Given that I¹m just about to try this, I¹d like to hear the funny story just
as soon as you¹re ready to relate it.
--
.~.Robert P. Nix Mayo Foundation
/V\RO-OE-5-55 200 First Street SW
/ ( ) \ 507-284-0844 Rochester, MN 55905
^^-^^ -
I have 5301
--
.~.Robert P. Nix Mayo Foundation
/V\RO-OE-5-55200 First Street SW
/( )\ 507-284-0844 Rochester, MN 55905
^^-^^ -
In theory, theory and practice are the same, but
in practice, theory and practice
(Slightly off topic)
I had a user request additional memory because their image was slow, and I
tried to explain that it wouldn't help because they weren't CPU bound. They
insisted and I finally gave in and increased the image from 2gig to 3gig.
I waited a couple of days, and then asked if it
Find searches for things at the beginning of a record. Locate searches
anywhere within a record. Given that you didn¹t do a split to put the
userids on separate lines, I¹d think that ³pipe cms q n | locate /ESA/ |
console² would work better.
Also, q n is a CP command, not a CMS, so actually,
And on that note: When is IBM going to come up with a non-tape version of
SPXTAPE? One of my two systems has no tape drives anymore, and since we only
use them during the install, I can see a time when the other system loses
its access to tape as well (I have a z/VM 5.3 DVD sitting on my desk).
Actually, I was very glad to get the page and spool out of the res pack.
It¹s such a pain to remove later on. The way we install, the starter system
³becomes² the production system after the customization. So the concept of a
starter system disappears.
If you¹re installing your starter system
Will a cp command correctly copy links and other special files?
What I¹ve always seen suggested would be to mount the filesystems as you
described, and then use a command such as:
cd /mnt/3390-3 ; tar cf - * | (cd ../3390-9 tar xf -)
tar will copy everything correctly. This will work
We¹ve been running PerfKit for quite some time, and the 191 disk got too
small for the trend data. I brought down both PERFSVM virtual machines (we
run two CECs and CSE), created new, larger disks, and set up the PERFSVM¹s
to use them, and then xautolog¹ed both PERFSVMs.
Now, both are getting the
itself will not be listed, but
if you issue MONITOR STOP, CP will stop using it too.
2007/7/11, RPN01 [EMAIL PROTECTED]:
We've been running PerfKit for quite some time, and the 191 disk got too
small for the trend data. I brought down both PERFSVM virtual machines (we
run two CECs and CSE
Unless the Linux admin takes specific action to label the disk with the
current label, it will be overwritten with the Linux volume name, and will
then not be available the next time the userid is logged out and back in.
Two courses of action: Either change the minidisk to exclude cylinder zero,
In the second level system, do a Q DASD ALL and look at the results; Do
you see your disks there, and what is their state?
In the first level system, do a Q DASD label for the two disks in
question, and see what their state is there.
There has to be a reason that the disk isn't finding its way
I never thought about this, but it's a great idea. There's no telling what
you'll learn when you watch this group... :-)
--
.~.Robert P. Nix Mayo Foundation
/V\RO-OE-5-55200 First Street SW
/( )\ 507-284-0844 Rochester, MN 55905
^^-^^
I like the clean system idea... On the hardware encryption: This would
assume that your disaster recovery facility had those same encrypting tape
drives in place, wouldn't it? It would limit your possible recovery sites
considerably. And even if your DR vendor installed the same drives on one of
There is one situation where the virus scan is potentially useful: The
mainframe Linux doesn¹t exist in a vacuum. It talks to and from other
computers, which may be subject to a virus being passed on in files handled
by Linux. It isn¹t good PR for someone to be walking around saying ³Yeah, I
Actually, your zVM doesn¹t even need FTPSERVE up; you only need the FTP
client, which doesn¹t depend on the server at all to go out to another box.
So the list is reduced to z/VM, TCP/IP and the FTP program, plus CMSDDR, if
that is needed to deconstruct the package.
Setting up an FTP / NFS
Check the 191 disk for userid and see if it's full. If so,
archive (or just delete) the data files there and autolog it to clean out
the backlog of accounting data.
--
.~.Robert P. Nix Mayo Foundation
/V\RO-OE-5-55200 First Street SW
/( )\
Are you dealing with a VIRTUAL CTC, or a real device?
If virtual, then your SEND CP TCPIP DET 2030 took it completely out of
existence. You'd need to redefine it and couple it to the other virtual
machine:
SEND CP TCPIP DEFINE CTC 2030 USER VMTEST
Then:
SEND CP TCPIP COUPLE 2030 TO VMTEST 2000
I'd infer that, based on all the conversation so far, you have a real CTCA
at 2030. For this exercise, just totally forget about the real device; it
has no bearing on the outcome.
What you're trying to do is connect your first-level TCPIP machine with your
first-level zVM guest machine. Both
Our site is considering GDPS, and we've come to the conclusion that z/VM
must be a non-participant. Our current plan is to move z/VM on to its own
logical controller, outside the grasp of GDPS. I think that the last straw
was that we could participate in GDPS only if our two z/VM systems did not
, but
in practice, theory and practice are different.
On 4/10/07 3:36 PM, Rob van der Heij [EMAIL PROTECTED] wrote:
On 4/10/07, RPN01 [EMAIL PROTECTED] wrote:
Necessity is the Mother of all...
He said, while cleaning his ears with the barrel of a gun :-)
There's always another way to do
Actually, there's nothing that says the directory has to be on the res pack
as well. You could move it to another CP Owned volume. While not ideal, ours
live at the beginning of our page packs, and the two systems share the res
pack.
--
.~.Robert P. Nix Mayo Foundation
/V\
From: The IBM z/VM Operating System on behalf of RPN01
Sent: Tue 4/10/2007 2:51 PM
To: IBMVM@LISTSERV.UARK.EDU
Subject: Re: [IBMVM] Time to move
Actually, there's nothing that says the directory has to be on the res pack
as well. You could move it to another CP Owned volume. While
I¹m getting the same messages though, even after doubling the size of the
MONDCSS segment. It would be nice if, along with the message that it is too
small, it would indicate just how big it wants it to be...
--
.~.Robert P. Nix Mayo Foundation
/V\RO-OC-1-13
Since our environment is fairly dynamic, what I've done is to let SYSTEM
CONFIG vary on all the devices that are visible to our LPAR, and then parse
through a Q DASD ALL in AUTOLOG1, looking for CP OWNED and CP SYSTEM. Any
other volume I find, I vary offline, keeping my zOS peers happy and off my
Did you format the entire paging space? It needs to be pre-defined with 4k
page blocks before it can be used.
CMS minidisk volumes can be defined using ICKDSF by just formatting cyl 0 to
contain the correct label, and everything will handle formatting their small
piece of the world when the time
Actually, it isn't necessarily IBM's call for a lot of the packages listed
there. Many of these packages were written by customers using zVM, and they
have the right to decide how to distribute their intellectual property.
Generally, the submitter of the package is listed along with each package,
I haven¹t done it, but the zAPP processor would only be usable by the zOS
LPAR; zVM has no use for zAAPs.
--
.~.Robert P. Nix Mayo Foundation
/V\RO-OC-1-13200 First Street SW
/( )\ 507-284-0844 Rochester, MN 55905
^^-^^ -
We¹ve been running the Performance Toolkit since October, and the sysname
FCXTREND file is getting a bit large. I wrote a pipeline to break this file
down into monthly segments, and thought I had a way of using them via the
FCXSUM command, documented in the manual... But once I got to that point,
The settings in SYSTEM CONFIG are only used during an IPL. To change the
timezone without the IPL, you must use the SET command.
--
.~.Robert P. Nix Mayo Foundation
/V\RO-OC-1-13 200 First Street SW
/ ( ) \ 507-284-0844 Rochester, MN 55905
not leaving
the user where (s)he might expect to be.
On 3/8/07 7:38 PM, Alan Altmark [EMAIL PROTECTED] wrote:
On Thursday, 03/08/2007 at 11:25 CST, Rich Smrcina [EMAIL PROTECTED]
wrote:
That should do it... :)
RPN01 wrote:
How ?bout just a plain old ?exit??
Perhaps you guys are really just
How bout just a plain old ³exit²?
--
.~.Robert P. Nix Mayo Foundation
/V\RO-OC-1-13 200 First Street SW
/ ( ) \ 507-284-0844 Rochester, MN 55905
^^-^^ -
In theory, theory and practice are the same, but
in practice, theory
Actually, the GCS build isn¹t quite that trivial.
SPOOL PUN * CLASS I
SPOOL RDR CLASS I
VMFSETUP ZVM GCS
VMFBLD PPF ZVM GCS GCTLOAD (ALL
I C
Your VMFBLD command was also missing the PPF parameter.
--
.~.Robert P. Nix Mayo Foundation
/V\RO-OC-1-13 200 First
There may not be a physical 3390 Mod 27. As newer DASD hardware has been
created, and with the need for larger and larger disks, the new hardware
emulates the 3390 in larger and larger sizes. We regularly use 3390 mod 27
DASD, with 32,760 cylinders on each device. These are emulated on portions
of
Glad to know that other¹s memorys work no better than mine.
We should all switch to emulated 2311¹s and be done with it.
--
.~.Robert P. Nix Mayo Foundation
/V\RO-OC-1-13 200 First Street SW
/ ( ) \ 507-284-0844 Rochester, MN 55905
^^-^^ -
There is an exit you can write called fcxrenam exec (if memory serves). In
this script, you can do more than the standard three file roll, and do any
cleanup needed to keep the disk from filling. You could send the files to a
service machine to be archived, or just move them to a large disk to
Having just added two more grants to my SYSTEM CONFIG file, I¹m wondering if
there is any advantage to having them all here as opposed to doing them
dynamically in AUTOLOG1¹s profile.
The advantage to doing them in AUTOLOG1 is that I already have a list of
images there that I want to bring up,
I tried ³Q IPLPARMS² on my system, and get:
q iplparms
No IPL parameters are currently defined
Ready; T=0.01/0.01 08:56:31
How does this tell me what system I¹m on?
However, using ³IDENTIFY², I get:
identify
TS00086 AT POLARVIA RSCS 01/25/07 08:57:50 CST THURSDAY
Ready;
I am a class A user; What version of zVM are you running, because your Q
IPLPARMS looks nothing like mine...
--
.~.Robert P. Nix Mayo Foundation
/V\RO-OC-1-13 200 First Street SW
/ ( ) \ 507-284-0844 Rochester, MN 55905
^^-^^ -
In
The SYSTEM CONFIG file can have several CPUIDs specified, and can then be
sectioned off based on the system name associated with the CPUID.
Specifically, you get to specify the CP Owned volumes, so you have two
separate lists; one for SYSA and another for SYSB. In our case, we move the
CP
Q IPLPARMS only works if you have separate config files to specify (all our
systems run from one), and IDENTIFY only works if you've actually set up
SYSTEM NETID correctly. It would appear that the Q GATEWAY is the only
method that would work consistently.
Since we do this in several execs,
Hi Edward; Welcome to the game
The cost of Linux may all be related to what you plan to do with it. For a
proof of concept, and if you¹re familiar with Linux already, you could try
the Debian distribution, which has no cost. There are one or two others as
well which should be free
machine from one physical server to another
be
too much like Waterloo's Single-System-Image. That is so 'old mainframe'.
(Sorry, Phil, Romney, but I really hate it when our PC people think that
virtualization is NEW)
/Tom Kern
--- RPN01 [EMAIL PROTECTED] wrote:
We also run two z9 CECs
We also run two z9 CECs with a zVM LPAR in each, 2 IFLs each and 16gig of
storage each. Both run a vSwitch and share a subnet (key to being able to
move images from one LPAR to the other). Both have access to all of the
images¹ DASD. We run a load on both, but when one of the zVM¹s or CECs must
Just slightly ³outside the box²... Could you use the eight digit account
field as a numeric index into a publicly accessable file and keep the actual
parameter lines needed in that file? The file lines could be as long as
necessary, and with eight digits to index with, you could have quite a lot
201 - 244 of 244 matches
Mail list logo
