[egr-rg] GGF13 EGR-RG Use Case Template ...

[Subramaniam, Ravi]

Hi Toshi,

Thanks and I understand. 

I missed Robert Fogel's original document. Was there something other than the first document that I put together for GGF13? Robert F had  input that I incorporated amongst others. I note that your GGF presentation took from that document that I worked on so I could be missing something after that.

Thanks!

Ravi

--------------------------
Ravi Subramaniam
Intel


-----Original Message-----
From: Toshiyuki Nakata <nakata at mtg.biglobe.ne.jp>
To: Subramaniam, Ravi <ravi.subramaniam at intel.com>
CC: Robert Cohen <bcohen at bway.net>; Satoshi Itoh <satoshi.itoh at aist.go.jp>; t-nakata at cw.jp.nec.com <t-nakata at cw.jp.nec.com>; Fogel, Robert <robert.fogel at intel.com>; egr-rg at gridforum.org <egr-rg at gridforum.org>
Sent: Wed Mar 23 23:58:02 2005
Subject: Re: [egr-rg] GGF13 EGR-RG Use Case Template ...

Hi:

> I hope the newer version of the documents incorporates your feedback and
> the feedback from GGF13 (that I had to miss). I have not read the newer
> version with the recent updates. Satoshi?
> 

March is the end of fiscal year 2004 in Japan so both Satoshi and I are 
up to our necks with various things.

I hope to edit Robert (Fogel)'s original doc to reflect the
comments in GGF13 as well as Robert (Cohen)'s comments early in April.

Thanks & Best Regards
Toshi

> Thanks!
> 
> Ravi
> 
> -----Original Message-----
> From: Robert Cohen [mailto:bcohen at bway.net] 
> Sent: Monday, March 14, 2005 12:43 PM
> To: Subramaniam, Ravi
> Cc: Satoshi Itoh; t-nakata at cw.jp.nec.com; Fogel, Robert;
> egr-rg at gridforum.org
> Subject: Re: [egr-rg] GGF13 EGR-RG Use Case Template ...
> 
> Ravi,
> 
> I read through Rev 0 of the template and it covers many of the main
> topics that should be addressed in a case study.
> 
> What would make this case analysis far more valuable in terms of ROI or
> some estimate of impacts would be to try to collect some quantitative
> data that would take the case studies to a more useful level.
> 
> For instance, if the user was able to provide some type of comparison
> between the before and after situation, that would be particularly
> helpful. One example might be:
> 
>     1. I did product simulation analysis that previously took 8 minutes
> for each simulation and now it takes two minutes.
> 
> More useful than this would be some detail in terms of what was done
> previously and what the new cluster or grid can do:
> 
>     1. When my department did product simulations, we used a Unix server
> with x CPUs and y Gig of memory. Replacing this with a 20 CPU Linux
> cluster let us take the time from 8 minutes to 2 minutes for each
> simulation. We did z simulations a week, so this saved a total of z*6
> minutes of time each week.
> 
> Even better would be:
> 
>     1. My department did product simulations using a Unix server with x
> CPUs and y Gig of memory. It cost "a" for Linux hardware, "b" for
> middleware and "c" for systems integration (by in-house or outside
> integrators) so the total new system cost was "d." This new system was
> able to reduce the time for product simulations from 8 minutes to 2
> minutes. The new hardware and sofware plus integration cost was "d" 
> compared to the previous cost of "e." In addition, we had additional
> cost savings because: we saved the cost of "f" for a programmer that was
> not needed and "g" for storage, software licenses, etc., that were not
> purchased, etc.
> 
> Note: With these parameters, it would be possible to do type of cost
> savings analysis that would approximate an ROI and make the collection
> of data more valuable.
> 
> Some of this data might be collected by adding a few short queries on
> the cost savings items in the relevant sections.
> 
> 
> Bob
> 
> 
> Subramaniam, Ravi wrote:
> 
> 
>>Hi Satoshi/Toshi,
>>
>>Here is the Rev 0 of the template for use cases for discussion at the
>>GGF-13 EGR session. This is based on my thinking and input from Robert
>>and my participations in discussions on OGSA use cases. The expectation
>>is that further discussion with the group will help us refine this.
>>
>>1. We need to come up with a way to categorize the use cases. We need
>>taxonomy so that we can organize the use cases as they come. I have
>>taken a stab at this in the document but there needs to be discussion
> 
> at
> 
>>the session.
>>
>>2. We also need to recruit volunteers to
>>
>>a) develop and submit use cases
>>b) visit the email archives of the different WGs and RGs in GGF and
>>submit any use case related activity to EGR-RG (preferably reps from
> 
> the
> 
>>groups but others could be welcome). We already have the OGSA use cases
>>but it would be necessary to get the ones from other groups.
>>c) take previously submitted use cases and transform them to the
>>"EGR-RG" template
>>
>>Hope you are having a good GGF. I was hoping I would be there but an
>>important meeting with my management has been planned for this week. I
>>had to cancel my plans. Sorry for the late notice and inconvenience.
>>
>>Ravi
>> 
>>
> 
> 
> 

-- 
Toshiyuki Nakata
t-nakata at cw.jp.nec.com
+81-44-431-7653
(NEC Internal 8-22-60210)