Actually, Alan, it's the HRM003 after the DB2 abend, not the FRF002, that
VM63978 should fix. The FRF002 is fixed (circumvented) by the DB2 APAR y
ou
previouly applied, and we're still investigating the CP aspects under APA
R
VM63975 (yet another nearby number).
- Bill Holder, IBM z/VM Develop
Yes, that's the one, Alan, thanks. With the numbers so similar, and work
ing
on both at the same time, both involving some of the same code (though
totally different scenarios and symptoms), I should've known to double ch
eck
everything before posting. We should have VM63978 turned around quickly
This weekend we successfully tested a fixtest for this problem. The APAR
is VM63978. Seems to
be related to heavy DB2 use on a z990. (Perhaps also a z9-109 or z890.) O
nly happens first level.
IBM thinks the APAR will fix the FRF002 abend we had, as well as the DB2
crash mentioned below.
On Sa
Is the HPMA problem our APAR VM63978? (Only one digit different.) We had
a successful fix test on
that one this weekend.
I have to apologize, I'm confusing two different problems I've been worki
ng
on. The hang condition that Martha reported is not the one related to
HPMA enabled hardware or "logically zero" page states. Rather, it has to
do
with some unusual behavior by DB2 for VM (with dataspaces) in its usage o
Not even SFS with dataspace enabled directories?
Regards,
Richard Schuh
-Original Message-
From: The IBM z/VM Operating System [mailto:[EMAIL PROTECTED] On Behalf Of
Bill Holder
Sent: Monday, March 27, 2006 2:11 PM
To: IBMVM@LISTSERV.UARK.EDU
Subject:Re: Update on z/VM
Marist has a job opening for someone with Linux experience, especially
Linux on zSeries (we have over 500 servers). If you are interested,
please go to http://www.marist.edu/humanresources/employ.html?id=369
Martha
On Mon, 27 Mar 2006 16:11:09 -0600, Bill Holder <[EMAIL PROTECTED]> wrot
e:
>There are two aspects of the problem, only one of which is directly Diag
nose
>x'10' / release processing related. The other aspect has to do with
>handling of the "logically 0" page state created by Diagnose x'10', when
There are two aspects of the problem, only one of which is directly Diagn
ose
x'10' / release processing related. The other aspect has to do with
handling of the "logically 0" page state created by Diagnose x'10', when
the
HPMA function is enabled and substantial paging load is present. I reall
We haven't IPLd for a time change in 3 years. In fact, if it wasn't
for a CP upgrade and the occasional power problem, we wouldn't re-IPL VM
at all. If there is an app that doesn't like it, that gets recycled
individually. We have a wide variety of z/OS, Linux and CMS based apps and
only a coupl
Chuck,
I got something from Kris Beulens at IBM that works
very nicely for us. These ir a package from VM
Downloads called RXSERVER. Part of it is a VMUTIL
EXEC: create a VMUTIL userid, place the sample PROFILE
EXEC, VMUTIL EXEC and VMUTIL PROFILE on its 191 disk,
define your userid's in the AUTH_
Then this is strictly a DIAG 10 problem? If so, that means that my system
hosting 100+ TPF and 0 BD2 is not susceptible, right?
We will be upgrading on 8 April. We tried a week ago but ran into a severe
problem in other vendor code and had to fall back. We have the fix for that
problem and are
I have tested Sunday's change with CMS and linux guests in 2nd level
z/VM and a test LPAR. The SET TIMEZONE works great! I IPL'ed at
something like 04/02/06 01:58:00, saw my WAKEUP SET TIMEZONE EDT issued
two minutes later, watched my CMS servers do well, and checked linux
time with the date co
The other
option is to run the system on GMT so that you do not have to bow to the
vagaries of government, state or federal J That works for us because we have datacenters and users scattered
around the globe.
Regards,
Richard Schuh
-Original
Message-
From: The IBM z/VM Ope
Nice approach - if you can use it. IPLing is not an option for some
shops. My client has lotsa linux v.m.s and way lighter on the CMS side.
IPLing is not an option. With the light CMS load, critical as it may be,
my testing of SFS, RACF, WAKEUP servers, backups all look OK. But it
does take o
Title: RE: Time Change
Thanks for the quick responses. We only have Linux guests so it appears the SET TIMEZONE command will work for the VM side of things.
_
From: Kreiter, Chuck
Sent: Monday, March 27, 2006 3:55 PM
To: IBMVM
Sub
Thanks, Martha. Running DB2 for VM with plenty of other load is the main
exposure, as the behavior which triggers the problem (contact me directly
if
you want the gory details) is quite unusual and unique to DB2, as far as
we
know. There's a small possibility of hitting it while running a Linu
Chuck,
Dig out your "z/VM CP Planning
and Administration" manual (or see: http://publibz.boulder.ibm.com/epubs/pdf/hcsg2b10.pdf
)
Look up "TIMEZONE_BOUNDARY Statement",
see "Usage Notes" item #2.
In short, your easiest and most reliable
course of action is to IPL after 03:15 AM in the spring and
Ahh ... I remembered incorrectly about the SFS servers then.
>The settings in SYSTEM CONFIG are read only at IPL time. CP will look
>for the most recent defiintion that has passed and set the tz to that.
>You can dynamically issue the SET TIMEZONE EDT command this Sunday at 2
>a.m. to accomplish
The Sine Nomine presentations from the 2006 European zSeries
Expo in Dusseldorf
held last week are available for download from http://www.sinenomine.net/downloads/zexpo-dusseldorf.
Note that sessions L06 and L56 are substantially updated from the versions on
the conference CD, so I’d recom
Chuck, the System Config is only read at boot-time. At MSU, we
IPL weekly (on Sunday night), so it takes effect at that time.
We also issue the CP SET TIMEZONE command at the
actual time-change using the VMUTIL service machine via WAKEUP).
Not all products/guests handle the time-change properly ho
The settings in SYSTEM CONFIG are read only at IPL time. CP will look
for the most recent defiintion that has passed and set the tz to that.
You can dynamically issue the SET TIMEZONE EDT command this Sunday at 2
a.m. to accomplish it as well. Virtual machines that have requested the
tz signa
Title: Time Change
Being new to z/VM, I was wondering what most users do for the time change? I noticed several definitions SYSTEM CONFIG that appear to handle that automagically, but wanted to see if there was a procedure that needed to be followed to activate it.
Thanks,
Chuck Kreiter
Judson,
Add the line
using namespace std;
after the "#include " in the hello cpp source, recompile, and
rebuild.
This should make all those error messages disappear.
Mike Donovan
"West, Judson"
I had promised to update those attending our session via the list, which
is why I posted the note.
Kurt was kind enough to post the APAR description of the problem. The symptoms
are very bizarre. It begins with virtual machines being unable to logoff
the system (going into logoff/force pending m
Thank you
--- "Stracka, James (GTI)" <[EMAIL PROTECTED]>
wrote:
> APAR VM63968 says:
>
>ERROR DESCRIPTION:
> This was originally seen as described in PITS
> 2T0Q191.
> This is a deadlock between 2 tasks where one
> holds
> PTIL-X and the other wants a share of PTIL. The
> task t
APAR VM63968 says:
ERROR DESCRIPTION:
This was originally seen as described in PITS 2T0Q191.
This is a deadlock between 2 tasks where one holds
PTIL-X and the other wants a share of PTIL. The task that holds
PTIL-X is deferred on PTE serialization and won't run until
a redr
I don't believe I have seen a description of their
problem. We are planning on an install of zVM 5.2 for
a new client. Is there a description of the problem
available?
Thks,
Bill J.
P.S. Some of us are not afforded the opportunity to
attend SHARE.
--- Kurt Acker <[EMAIL PROTECTED]> wrote:
> Th
Larry--My "guess" is that your linux guests do not have RUN set
ON. You could verify that by entering #CP Q SET from a linux guest
to which you are connected at a VM console (not putty). You should
be able to see, RUN ON in response to the query.
q
set
The code for this was written as a circumvention
and is still under review. As initial performance testing has been
very positive, it may wind up becoming the official fix, however it is
still to early for us to make that decision. No matter what, the
fix test can be obtained by opening a PMR (t
We have been having a problem with our linux guests on VM. We lose
connectivity through putty and have to shutdown using the signal shutdown
command to the linux guest. IBM has found the problem and we need to
upgrade the kernel..I guess to a general...I know don't quit my day job.
Anyway wha
Thanks Eginhard, your solution is easy to implement and I can code the
EXEC to check an exempt list to avoid disconnecting OPERATOR's console
or some other exempt userids.
This e-mail, including any attachments, may be confidential, privi
The IDLEUSER report is accurate and up-to-date as of its timestamp.
But I'm using it to implement a security policy to disconnect userids
(people) that have been idle more than a specified number of minutes.
So I need to look at the IDLEUSER report immediately after it's
generated otherwise I co
SEND was just suggested as an alternative
to logging onto and gracefully shutting down/logging off the machine. FORCE
is kind of brutal as if their is any outstanding work that needs to complete,
you wind up risking putting a guest (and this is true for any guest) into
logoff force pending status
On Monday, 03/27/2006 at 08:09 PST, "Schuh, Richard" <[EMAIL PROTECTED]>
wrote:
> SEND does not require any programming. A class C user can issue the SEND
> command to any user without having to be SECUSER; non-C users must be
the
> SECUSER before they can send a command to be executed by the
SEND does not require any programming. A class C user can issue the SEND
command to any user without having to be SECUSER; non-C users must be the
SECUSER before they can send a command to be executed by the target machine.
Regards,
Richard Schuh
-Original Message-
From: The IBM z/VM
On 3/27/06, Kurt Acker <[EMAIL PROTECTED]> wrote:
> Just log onto guest and issue "logoff" (You could also send it a #CP LOGOFF
> from a secuser id if you wanted to avoid logging into that user).
Any reason why SEND CP PERFSVM LOGOFF is better than the proposed
FORCE PERFSVM ? Doing the SEND p
We are also planning to upgrade to z/VM 520 within the next few weeks - any
idea when the fix will close so it can be ordered.
Martha McConaghy
<[EM
Those of you who attended our z/VM 5.2.0 Early Experiences presentation
at SHARE know that Marist had some serious problems that forced us to back
out our upgrade. An APAR has been opened on the problem, VM63968. Not
long after SHARE, another customer also had a similar problem, so it is
not iso
Just log onto guest and issue "logoff"
(You could also send it a #CP LOGOFF from a secuser id if you wanted
to avoid logging into that user).
Kurt Acker
"Romanowski, John
(OFT)" <[EMAIL PROTECTED]>
Sent by: The IBM z/VM Operating System
03/27/2006 10:02 AM
Please respond to
The IBM
Anyone know how to tell the PERFSVM disconnected service machine to stop
and logoff?
Or should I just CP FORCE it? I'm looking for a command to shut it
down gracefully as part of my VM shutdown sequence.
This e-mail, including any attachm
41 matches
Mail list logo