Re: IOCP for MP2003-204

[Roland P. Chung]

Hello All, thanks to all who responded. After hours of
struggling, I manage to set up a IOCDS which I was able to
get at the tape drives, DASDs and 3174. IPLed the
standalone ICKDSF and DDR. I was very happy to be able to
restore all 5 CP OWNED volumes. 

Came to the monent of truth IPL'ed VM hit the
terminal... nothing...

After spent hours investigating and trial many different
address on the 3174, I only can figure out that VM can't
get at the 3174 because in the SYSTEM CONFIG file, it says:

Operator_Consoles  001F 0020 0021 0022 0023
 Emergency_Message_Consoles   0020 0021 0022 0023

And in MP2003 world, I think there are no channel 0 anymore
and there re no good old 01F(the integrated console) hangs
on that channel either. 

I think the address (at least the UU) has been hardcoded as
40-5F on the 3174 controler.

I was very tire my brian can only think of a few
solutions:

1) find an IBM "Productivity Centre" near Washinging, DC
(if there are such thing in US!), restore the 5 pack VM
system, IPL it and fix the SYSTEM CONFIG to include 040 -
043 (the address on that 3174); Backup the 230RES, restore
it back here and I should be in business

2) reconfigure the 3174 to start the UU from 20 - well, the
Utility and a backup disk are there only thing I need
is a manual. Anyone has a LINK I can get a softcopy? 

3) get a zVM DVD? How and where?

Anyone has a better idea on how to resolve this problem?
TIA

Regards,

...Roland
--- Alan Altmark <[EMAIL PROTECTED]> wrote:

> On Friday, 06/16/2006 at 09:46 AST, George Haddad
> <[EMAIL PROTECTED]> wrote:
> > Our 2003-205 can be IMLed to do Dynamic I/O. We never
> did implement that
> > feature --- we just never made many I/O changes over
> the yrs -- so I
> > don't know if that negates the need for an inital IOCDS
> load or not.
> 
> No, it doesn't.  To IPL CP you have to IPL *something*. 
> That *something* 
> must be in the active I/O configuration.  You need a tape
> drive and enough 
> dasd volumes to restore a basic VM system, plus a
> console.
> 
> This is the nice thing about having an emergency 1-pack
> micro-system. 
> Sysres, paging, spooling, a directory, OPERATOR and
> MAINT, DDR, IOCP, all 
> on one volume, with the console pointing to the
> integrated console (on 
> modern machines, integrated 3270).  Or, also on modern
> machines, the 
> ability to IPL the z/VM DVD.
> 
> IF you authorized the LPAR to perform dynamic I/O
> operations and you have 
> dynamic I/O enabled in VM, then CP, once up, can start to
> add chpids, 
> control units, and devices to the I/O config.
> 
> Alan Altmark
> z/VM Development
> IBM Endicott
> 

Re: ICC problem on z890

[Shimon Lebowitz]

That's exactly what you would hear from me.
Well. maybe not *that* bad, but... 

Ever since we installed the z890 we have had console
outages that have included ALL of the PCs on the private
ICC network, and that have only been corrected by 
varying off/on the CHPID at the HMC. (We have not
seen a situation where only one or more LUs were
disabled, but others worked. All or nothing here).

We have opened multiple hardware incidents with
IBM here in Israel, and our hardware support technician
has tried all sorts of changes (the OSA card, the
cableing, the network switch) but every month or so 
it happens again. "Get IBM to fix them" is easier 
said than done.

Luckily, our ops do have one 3270 emulator on an 
old  3174, which has been very helpful!

Interestingly, the network has *two* OSA cards,
but only the VM/VSE one with 4 terminals has 
caused problems. The other system, running 1 z/OS 
console, has never gotten stuck.

Shimon

On 16 Jun 2006 at 15:01, Alan Altmark wrote:

> On Friday, 06/16/2006 at 02:52 AST, "Hooker, Don - OIT" 
> <[EMAIL PROTECTED]> wrote:
> > Thomas, though I have not personally seen it, I?ve been told we 
> experienced the 
> > same problem here with our 
> > 
> > VSE consoles.  We no longer use the ICC sessions for them.  Apparently 
> the VM 
> > consoles did not suffer.
> > 
> > I believe several people at the last WAVV mentioned having the same 
> problem.  I 
> > think they said that if they varied the
> > 
> > CHIPD  offline and then online again that the consoles could recover.
> 
> Hopefully each installation that experiences this problem opens up a 
> hardware PMR.  If they don't work, get IBM to fix them.  I don't want to 
> hear, 10 years from now, how "I tried the ICC 10 years ago and it didn't 
> work for me - I'll take an old-fashioned 3277 any day!"  :-)
> 
> Alan Altmark
> z/VM Development
> IBM Endicott

Re: ICC problem on z890

[Rick Troth]

On Fri, 16 Jun 2006, Alan Altmark wrote:
>I don't want to
> hear, 10 years from now, how "I tried the ICC 10 years ago and it didn't
> work for me - I'll take an old-fashioned 3277 any day!"  :-)

I do wish that DISCONN without HOLD would do via ICC what it
does for TELNET and ye olde dial-up and actually drop the session.

Now where'd I leave that 7171??

-- R;