interesting to try and locate the knee.
I found a copy of the Reed Mullen chart I mentioned.
On a z990, the MP effect for VM kicked in seriously at 14 processors for VM
workload. For z/OS workload the MP effect became serious at 9 processors.
I suspect that the z10 is a very different animal,
...@listserv.uark.edu] On Behalf
Of David Boyes
Sent: Thursday, August 13, 2009 11:20 AM
To: IBMVM@LISTSERV.UARK.EDU
Subject: Re: [IBMVM] MP effect on z/VM Linux hosting
interesting to try and locate the knee.
I found a copy of the Reed Mullen chart I mentioned.
On a z990, the MP effect for VM kicked
On 8/13/09 2:22 PM, Marcy Cortes marcy.d.cor...@wellsfargo.com wrote:
Can you email me that chart?
Or is it paper?
And thanks Alan A for that link. That helps.
The copy I have is paper, but I'll try to get it scanned and send it to you
offlist.
@LISTSERV.UARK.EDU
Subject: Re: MP effect on z/VM Linux hosting
Can you email me that chart?
Or is it paper?
And thanks Alan A for that link. That helps.
Marcy
This message may contain confidential and/or privileged information. If you
are not the addressee or authorized to receive
@LISTSERV.UARK.EDU
Subject: Re: MP effect on z/VM Linux hosting
That's the first thing I thought of, too -- but note that
there is no z/V= M in this picture, just z/OS and PR/SM. z/OS
does recognize the existence of z/VM but it still does no= t
handshake with z/VM as far as I know. Have I
That's the first thing I thought of, too -- but note that there is no z/V
M in this picture, just z/OS
and PR/SM. z/OS does recognize the existence of z/VM but it still does no
t handshake with z/VM
as far as I know. Have I missed something?
Alan Ackerman
Alan (dot) Ackerman (at) Bank of
On Sunday, 08/09/2009 at 11:01 EDT, Alan Ackerman
alan.acker...@earthlink.net wrote:
The MP effect is determined by the operating system and by the workload.
It is not simply a hardware effect at all. If two or more processors
access the same memory location -- one will have to wait.
True
On Wednesday, 08/12/2009 at 01:01 EDT, Alan Ackerman
alan.acker...@earthlink.net wrote:
That's the first thing I thought of, too -- but note that there is no
z/VM
in this picture, just z/OS and PR/SM.
z/OS does recognize the existence of z/VM but it still does not
handshake with z/VM as far
Does this mean that z/OS is finally catching up to the VM/DOS handshaking of
years gone by?
Regards,
Richard Schuh
z/OS has a new feature HiperDispatch to help reduce the MP
and NUMA effec= t. In the Austin SHARE proceedings, see 2831
- System z10: HiperDispatch From a Sysprog=
How many IFLs are we talking about here? At a guess, if you are going fro
m 4 to 8 IFLs, you won't
see any MP effect, but if you are going from 30 to 60 IFLs, you will defi
nitely see an MP effect.
What I can remember about the MP effect:
The MP effect is determined by the operating system and
-Original Message-
From: The IBM z/VM Operating System [mailto:ib...@listserv.uark.edu] On
Behalf Of John McKown
Sent: 08 August 2009 04:04
To: IBMVM@LISTSERV.UARK.EDU
Subject: Re: MP effect on z/VM Linux hosting
On Fri, 7 Aug 2009, Marcy Cortes wrote:
I'm getting conflicting
missing something?
Paul Feller
AIT Mainframe Technical Support
-Original Message-
From: The IBM z/VM Operating System [mailto:ib...@listserv.uark.edu] On Behalf
Of Marcy Cortes
Sent: Friday, August 07, 2009 7:50 PM
To: IBMVM@LISTSERV.UARK.EDU
Subject: MP effect on z/VM Linux hosting
I'm
is floating around out
there somewhere. I did not see or hear of anything about the UP/MP effect
inside Linux itself.
From: The IBM z/VM Operating System on behalf of Marcy Cortes
Sent: Fri 8/7/2009 8:49 PM
To: IBMVM@LISTSERV.UARK.EDU
Subject: MP effect on z/VM
First, IBM has never measured this for z/VM, their resources devoted to
things like LSPR for VM are long gone. So you are either getting a
sale's person's guess, or a z/OS guess and I guess they guess different?
Amdahl's Law of Multiprogramming from a software perspective states
that a
On 8/7/2009 at 8:49 PM, Marcy Cortes marcy.d.cor...@wellsfargo.com
wrote:
-snip-
(z/OS per
engine pricing actually goes down to compensate for this, z/VM's does not).
Are you talking about z/VM itself, or products that run on z/VM? Just to
clarify for others reading this, the one-time
effect on z/VM Linux hosting
On 8/7/2009 at 8:49 PM, Marcy Cortes marcy.d.cor...@wellsfargo.com
wrote:
-snip-
(z/OS per
engine pricing actually goes down to compensate for this, z/VM's does not).
Are you talking about z/VM itself, or products that run on z/VM? Just to
clarify for others
: The IBM z/VM Operating System [mailto:ib...@listserv.uark.edu] On Behalf
Of Marcy Cortes
Sent: Saturday, August 08, 2009 9:08 AM
To: IBMVM@LISTSERV.UARK.EDU
Subject: Re: [IBMVM] MP effect on z/VM Linux hosting
I'm not so concerned about the pricing - just wondered if it indicated that it
didn't have
for
your cooperation.
-Original Message-
From: The IBM z/VM Operating System [mailto:ib...@listserv.uark.edu] On Behalf
Of Marcy Cortes
Sent: Saturday, August 08, 2009 9:09 AM
To: IBMVM@LISTSERV.UARK.EDU
Subject: Re: [IBMVM] MP effect on z/VM Linux hosting
And I mistated the pricing thing
On 8/8/2009 at 12:07 PM, Marcy Cortes marcy.d.cor...@wellsfargo.com
wrote:
-snip-
The real concern is whether to put 6 IFLS on 7 different boxes or do we lose
anything by having say 14 IFLs on 3 boxes or 10-10-11-11 on 4 boxes.
Based on what Jim Vincent and Rick Barlow have been saying for
I'm getting conflicting answers from IBM.
How does VM scale with regards to multiprocessors? In the z/OS world the 1st
z/10 engine is like 900 something MIPS and the 60th on the box is something
like 359. Does z/VM hosting Linux suffer the same fate? (z/OS per engine
pricing actually goes
On Fri, 7 Aug 2009, Marcy Cortes wrote:
I'm getting conflicting answers from IBM.
How does VM scale with regards to multiprocessors? In the z/OS world
the 1st z/10 engine is like 900 something MIPS and the 60th on the box
is something like 359. Does z/VM hosting Linux suffer the same
Reed Mullen used to have a chart in one of his presentations that
showed that the MP effect was still present for VM workload but that
the impact was smaller than for Z/OS and kicked it later in the curve.
I don't remember the precise numbers (and the z10 probably has changed
everything)
22 matches
Mail list logo
