BMC vs. IBM Utility Benchmarks....revisited

Gerald Hodge

BMC vs. IBM Utility Benchmarks....revisited


Think of any utility suite as a very sophisticated tool kit. You will never
use all of the tools, even if you want too. What is of interest is which of
the tools you will use. You need to get this point to management as soon as
possible or they will make what is simply a financial decision, because
everything else is equal and the "vendor" has told them so.



I am aware of a friend who was told that he had to accept the deal made
because it was "made." He demonstrated that the "new" vendor's utilities
could not complete in 24 hours work that needed to be turned around
overnight. Several of these overnight jobs had to be canceled because they
took days and no end was insight. The company ended up paying for both
suites because the decision was made without technical input.



Set up some reasonable examples and have competing vendors prove that they
can do the work in the time allotted. Vendors play games with CPU time, but
have difficulty fudging elapsed time. The business requirements should be
the telling factor. Saving money and going out of business are not
acceptable to almost anyone but the vendor cashing the first check.



Gerald Hodge

HLS Technologies, Inc.

_____

From: DB2 Data Base Discussion List [mailto:[login to unmask email] On Behalf
Of Joel Goldstein - Responsive Systems
Sent: Friday, December 14, 2007 3:19 PM
To: [login to unmask email]
Subject: Re: [DB2-L] BMC vs IBM Utility Benchmarks....revisited



Phil hits it right on the head here.



Results will vary at every installation, and there are too many variables
for anyone else's benchmark numbers to bear relevance to your environment.



Since these queries often originate from some bean counter trying to
$1.98...., there is much more that should be considered than just money.

It should be more about features, ease of use, and making the DBA staff more
efficient.



If somebody is thinking about conversion, that "somebody" should also factor
in the people and machine time cost of conversion.

For large installations, this can easily be a 7 figure cost...



Regards,

Joel







Joel Goldstein
Responsive Systems
Buffer Pool Tool for DB2, the worldwide industry standard
Performance software that works......
Predicts Group Buffer Pool performance too!
www.responsivesystems.com
tel. (732) 972-1261
fax.(732) 972-9416


The IDUG DB2-L Listserv is only part of your membership in IDUG. DB2-L list
archives, the FAQ, and delivery preferences are at
< http://www.idug.org/lsidug > www.idug.org under the Listserv tab. While at
the site, you can also access the IDUG Online Learning Center, Tech Library
and Code Place, see the latest < http://www.idug.org/lsconf > IDUG conference
information, and much more.
If you have not yet signed up for Basic Membership in IDUG, available at no
cost, click on < http://www.idug.org/lsms > Member Services

The IDUG DB2-L Listserv is only part of your membership in IDUG. DB2-L list archives, the FAQ, and delivery preferences are at http://www.idug.org/lsidug under the Listserv tab. While at the site, you can also access the IDUG Online Learning Center, Tech Library and Code Place, see the latest IDUG conference information, and much more. If you have not yet signed up for Basic Membership in IDUG, available at no cost, click on Member Services at http://www.idug.org/lsms

Nick Cianci

Re: BMC vs IBM Utility Benchmarks....revisited
(in response to Gerald Hodge)
It may be wrong, but if you are trying to justify the time to run a set
of comparative analysis between various products, often it helps to
point out to you boss (or more importantly his\her boss) the type of
savings that can be made! So public benchmarks are not a total waste of
time.

Whilst I agree with the sentiment that you got to analyse the products
on your on environment, you also need to be able to simulate your
production environment to get accurate benefit analysis. The trouble
here is I've only worked at one site in over 20 years where they've
actually kept a decent volume in their Stress & Volume environment.


Regards,

Nick CIANCI

________________________________

From: DB2 Data Base Discussion List [mailto:[login to unmask email] On
Behalf Of Rasmussen, Steen
Sent: Monday, 17 December 2007 7:14 AM
To: [login to unmask email]
Subject: Re: [DB2-L] BMC vs IBM Utility Benchmarks....revisited



I like to throw in my few Cents - whether these are US or
European Cents.



Since more and more DB2 sites have been are being outsourced - I
have notices more and more interest in looking into the CPU consumption.
Elapsed time can be important, and so can the features being available
to accomplish whatever you need to accomplish - but if you are being
charged by the CPU consumption - some other figures might become less
interesting.



My point is - like Grainger, Raymond and others have pointed out
- you will need to look at your specific scenarios, requirements, how
you pay and needed features.



Have fun - doing benchmarks is a lot of fun :-)



Steen Rasmussen
CA

Principal Technical Specialist DB2 tools

IBM Certified Database Associate -- DB2 9 Fundamentals

IBM Certified Database Administrator - DB2 9 DBA for z/OS

Tel : +1-630-505-6673 (US direct)
Tel : +1-815-274-9589 (US mobile)

Tel : +45-22 15 44 98 (Europe mobile)
[login to unmask email]
< http://www.ca.com >


________________________________


The IDUG DB2-L Listserv is only part of your membership in IDUG.
DB2-L list archives, the FAQ, and delivery preferences are at
www.idug.org < http://www.idug.org/lsidug > under the Listserv tab. While
at the site, you can also access the IDUG Online Learning Center, Tech
Library and Code Place, see the latest IDUG conference information
< http://www.idug.org/lsconf > , and much more.
If you have not yet signed up for Basic Membership in IDUG,
available at no cost, click on Member Services
< http://www.idug.org/lsms >
The IDUG DB2-L Listserv is only part of your membership in IDUG.
DB2-L list archives, the FAQ, and delivery preferences are at
www.idug.org < http://www.idug.org/lsidug > under the Listserv tab. While
at the site, you can also access the IDUG Online Learning Center, Tech
Library and Code Place, see the latest IDUG conference information
< http://www.idug.org/lsconf > , and much more.
If you have not yet signed up for Basic Membership in IDUG,
available at no cost, click on Member Services
< http://www.idug.org/lsms >


The IDUG DB2-L Listserv is only part of your membership in IDUG. DB2-L list archives, the FAQ, and delivery preferences are at http://www.idug.org/lsidug under the Listserv tab. While at the site, you can also access the IDUG Online Learning Center, Tech Library and Code Place, see the latest IDUG conference information, and much more. If you have not yet signed up for Basic Membership in IDUG, available at no cost, click on Member Services at http://www.idug.org/lsms

Isaac Yassin

Re: BMC vs IBM Utility Benchmarks....revisited
(in response to Nick Cianci)
Hi,



Just to complicate things.

In one of those outsourced there was a price for I/O.

The outsourcer changed all files to be lrecl=blksize …



Isaac Yassin

From: DB2 Data Base Discussion List [mailto:[login to unmask email] On Behalf Of Rasmussen, Steen
Sent: Sunday, December 16, 2007 10:14 PM
To: [login to unmask email]
Subject: Re: [DB2-L] BMC vs IBM Utility Benchmarks....revisited



I like to throw in my few Cents – whether these are US or European Cents.



Since more and more DB2 sites have been are being outsourced – I have notices more and more interest in looking into the CPU
consumption. Elapsed time can be important, and so can the features being available to accomplish whatever you need to accomplish –
but if you are being charged by the CPU consumption – some other figures might become less interesting.



My point is – like Grainger, Raymond and others have pointed out – you will need to look at your specific scenarios, requirements,
how you pay and needed features.



Have fun – doing benchmarks is a lot of fun J



Steen Rasmussen
CA

Principal Technical Specialist DB2 tools

IBM Certified Database Associate -- DB2 9 Fundamentals

IBM Certified Database Administrator - DB2 9 DBA for z/OS

Tel : +1-630-505-6673 (US direct)
Tel : +1-815-274-9589 (US mobile)

Tel : +45-22 15 44 98 (Europe mobile)
HYPERLINK "mailto:[login to unmask email]"[login to unmask email]
HYPERLINK "http://www.ca.com"

_____


The IDUG DB2-L Listserv is only part of your membership in IDUG. DB2-L list archives, the FAQ, and delivery preferences are at
HYPERLINK "http://www.idug.org/lsidug"www.idug.org under the Listserv tab. While at the site, you can also access the IDUG Online
Learning Center, Tech Library and Code Place, see the latest IDUG conference information, and much more.
If you have not yet signed up for Basic Membership in IDUG, available at no cost, click on Member Services
The IDUG DB2-L Listserv is only part of your membership in IDUG. DB2-L list archives, the FAQ, and delivery preferences are at
HYPERLINK "http://www.idug.org/lsidug"www.idug.org under the Listserv tab. While at the site, you can also access the IDUG Online
Learning Center, Tech Library and Code Place, see the latest IDUG conference information, and much more.
If you have not yet signed up for Basic Membership in IDUG, available at no cost, click on Member Services

No virus found in this incoming message.
Checked by AVG Free Edition.
Version: 7.5.503 / Virus Database: 269.17.4/1188 - Release Date: 17/12/2007 14:13


No virus found in this outgoing message.
Checked by AVG Free Edition.
Version: 7.5.503 / Virus Database: 269.17.4/1188 - Release Date: 17/12/2007 14:13


The IDUG DB2-L Listserv is only part of your membership in IDUG. DB2-L list archives, the FAQ, and delivery preferences are at http://www.idug.org/lsidug under the Listserv tab. While at the site, you can also access the IDUG Online Learning Center, Tech Library and Code Place, see the latest IDUG conference information, and much more. If you have not yet signed up for Basic Membership in IDUG, available at no cost, click on Member Services at http://www.idug.org/lsms

Avram Friedman

Re: BMC vs IBM Utility Benchmarks....revisited
(in response to Isaac Yassin)
List

To throw in my 2 cents ...
Assuming the interest in a benchmark is to evaluate
CPU
I/O counts
elapsed run time

Might I suggest a paper and pencil approach instead of an artificial bench
mark.

The biggest contributing factor is
Inside Vs Outside DB2

There are a lot of information sources on this
For example
reviews on VSAM vs DB2 performance

Some VERY rough numbers and comments

DB2 Utility CPU will be (if you can mesure it) will be about 2x VSAM REPRO CPU
for COPY. This is due to the overhead of Locking, multi address space
management, and shared buffer management.

I/O counts (if you can mesure it) will be reduced for well tuned VSAM vs DB2.
This is because VSAM can reliably read many CI's with a single EXCP. DB2
prefetch can do this too but prefetch is crippled by several factors (a cap on
the number of CIs, and cooperation with other pool users to name two)

Elapsed time is largly controlled by I/O counts.

The example comments are based on copy only. Copy is not the only utility.
One more suggestion would be to compare IBM classic REORG to IBM OnLine
Reorg .... I bet (I have not looked) this is discussed in the DB2 V?
performance topics redbook for the version that introduced online reorg and
lots of other places. I know this comparision is not vendor vs vendor but will
provide a paper and pencil hint as to what happens as you move away for full
MSTR and DBM1 control.

To the extent toat one might consider the outside DB2 solution to be far
superior, this can be demostrated with out adding the extra legal complication
of vendor vs vendor. Why not ask about performance diffrences between the
basic IBM utility suite and the advanced IBM tools. Once you determine the
magnitue of the improvement if any you can then concentrate on some of the
other non run time perforance issues like license fees, ease of use, quality of
support, currency with new releases, who has the best gifts etc.

I would like to stress that while this benchmark, paper and pencil estimate,
actual mesurement question is very interesting as well as pointing out outside
of DB2 performance almost always wins it is not only issue to be considered in
selecting a utility product or suite.

The question in general falls into a group of querries that interest me
proffessionally ... "Finding the proper issue in public Q&A"

Best wishes
Avram Friedman

The IDUG DB2-L Listserv is only part of your membership in IDUG. DB2-L list archives, the FAQ, and delivery preferences are at http://www.idug.org/lsidug under the Listserv tab. While at the site, you can also access the IDUG Online Learning Center, Tech Library and Code Place, see the latest IDUG conference information, and much more. If you have not yet signed up for Basic Membership in IDUG, available at no cost, click on Member Services at http://www.idug.org/lsms