of the disk as free where a >>> SysAffin is coded for another system. >>> >>> IMHO this will lead to confusion when somebody who is not aware of the >>> SysAffin is editing the directory. He can by accident assign such space t >>> o a >>> minidisk in the meaning that this is an empty space. >>> >>> Is there a way to mark such areas as alloced? >>> >>> Best regards, >>> Florian >> > >
