Upgrading the mainframe processor questions

JH

Upgrading the mainframe processor questions
To List:
We are upgrading the mainframe processor from a
2064-1C7 to a 2084-307. The CPU serial number will not
change. Our z/OS version is at 1.4. Our z/OS DB2 is
at V71.
Is there any potential problem for our DB2 systems?

Thanks
Joe



__________________________________
Do you Yahoo!?
Yahoo! Mail - now with 250MB free storage. Learn more.
http://info.mail.yahoo.com/mail_250

---------------------------------------------------------------------------------
Welcome to the IDUG DB2-L list. To unsubscribe, go to the archives and home page at http://www.idugdb2-l.org/archives/db2-l.html. From that page select "Join or Leave the list". The IDUG DB2-L FAQ is at http://www.idugdb2-l.org. The IDUG List Admins can be reached at [login to unmask email] Find out the latest on IDUG conferences at http://conferences.idug.org/index.cfm

Tom Moulder

Re: Upgrading the mainframe processor questions
(in response to JH)
There are a number of things that you should consider. Financially, DB2
will cost more since you will go from 247 MSUs to 402 MSUs. This is over a
60% increase in MSUs, but I suspect that the cost of DB2 will not increase
by 62%. Physically, you will still have 7 engines, so if you are currently
making use of parallelism, these processes should not change significantly.
The engines will be faster, but the number remains the same. If you have
not re-bound any packages/plans, then I would ocnsider this carefully as the
optimizer will use the faster engines in determining access paths and these
could change. If you wanted to be careful, you could trial Path Checker
from IBM which will allow you to run explains on the new processor and then
compare the entries in the plan table with entries from the time that the
previous processor was installed. You could see before the bind/rebind what
would change in access paths.

You said nothing about your real storage for each one. If that changed,
then you would need to take that into account for performance
considerations. You should get some performance improvement from
Bufferpools in data spaces, so if you have not done that in the past you
might consider making that change.

Tom Moulder
TREX Associates, Inc.
9728 Delmonico
Keller, TX. 76248-9559
+1 817-741-5549 Office
+1 817-741-5548 Fax
+1 832-858-4279 Mobile
www.t-rex-associates.com Web Site


-----Original Message-----
From: DB2 Data Base Discussion List [mailto:[login to unmask email] On Behalf
Of Joe Huang
Sent: Tuesday, January 11, 2005 12:56 PM
To: [login to unmask email]
Subject: [DB2-L] Upgrading the mainframe processor questions

To List:
We are upgrading the mainframe processor from a
2064-1C7 to a 2084-307. The CPU serial number will not
change. Our z/OS version is at 1.4. Our z/OS DB2 is
at V71.
Is there any potential problem for our DB2 systems?

Thanks
Joe



__________________________________
Do you Yahoo!?
Yahoo! Mail - now with 250MB free storage. Learn more.
http://info.mail.yahoo.com/mail_250

----------------------------------------------------------------------------
-----
Welcome to the IDUG DB2-L list. To unsubscribe, go to the archives and home
page at http://www.idugdb2-l.org/archives/db2-l.html. From that page select
"Join or Leave the list". The IDUG DB2-L FAQ is at http://www.idugdb2-l.org.
The IDUG List Admins can be reached at [login to unmask email] Find
out the latest on IDUG conferences at http://conferences.idug.org/index.cfm

--
No virus found in this incoming message.
Checked by AVG Anti-Virus.
Version: 7.0.300 / Virus Database: 265.6.10 - Release Date: 1/10/2005


--
No virus found in this outgoing message.
Checked by AVG Anti-Virus.
Version: 7.0.300 / Virus Database: 265.6.10 - Release Date: 1/10/2005

---------------------------------------------------------------------------------
Welcome to the IDUG DB2-L list. To unsubscribe, go to the archives and home page at http://www.idugdb2-l.org/archives/db2-l.html. From that page select "Join or Leave the list". The IDUG DB2-L FAQ is at http://www.idugdb2-l.org. The IDUG List Admins can be reached at [login to unmask email] Find out the latest on IDUG conferences at http://conferences.idug.org/index.cfm

Roger Miller

Re: Upgrading the mainframe processor questions
(in response to Tom Moulder)
That is a huge boost in processing power - nearly double - depending upon
the exact work load. That's the kind of change which causes some access
path changes, which are usually for the better, but not always. That much
more horsepower often means more work to do. Additional complexity and
load is a common cause of performance problems. We still see some
customers using MIPS (Misleading Indicator of Processor Speed) and
expecting the times to be in that ratio, but the reality is that various
work loads differ substantially. YMMV.

Roger Miller

On Tue, 11 Jan 2005 10:55:58 -0800, Joe Huang <[login to unmask email]> wrote:

>To List:
>We are upgrading the mainframe processor from a
>2064-1C7 to a 2084-307. The CPU serial number will not
>change. Our z/OS version is at 1.4. Our z/OS DB2 is
>at V71.
>Is there any potential problem for our DB2 systems?
>
>Thanks
>Joe
>
>__________________________________
>Do you Yahoo!?
>Yahoo! Mail - now with 250MB free storage. Learn more.
>http://info.mail.yahoo.com/mail_250
>

---------------------------------------------------------------------------------
Welcome to the IDUG DB2-L list. To unsubscribe, go to the archives and home page at http://www.idugdb2-l.org/archives/db2-l.html. From that page select "Join or Leave the list". The IDUG DB2-L FAQ is at http://www.idugdb2-l.org. The IDUG List Admins can be reached at [login to unmask email] Find out the latest on IDUG conferences at http://conferences.idug.org/index.cfm

Richard Humphris

Re: Upgrading the mainframe processor questions
(in response to Roger Miller)
Hi Roger,

Does this mean, in a datasharing environment with two different size
processors, you could get different access paths depending on which
processor does the bind?

Thanks,
Rich

-----Original Message-----
From: DB2 Data Base Discussion List [mailto:[login to unmask email] On
Behalf Of Roger Miller
Sent: Tuesday, January 11, 2005 3:38 PM
To: [login to unmask email]
Subject: Re: [DB2-L] Upgrading the mainframe processor questions


That is a huge boost in processing power - nearly double - depending
upon
the exact work load. That's the kind of change which causes some access
path changes, which are usually for the better, but not always. That
much
more horsepower often means more work to do. Additional complexity and
load is a common cause of performance problems. We still see some
customers using MIPS (Misleading Indicator of Processor Speed) and
expecting the times to be in that ratio, but the reality is that various
work loads differ substantially. YMMV.

Roger Miller

On Tue, 11 Jan 2005 10:55:58 -0800, Joe Huang <[login to unmask email]> wrote:

>To List:
>We are upgrading the mainframe processor from a
>2064-1C7 to a 2084-307. The CPU serial number will not
>change. Our z/OS version is at 1.4. Our z/OS DB2 is
>at V71.
>Is there any potential problem for our DB2 systems?
>
>Thanks
>Joe
>
>__________________________________
>Do you Yahoo!?
>Yahoo! Mail - now with 250MB free storage. Learn more.
>http://info.mail.yahoo.com/mail_250
>

------------------------------------------------------------------------
---------
Welcome to the IDUG DB2-L list. To unsubscribe, go to the archives and
home page at http://www.idugdb2-l.org/archives/db2-l.html. From that
page select "Join or Leave the list". The IDUG DB2-L FAQ is at
http://www.idugdb2-l.org. The IDUG List Admins can be reached at
[login to unmask email] Find out the latest on IDUG conferences
at http://conferences.idug.org/index.cfm

E-MAIL CONFIDENTIALITY NOTICE: The contents of this e-mail message and any attachments are intended solely for the
addressee(s) and may contain confidential and/or legally privileged information. If you are not the
intended recipient of this message or if this message has been addressed to you in error, please
immediately alert the sender by reply e-mail and then delete this message and any attachments. If you
are not the intended recipient, you are notified that any use, dissemination, distribution, copying, or
storage of this message or any attachment is strictly prohibited.

---------------------------------------------------------------------------------
Welcome to the IDUG DB2-L list. To unsubscribe, go to the archives and home page at http://www.idugdb2-l.org/archives/db2-l.html. From that page select "Join or Leave the list". The IDUG DB2-L FAQ is at http://www.idugdb2-l.org. The IDUG List Admins can be reached at [login to unmask email] Find out the latest on IDUG conferences at http://conferences.idug.org/index.cfm

Tom Moulder

Re: Upgrading the mainframe processor questions
(in response to Richard Humphris)
Yes. I have even seen access path changes depending upon the CPU busy of
the machine at the time of the bind.

Tom Moulder
TREX Associates, Inc.
9728 Delmonico
Keller, TX. 76248-9559
+1 817-741-5549 Office
+1 817-741-5548 Fax
+1 832-858-4279 Mobile
www.t-rex-associates.com Web Site


-----Original Message-----
From: DB2 Data Base Discussion List [mailto:[login to unmask email] On Behalf
Of Humphris,Richard P.
Sent: Friday, January 14, 2005 9:05 AM
To: [login to unmask email]
Subject: Re: [DB2-L] Upgrading the mainframe processor questions

Hi Roger,

Does this mean, in a datasharing environment with two different size
processors, you could get different access paths depending on which
processor does the bind?


Thanks,
Rich

-----Original Message-----
From: DB2 Data Base Discussion List [mailto:[login to unmask email] On Behalf
Of Roger Miller
Sent: Tuesday, January 11, 2005 3:38 PM
To: [login to unmask email]
Subject: Re: [DB2-L] Upgrading the mainframe processor questions


That is a huge boost in processing power - nearly double - depending upon
the exact work load. That's the kind of change which causes some access
path changes, which are usually for the better, but not always. That much
more horsepower often means more work to do. Additional complexity and load
is a common cause of performance problems. We still see some customers
using MIPS (Misleading Indicator of Processor Speed) and expecting the times
to be in that ratio, but the reality is that various work loads differ
substantially. YMMV.

Roger Miller

On Tue, 11 Jan 2005 10:55:58 -0800, Joe Huang <[login to unmask email]> wrote:

>To List:
>We are upgrading the mainframe processor from a
>2064-1C7 to a 2084-307. The CPU serial number will not
>change. Our z/OS version is at 1.4. Our z/OS DB2 is
>at V71.
>Is there any potential problem for our DB2 systems?
>
>Thanks
>Joe
>
>__________________________________
>Do you Yahoo!?
>Yahoo! Mail - now with 250MB free storage. Learn more.
>http://info.mail.yahoo.com/mail_250
>

------------------------------------------------------------------------
---------
Welcome to the IDUG DB2-L list. To unsubscribe, go to the archives and home
page at http://www.idugdb2-l.org/archives/db2-l.html. From that page select
"Join or Leave the list". The IDUG DB2-L FAQ is at http://www.idugdb2-l.org.
The IDUG List Admins can be reached at [login to unmask email] Find
out the latest on IDUG conferences at http://conferences.idug.org/index.cfm

E-MAIL CONFIDENTIALITY NOTICE: The contents of this e-mail message and any
attachments are intended solely for the

addressee(s) and may contain confidential and/or legally privileged
information. If you are not the

intended recipient of this message or if this message has been addressed to
you in error, please

immediately alert the sender by reply e-mail and then delete this message
and any attachments. If you

are not the intended recipient, you are notified that any use,
dissemination, distribution, copying, or

storage of this message or any attachment is strictly prohibited.

----------------------------------------------------------------------------
-----
Welcome to the IDUG DB2-L list. To unsubscribe, go to the archives and home
page at http://www.idugdb2-l.org/archives/db2-l.html. From that page select
"Join or Leave the list". The IDUG DB2-L FAQ is at http://www.idugdb2-l.org.
The IDUG List Admins can be reached at [login to unmask email] Find
out the latest on IDUG conferences at http://conferences.idug.org/index.cfm

--
No virus found in this incoming message.
Checked by AVG Anti-Virus.
Version: 7.0.300 / Virus Database: 265.6.11 - Release Date: 1/12/2005


--
No virus found in this outgoing message.
Checked by AVG Anti-Virus.
Version: 7.0.300 / Virus Database: 265.6.11 - Release Date: 1/12/2005

---------------------------------------------------------------------------------
Welcome to the IDUG DB2-L list. To unsubscribe, go to the archives and home page at http://www.idugdb2-l.org/archives/db2-l.html. From that page select "Join or Leave the list". The IDUG DB2-L FAQ is at http://www.idugdb2-l.org. The IDUG List Admins can be reached at [login to unmask email] Find out the latest on IDUG conferences at http://conferences.idug.org/index.cfm

Patrick Bossman

Re: Upgrading the mainframe processor questions
(in response to Tom Moulder)
Hello,
The optimizer considers the processor speed at bind time. The optimizer
also considers ridpool, sort pool, and bufferpool sizes.

The optimizer consideres the number of processors from a degree of
parallelism perspective only.

The optimizer does not pay attention to the CPU busy, just the processor
speed when determining the sequential path.

In my experience - assuming the SQL and physical design have not changed
The queries most likely to regress due to some optimizer input change are
those where the optimizer does not have a very accurate estimate to begin
with.
Eg. Default statistics, lack of correlation / skew statistics, or lack of
ability to use available statistics (host variable, parameter marker,
special register, expressions around columns).

So in cases where the optimizer has some deception or unknown, efficient and
inefficient access paths can have cost estimates which are very close. Then
some optimizer input changes (RUNSTATS, CPU speed, maintenance / new
release, etc.) enough so that the inefficient access path now has a cheaper
cost estimate, and there can be a wild swing in the actual performance.

Whenever we see these cases, our first check is to look at the statistics
and predicates and address potential pitfall areas. Address default
statistics, investigate possible correlation / skew, and look at any range
predicates which don't use literals (default filter factor often
significantly different from actual selectivity). In DB2 V8, this process
is automated by Statistics Advisor - it will even highlight range predicate
with markers to suggest that REOPT may be beneficial.

Regards,
Pat Bossman

---------------------------------------------------------------------------------
Welcome to the IDUG DB2-L list. To unsubscribe, go to the archives and home page at http://www.idugdb2-l.org/archives/db2-l.html. From that page select "Join or Leave the list". The IDUG DB2-L FAQ is at http://www.idugdb2-l.org. The IDUG List Admins can be reached at [login to unmask email] Find out the latest on IDUG conferences at http://conferences.idug.org/index.cfm