Hi Greg,
Yep, that was it - and I think I understand the syntax a little better now
that we have a working set of parms (I was also confusing the values of CSS
and RCSS).
So now I have all of the CHPIDS online to my TEST LPAR:
q chpids
0 1 2 3 4 5 6 7 8 9 A B C D E F
0x + + + + + + + + . . . . . . . .
1x . . . . + + + + . . . . . . . .
2x + + + + + + + + + + + + + + + +
3x . . . . . . . . . . . . . . . .
4x . . . . . . . . . . . . . . . .
5x . . . . . . . . . . . . . . . .
6x . . . . . . . . . . . . . . . .
7x . . . . . . . . . . . . . . . .
8x + + + + + + + + + + + + + + + +
9x + + + + + + + + + + + + + + + +
Ax . . . . . . . . . . . . . . . .
Bx . . . . . . . . . . . . . . . .
Cx . . . . . . . . . . . . . . . .
Dx . . . . . . . . . . . . . . . .
Ex . . . . . . . . . . . . . . . .
Fx . . . . . . . . . . . . . . . .
I modified the control units, eg:
cp modify cu 4000 css 1 add_css_access chpid 80 84 88 8C 81 85 89 8D cuadd 0
And added all of the IODevs, eg:
cp def dev 4000-4080 unitadd 00 cu 4000 css 1 dasd par TEST
All of the DASD and control units came online normally. :)
Now, my last (hopefully!) question is this. Are these modifications stored
in the HSA, and/or will they survive an IPL of z/VM in the TEST LPAR? I
realize we'll still need a POR to make these permanent (absent using HCD,
which we haven't deployed yet) - but for the time being are these changes
"permanent" unless/until a POR is performed, or will they disappear when the
z/VM TEST LPAR is IPL'd?
Thanks for your help Greg, MUCH appreciated. :)
-Mike
-----Original Message-----
From: The IBM z/VM Operating System [mailto:[email protected]] On
Behalf Of Gregg
Sent: Thursday, March 18, 2010 1:27 PM
To: [email protected]
Subject: Re: Adding a second CSS
Hi Mike, yeppurr it looks like I had my ones and zeros backwards
CP MODify CHPID 80 CSS 1 ADD_CSS_access RCSS 0 INITial_access TEST
ACCess_list TEST
On Thu, Mar 18, 2010 at 1:15 PM, Michael Coffin <[email protected]>
wrote:
> Hi Greg,
>
> TEST is in CSS 1. The DASD on CHPID 80 that I'm trying to give TEST
access
> to is in CSS 0. Here are snips from our IOCP:
>
> RESOURCE PARTITION=((CSS(0),(sys1,2),(sys2,1),(sys3*
> T,4),(sys4,3),(*,5),(*,6),(*,7),(*,8),(*,9),(*,A),(*,*
> B),(*,C),(*,D),(*,E),(*,F)),(CSS(1),(TEST,1),(*,2),(*,3)*
> ,(*,4),(*,5),(*,6),(*,7),(*,8),(*,9),(*,A),(*,B),(*,C),(*
> *,D),(*,E),(*,F)))
> CHPID PATH=(CSS(0),80),SHARED, *
> PARTITION=((sys1,sys2,sys3,sys4),(=)), *
> PCHID=290,TYPE=FC
>
> CNTLUNIT CUNUMBR=4000,PATH=((CSS(0),80,84,88,8C,81,85,89,8D)),*
> UNITADD=((00,256)),CUADD=0,UNIT=2107
> IODEVICE ADDRESS=(4000,128),CUNUMBR=(4000),STADET=Y,UNIT=3390
>
> Thanks for the info about "adding" access vs. "replacing", I was pretty
> concerned about that. :)
>
> -Mike
>
>
> -----Original Message-----
> From: The IBM z/VM Operating System [mailto:[email protected]] On
> Behalf Of Gregg
> Sent: Thursday, March 18, 2010 12:16 PM
> To: [email protected]
> Subject: Re: Adding a second CSS
>
> Hi Mike,
> what CSS is Test in? Maybe I got my 1's and 0's backwards... would
> you paste, from the IOCP the RESOURCE PARTITION statement, CHPID 80
> and the CNTLUNIT statement for CU 4000?
> My read earlier was that CHPID 80 is currently defined in CSS 1 and I
> assumed TEST was in CSS 0.
> The modify adds the lpnames to the ACC/INIT lists.
> We've 'spanned' channels channels here, where we first had to DEFine
> in one CSS and then MODify for the second.
>
>
> On Thu, Mar 18, 2010 at 7:31 AM, Michael Coffin <[email protected]>
> wrote:
>> Hi Greg,
>>
>> Ooops, yes - CAPS is a typo in this email (not in the actual command),
>> should be TEST. :)
>>
>> I tried the CP MODIFY with the parms you suggested, but I'm still getting
>> HCP0864E:
>>
>> CP MODify CHPID 80 CSS 0 ADD_CSS_access RCSS 1 INITial_access TEST
>> ACCess_list TEST
>> HCPCCO864E Partition TEST does not have access to CSS 0
>>
>>
>> Perhaps, does this command need to be executed on a system that already
> has
>> access to CSS 0? Are we in a "chicken and egg" scenario where I'm
>> requesting access to the CSS, but I'll never get it because I don't
> already
>> have access to the CSS?
>>
>> PS: Do you know if the ACCESS_LIST needs to be the COMPLETE list of
> systems
>> to have access to the CSS, or only the system you are adding via the CP
>> MODIFY command? In other words, will the system(s) specified in
> ACCESS_LIST
>> be added to the existing access list, or REPLACE the existing access
list?
>>
>> -Mike
>>
>> -----Original Message-----
>> From: The IBM z/VM Operating System [mailto:[email protected]] On
>> Behalf Of Gregg
>> Sent: Thursday, March 18, 2010 5:50 AM
>> To: [email protected]
>> Subject: Re: Adding a second CSS
>>
>> Mike, is "init CAPS" a typo/some email malfunction?
>>
>> The CSS value is the target CSS and the RCSS value "is the reference
>> channel-subsystem image from which the channel-path characteristics
>> are to be copied"
>> INITial_access is optional, but ACCess_list is required and as you
>> know the INIT and ACC lpname(s list) do not need to be the same in the
>> INIT and ACC lists
>> I believe the modify chpid should be
>> CP MODify CHPID 80 CSS 0 ADD_CSS_access RCSS 1 INITial_access TEST
>> ACCess_list TEST
>>
>>> The spindle Im trying to reach is 4030 on CU 4000 on CHPID 80. So
> first
>> I
>>> try to define the CHPID for our LPAR (TEST):
>>>
>>> CP DEFINE CHPID 80 PCHID 290 CSS 0 TYPE FC SHARED ACC othersys1
othersys2
>>> othersys3 TEST INIT TEST
>>>
>>> But I get:
>>>
>>> HCPCCO864E Partition TEST does not have access to CSS 0
>>>
>>> So I try to authorize our partition:
>>>
>>> cp modify chpid 80 add access_list TEST
>>> HCPCCO864E Partition TEST does not have access to CSS 0
>>>
>>> Hmmmm, well the chpid exists (even though it shows as undefined in QUERY
>>> CHPIDS), so I try to modify the chpid to add the CSS:
>>>
>>> cp modify chpid 80 css 0 add_css rcss 0 init CAPS
>>> HCPZPM6704E Missing token at end of line
>>
>> --
>> Gregg Reed
>> "No Plan, survives execution"
>>
>
>
>
> --
> Gregg Reed
> "No Plan, survives execution"
>
--
Gregg Reed
"No Plan, survives execution"
