On Tue, Mar 11, 2008 at 1:59 AM, Alan Altmark [EMAIL PROTECTED] wrote:
While trying to reconstruct a source directory from an object directory
might seem logical, far too much information is lost in translation.
Obviously it would be really easy to write a wrapper around DIRECTXA
that also
On Thu, 6 Mar 2008 08:32:32 -0500, Alan Altmark [EMAIL PROTECTED]
wrote:
The information you are looking for is in a type 04 accounting record.
Alan Altmark
z/VM Development
IBM Endicott
=
But I understand that type 04
Thanks for the responses to my DR question.Helpful information!
So basically, I have 2 methods of bringing z/VM up at our DR site:
1) Run it under the z/VM Floor System (we use Sungard as our DR service).
2) Bring z/VM up in an LPAR.
To me, option one is probably the BEST and EASIEST to
Hi,
I was wondering if someone could provide clarification on the set up to use
SNMP to manage virtual switches with z/VM 5.3.0.
In the z/VM TCP/IP Planning and Customization book (chapters 21 23), it
mentions the requirement to define a manager for the virtual switch by adding
the VSWITCH
1) I need to change my VM directory so that I have DEDICATE VOLID
yy rather than DEDICATE (which is what we're doing
now).
One possible preparation: If you consistently do not use the last
cylinder of every volume, you could restore your disks into minidisks on
the VM
On Mar 11, 2008, at 6:59 AM, Karl Kingston wrote:
Thanks for the responses to my DR question.Helpful information!
So basically, I have 2 methods of bringing z/VM up at our DR site:
1) Run it under the z/VM Floor System (we use Sungard as our DR
service).
2) Bring z/VM up in an
I missed this before. If you are going to have *AN LPAR* at the D/R
site and you have, it sounds like, a VM lpar and z/OS lpars to restore
for you D/R test, you really don't have any choice but to run your D/R
test under a VM system. That way the VM and z/OS systems you restore as
virtual
In addition to all the good advice already, I will add: Do not try to
run your VM system 2nd level if you have Linux guests. Performance of
the Linux guests will be terrible. There are only two levels of
hardware virtualization. That leaves your Linux guests being
virtualized in software
Jim Bohnsack wrote:
I missed this before. If you are going to have *AN LPAR* at the D/R
site and you have, it sounds like, a VM lpar and z/OS lpars to restore
for you D/R test, you really don't have any choice but to run your D/R
test under a VM system. That way the VM and z/OS systems you
VM is volume id based. (Many years ago, it was not). That being said,
VM, when it is IPL'ed, will go and look at all of the dasd that it can
touch and find out, among other things, what the volid is. Only if it
can't find a volid will inform you. It's up to you how to handle that
situation.
CP directory changes only influence the future, it defines the initial
configuration of virtual machine logging on, and what can be LINKed
afterwards. By no means will a directory change affect the config of
a logged on virtual machine. This seems a common misunderstanding.
E.g. when an MDISK is
It would be nice if there were some sample code to demonstrate connecting
to
this new system service (*VMEVENT) and some sample records so that we cou
ld
get a feel for the kind of data and the level of effort to begin
implementing this.
/Tom Kern
On Tue, 11 Mar 2008 10:57:26 -0400, Alan
While these may not help you if you format your directory source disk, you
could modify it to do some copies to other disks instead of or in addition
to the renames. These were written in the 80's and I'm sure a plumber
could make them pretty.
Bob.
This is incorrect or at least incomplete... (Sorry Adam)
If you are talking about CP's performance, or even CMS' performance,
then yes, it's a minor performance hit.
If you are talking about your Linux guests under your VM under their VM
under LPAR, then it is a very substantial hit.
On Tuesday, 03/11/2008 at 10:41 EDT, Lee Stewart
[EMAIL PROTECTED] wrote:
In addition to all the good advice already, I will add: Do not try to
run your VM system 2nd level if you have Linux guests. Performance of
the Linux guests will be terrible. There are only two levels of
hardware
Pipelines can connect to this. If you add IUCV *VMEVENT to your
directory entry, a pipe as simple as
pipe starmsg *vmevent|cons
will show you the records that are generated. This service is
documented in Chapter 23 of z/VM: CP Programming Services and it
shows the records that are created.
On
On Tuesday, 03/11/2008 at 12:02 EDT, Thomas Kern [EMAIL PROTECTED]
wrote:
It would be nice if there were some sample code to demonstrate
connecting
to
this new system service (*VMEVENT) and some sample records so that we
cou
ld
get a feel for the kind of data and the level of effort to
Simple samples work wonders.
Thanks for your quick help.
/Tom Kern
On Tue, 11 Mar 2008 12:15:15 -0400, Bruce Hayden [EMAIL PROTECTED] wro
te:
Pipelines can connect to this. If you add IUCV *VMEVENT to your
directory entry, a pipe as simple as
pipe starmsg *vmevent|cons
will show you the
OK we are running zLinux under zVM here. So from what I'm reading, z/vm
- z/vm - zlinux is not a very good idea???
Adam Thornton [EMAIL PROTECTED]
Sent by: The IBM z/VM Operating System IBMVM@LISTSERV.UARK.EDU
03/11/2008 09:53 AM
Please respond to
The IBM z/VM Operating System
On Tuesday, 03/11/2008 at 03:17 EDT, Rob van der Heij [EMAIL PROTECTED]
wrote:
On Tue, Mar 11, 2008 at 1:59 AM, Alan Altmark [EMAIL PROTECTED]
wrote:
While trying to reconstruct a source directory from an object
directory
might seem logical, far too much information is lost in
I would have to agree with Adam, it is a Disaster Recovery TEST,
performance is not your primary concern.
Bill Munson
VM System Programmer
201-418-7588
President MVMUA
http://www2.marist.edu/~mvmua/
Karl Kingston [EMAIL PROTECTED]
Sent by: The IBM z/VM Operating System
Well, it depends
If your disaster recovery box is near equal in size/resources as your recovery
workload, then you should think about a single VM solution.
However, if the disaster recovery box is much larger in size/resources, then
your recovery workload, then make it easy on your
While GLOBALV is useful for setting default behavior (e.g. DEFAULTS), I
wouldn't propose using it as a database of directory metadata. Adding
USER DIRECT metadata to the object directory seems useful and is
relatively straightforward to implement. I'd want filename, file
timestamp, and
Cross-posted to Linux-390, IBMVM and IBM-Main
For those unable to attend SHARE, or that don't have access to the restricted
areas of the SHARE web site, I've added 30 presentations from SHARE 100 in
Orlando, Florida to the linuxvm.org web site. You can find them at
I restored VM in an LPAR at DR tests twice a year for over 15 years and
never had a problem with it. Everybody says that performance is not an
issue at a DR test, until people complain about it or you can't finish th
e
test in the alloted time slot! Believe me, performance can be an issue,
folks ...
I don't see anywhere in the programming guide where DIAG 250 does and does
not work in terms of the whole disk. Clearly, the range of blocks to be
worked in a single D250 call must be standard sizes (eg: all 4K or all 512
or some such). But does the disk itself have to be formatted
On Tue, Mar 11, 2008 at 6:16 PM, Alan Altmark [EMAIL PROTECTED] wrote:
IMO, sanity checks are in the domain of directory maintenence products.
Fully agree, so for folks using XEDIT and DIRECTXA by hand (that we
were talking about) they *are* the directory maintenance product...
They may have
Are they reverting back to diag disabled after a reboot? I've had
this kind of problem making the diag driver stick for vdisks. The
Yast installer allows you to tell it that certain disks are diag
enabled, but then it seems to forget that when you first boot the
system. So - I enable it with
On Tue, Mar 11, 2008 at 8:49 AM, in message
[EMAIL PROTECTED],
Spracklen, Ken [EMAIL PROTECTED] wrote:
Hi,
I was wondering if someone could provide clarification on the set up to use
SNMP to manage virtual switches with z/VM 5.3.0.
In the z/VM TCP/IP Planning and Customization book
We run the IBM performance toolkit product.
I understand that MONWRITE feeds Perf tool kit. Is there a way I can
write MONWRITE files to disk as well as let them feed Perf tool kit?
Why do you ask, I want to use MXG to report on zVm.
Brian
On Tue, Mar 11, 2008 at 9:37 PM, Bruce Hayden [EMAIL PROTECTED] wrote:
Are they reverting back to diag disabled after a reboot? I've had
this kind of problem making the diag driver stick for vdisks. The
Yast installer allows you to tell it that certain disks are diag
enabled, but then it
On Tuesday, 03/11/2008 at 01:52 EDT, Bob Bolch [EMAIL PROTECTED] wrote:
There could be a lot of timestamps and source minidisks in a cluster
format
source directory.
Yeah, but who maintains such by using DIRECTXA directly?
Alan Altmark
z/VM Development
IBM Endicott
On Tuesday, 03/11/2008 at 08:49 EDT, Spracklen, Ken [EMAIL PROTECTED]
wrote:
I was wondering if someone could provide clarification on the set up to
use
SNMP to manage virtual switches with z/VM 5.3.0.
In the z/VM TCP/IP Planning and Customization book (chapters 21 23),
it
mentions the
Hi, Brian.
You don't need to run the MONWRITE virtual service machine simply to
feed the CP performance monitor data to PERFKIT. PERFKIT can collect the
raw monitor data directly from the MONDCSS saved segment.
You only need to run MONWRITE if you want it to collect the raw monitor
data to
Thanks Dave
Where would I find the instructions to setup and begin running monwrite?
Brian
-Original Message-
From: The IBM z/VM Operating System [mailto:[EMAIL PROTECTED] On
Behalf Of Dave Jones
Sent: Tuesday, March 11, 2008 5:48 PM
To: IBMVM@LISTSERV.UARK.EDU
Subject: Re: MONWRITE
On Tuesday, 03/11/2008 at 04:53 EDT, Mark Post [EMAIL PROTECTED] wrote:
VSWITCHes don't have any TCP/IP parameters associated with them, per se.
No IP
addresses, no routes, etc. The only reason z/VM TCP/IP is involved as
all is
as a controller of the OSA hardware. (The z/VM developers
On Tuesday, 03/11/2008 at 05:56 EDT, SUBSCRIBE IBMVM Ken Spracklen
[EMAIL PROTECTED] wrote:
Thanks for the info. So the example #5 in Chapter 23 of the zVM TCPIP
Planning and Customization book in the HOME section where it talks about
the VSWITCH keyword did not mean to imply that I had to
On Tue, Mar 11, 2008 at 5:15 PM, in message
[EMAIL PROTECTED], Rob van der Heij
[EMAIL PROTECTED] wrote:
-snip-
Since we're not speaking on Linux-390, I dare to say this ;-) A
Careful. The list has eyes everywhere, and we're watching you. :)
lot of this weird stuff is due to
Hi, Brian
Hamilton, Brian wrote:
Thanks Dave
Where would I find the instructions to setup and begin running monwrite?
Right here:
http://publibz.boulder.ibm.com/cgi-bin/bookmgr_OS390/BOOKS/hcsi1b20/3.2?SHELF=hcsh2a90.bksDT=20070424084605
(watch out for line wrap)
Brian
--
DJ
Thanks again...
And ideal on the amount of data were talking here on a typical day?
Typical mini disk size's??? 500 cyls, full 3390 mod 3, multiple
mod-3's...
This is currently what is being actively monitored
MONITOR DOMAIN ENABLED
PROCESSOR DOMAIN DISABLED
A different solution is to use ESAMON to feed MXG. That way, MXG gets all the network
data, Linux data, and the disk storage is about 1% of what raw data requires. ESAMON would
then be a complete replacement for Perf kit, monwrite, and a lot of disk space - not to
mention the cycles involved
On Tue, Mar 11, 2008 at 11:57 PM, Hamilton, Brian
[EMAIL PROTECTED] wrote:
And ideal on the amount of data were talking here on a typical day?
To my standards, a lot. For a decent z/VM with two dozen Linux
servers, think about one or more 3390-3 per day. My experience is that
installations who
I would estimate about 1 (or maybe a bit more) 3390-3 DASD per day to
hold the raw performance data stream from MONWRITE. Of course, if you
will want to keep more that one day's worth of raw data, you'll need
that much more extra DASD space reserved.
As Barton has already mentioned, it's a
We use a 3390 mod9 to hold the data. We also FTP it over to z/OS each
day.
Paul Feller
AIT Mainframe Technical Support
[EMAIL PROTECTED]
(319)-355-7824
-Original Message-
From: The IBM z/VM Operating System [mailto:[EMAIL PROTECTED] On
Behalf Of Hamilton, Brian
Sent: Tuesday,
On Tue, 11 Mar 2008, Karl Kingston wrote:
OK we are running zLinux under zVM here. So from what I'm reading, z/vm
- z/vm - zlinux is not a very good idea???
I have to back what Adam said from a managerial context.
There has been much discussion in recent weeks about virtualization
as to how
Installations using ESAMON might have a few hundred cylinders collecting the same data and
making MXG users VERY VERY VERY happy. And with the intersting monitor settings in use
by Brian, I would highly recommend attending a performance class to understand what they
mean. Could I recommend
On Wednesday, 03/12/2008 at 01:09 EDT, barton
[EMAIL PROTECTED] wrote:
MONWRITE was not even very well accepted 20 years ago when it became
available
because of the volume of data and mis-collection of data.
Eh? MONWRITE is a tool for collecting raw monitor data. It was not,
perhaps,
On Tuesday, 03/11/2008 at 04:39 EDT, Rob van der Heij [EMAIL PROTECTED]
wrote:
And on my wish list: the userid that issued the DIRECTXA command, so
we know where to put the monkey.
Ooooh! I like that one, too!
Alan Altmark
z/VM Development
IBM Endicott
Alan, look at what he's collecting. If you don't think that is miscollecting, you should
take the class too.
Alan Altmark wrote:
On Wednesday, 03/12/2008 at 01:09 EDT, barton
[EMAIL PROTECTED] wrote:
MONWRITE was not even very well accepted 20 years ago when it became
available
49 matches
Mail list logo