- ogsa-wg - lists.ogf.org

HPC Use Case ....
by Subramaniam, Ravi 21 Feb '06

21 Feb '06

Hi, Based on the discussion on the F2F today, I have an action item to resend the use case that I submitted to cover the HPC space. This was first sent Feb, 2003 to OGSA. I had called it infrastructure management but the first example is a HPC scenario. We may need to flesh this out and separate the relevant portions into a new use case definition for the HPC focus activities. Thanks! Ravi

2 1

RE: [ogsa-wg] Paper proposing "evolutionary vertical design efforts"
by Maguire_Tom＠emc.com 21 Feb '06

21 Feb '06

I appreciate and agree with the perspective that suggests that OGSA functionality be profiled using existing agreed upon specifications. Are you suggesting WS-Security, WS-Trust, WS-SecureConversation, and WS-ReliableMessaging as those common, agreed upon, non-controversial specifications? Some of these are at Standard (WS-Security), some are at committee draft (WS-ReliableMessaging) and some of them are at editor drafts (WS-Trust, WS-SecureConversation). The GGF guidelines for OGSA Profile Definition (GFD-I.59) would require that a profile based on these specifications be classified as an Informational Profile (given the current state of the specifications). Further I would like to point out that the normative thorn in the side here continues to be WS-Policy. Thanks, Tom Mobile: +1-845-729-4806 ________________________________ From: owner-ogsa-wg(a)ggf.org [mailto:owner-ogsa-wg@ggf.org] On Behalf Of Marvin Theimer Sent: Thursday, February 16, 2006 8:45 AM To: ogsa-wg(a)ggf.org Cc: Marvin Theimer; Savas Parastatidis; Tony Hey; Marty Humphrey; gcf(a)grids.ucs.indiana.edu Subject: [ogsa-wg] Paper proposing "evolutionary vertical design efforts" Enclosed is a paper that advocates an additional set of activities that the authors believe that the OGSA working groups should engage in. Broadly speaking, the OGSA and related working groups are already doing a bunch of important things: * There is broad exploration of the big picture, including enumeration of use cases, taxonomy of areas, identification of research issues, etc. * There is work going on in each of the horizontal areas that have been identified, such as EMS, data services, etc. * There is working going around individual specifications, such as BES, JSDL, etc. Given that individual specifications are beginning to come to fruition, the authors believe it is time to also start defining "vertical profiles" that precisely describe how groups of individual specifications should be employed to implement specific use cases in an interoperable manner. The authors also believe that the process of defining these profiles offers an opportunity to "close the design loop" by relating the various on-going protocol and standards efforts back to the use cases in a very concrete manner. This provides an end-to-end setting in which to identify holes and issues that might require additional protocols and/or (incremental) changes to existing protocols. The paper introduces both the general notion of doing focused vertical "design efforts" and then focuses on a specific vertical design effort, namely a minimal HPC design. The paper derives a specific HPC design in a "first principles" manner since the authors believe that this increases the chances of identifying issues. As a consequence, existing specifications and the activities of existing working groups are not mentioned and this paper is not an attempt to actually define a specifications profile. Also, the absence of references to existing work is not meant to imply that such work is in any way irrelevant or inappropriate. The paper should be viewed as a first abstract attempt to propose a new kind of activity within OGSA. The expectation is that future open discussions and publications will explore the concrete details of such a proposal. This paper was recently sent to a few key individuals in order to get feedback from them before submitting it to the wider GGF community. Unfortunately that process took longer than intended and some members of the community may have already seen a copy of the paper without knowing the context within it was written. This email should hopefully dispel any misconceptions that may have occurred. For those people who will be around on for the F2F meetings on Friday, Marvin Theimer will be giving a talk on the contents of this paper at a time and place to be announced. Marvin Theimer, Savas Parastatidis, Tony Hey, Marty Humphrey, Geoffrey Fox

3 2

FW: UFCG Demo
by Milojicic, Dejan S (HP Labs) 21 Feb '06

21 Feb '06

Hi Hiro, You asked what was the percentage of the new functionality in the UFCG demo compared to Boston. Please find Ayla's answer. I cc OGSA because of some other participants in the meeting. Thanks, Dejan. ________________________________ From: Ayla Débora [mailto:ayladebora@gmail.com] Sent: Monday, February 13, 2006 6:52 PM To: Milojicic, Dejan S (HP Labs) Cc: Rafaeli, Sandro; Saikoski, Katia (HP Brazil) Subject: Re: UFCG Demo Hi Dejan, Regarding the demo questions, in this demo we have explored notifications, which was not explored in the other one. Besides that, we have also explored deployment in more than a single node, using hostname hints. We have also included in CDDLM Web some other functionalities from CDDLM like the WSRF ones, but I did not include in the demo to avoid making it take more time than that. The version from the demo is also more stable than the previous one and has incorporated adjustments due to CDL changes and new tests that have pointed out misunderstandings. Thanks, Ayla PS: As we are having problems with the UFCG mailing server since Friday, please use this email address to contact me when necessary. On 2/13/06, Milojicic, Dejan S (HP Labs) <dejan.milojicic(a)hp.com> wrote: Thanks Ayla, Demo went well, people liked it (some, in a positive manner suggested that you could have added voice; we did try to fill in the gap by talking to it, but we also did not want to take focus away from the demo itself). Hiro asked, percentage-wise, how much was the demo different from the one you gave in Boston. I answered that it was not much different functionality-wise, but that it was much different wrt stability and completeness. Do you have any different answer that I can provide to him? Thank, Dejan. ________________________________ From: Ayla Débora [mailto:ayladebora@gmail.com] Sent: Sunday, February 12, 2006 4:16 PM To: Milojicic, Dejan S (HP Labs) Subject: Re: UFCG Demo Hi Dejan, I'm glad you liked it. We are still planning the OVUA submission, but maybe I'll only be able to send you a version probably on Wednesday because we took more time than expected to make the architectural changes in our implementation to make paralelism work as expected and to prepare the demo. Thanks and have a nice time in Athens, Ayla On 2/12/06, Milojicic, Dejan S (HP Labs) <dejan.milojicic(a)hp.com> wrote: Hi Ayla, G ot it and ran it. It is excellent. Thanks, Dejan. PS Are you still planning the OVUA submission. If so, and if interested, you can send to me the version of the paper for review, although I might be busy with the conference. However, I would do my best to provide some feedback at minimum. ________________________________ From: Ayla Débora [mailto:ayladebora@gmail.com] Sent: Saturday, February 11, 2006 4:13 PM To: Milojicic, Dejan S (HP Labs) Subject: UFCG Demo Hi Dejan, sorry for the delay. I just finished now the presentation. Please let me know if you want me to make any changes to the text or if you have any doubts or suggestions. I've generated a flash presentation that can be downloaded from: http://www.lsd.ufcg.edu.br/~ayla/cddlm/ufcg_demo.zip <http://www.lsd.ufcg.edu.br/%7Eayla/cddlm/ufcg_demo.zip> . For the presentation, you will need a browser with a Flash plugin. Please confirm when you download the file. Thanks, Ayla

2 1

Paper proposing "evolutionary vertical design efforts"
by Marvin Theimer 20 Feb '06

20 Feb '06

Enclosed is a paper that advocates an additional set of activities that the authors believe that the OGSA working groups should engage in. Broadly speaking, the OGSA and related working groups are already doing a bunch of important things: * There is broad exploration of the big picture, including enumeration of use cases, taxonomy of areas, identification of research issues, etc. * There is work going on in each of the horizontal areas that have been identified, such as EMS, data services, etc. * There is working going around individual specifications, such as BES, JSDL, etc. Given that individual specifications are beginning to come to fruition, the authors believe it is time to also start defining "vertical profiles" that precisely describe how groups of individual specifications should be employed to implement specific use cases in an interoperable manner. The authors also believe that the process of defining these profiles offers an opportunity to "close the design loop" by relating the various on-going protocol and standards efforts back to the use cases in a very concrete manner. This provides an end-to-end setting in which to identify holes and issues that might require additional protocols and/or (incremental) changes to existing protocols. The paper introduces both the general notion of doing focused vertical "design efforts" and then focuses on a specific vertical design effort, namely a minimal HPC design. The paper derives a specific HPC design in a "first principles" manner since the authors believe that this increases the chances of identifying issues. As a consequence, existing specifications and the activities of existing working groups are not mentioned and this paper is not an attempt to actually define a specifications profile. Also, the absence of references to existing work is not meant to imply that such work is in any way irrelevant or inappropriate. The paper should be viewed as a first abstract attempt to propose a new kind of activity within OGSA. The expectation is that future open discussions and publications will explore the concrete details of such a proposal. This paper was recently sent to a few key individuals in order to get feedback from them before submitting it to the wider GGF community. Unfortunately that process took longer than intended and some members of the community may have already seen a copy of the paper without knowing the context within it was written. This email should hopefully dispel any misconceptions that may have occurred. For those people who will be around on for the F2F meetings on Friday, Marvin Theimer will be giving a talk on the contents of this paper at a time and place to be announced. Marvin Theimer, Savas Parastatidis, Tony Hey, Marty Humphrey, Geoffrey Fox

3 3

Re: [ogsa-wg] INFO a hole???
by Hiro Kishimoto 20 Feb '06

20 Feb '06

Dieter's email bounced. -- Hiro Kishimoto Subject: Re: [ogsa-wg] INFO a hole??? Date: Mon, 20 Feb 2006 01:38:03 -0800 From: Dieter Gawlick <dieter.gawlick(a)oracle.com> Andreas, Please go to the INFOD website and look at the GGF16 folder.There you will find specifications and use cases which include patterns. When you look at the INFOD presentation at GGF16 you we see that the group plans a major cleanup of the interfaces in two months time. I think that OGSA has to look much more closely into the implications moving towards a /push world/. Some of the basics are well understood, e.g., those aspects captured in WS-Notification. But there is much, much more. For starters, I would highly recommend to look closely into the INFOD mechanism of matching publishers and consumers. It is quite sophisticated and seems to capture some difficult tasks that we tend to leave to developer as an implementation exercise. The current specifications reflect only step one of INFOD; OGSA members should also try to become familiar with what we intend to add later. Doing so, could provide significant guidance for some areas, i.e., data movement and transactions. Dieter Andreas Savva wrote: >>>>The INFOD spec has been simplified significantly - or rather we have >>>>decided to concentrate on a base specification before moving on to >>>>more advanced features. We plan to have the base spec going for public >>>> >>>> > >> >>Could you also post the reference to the INFOD specification(s)? Also >>references to any other material describing the more advanced features >>you expect to work on in the future would be nice. >> >>Abdeslem in his revision to the information services section for OGSA >>1.5 added mentions of INFOD and I would like to have a reference to a >>specification, even if an early working group draft. I would also like >>to understand better the potential role of INFOD. >> >>-- >>Andreas Savva >> >>Steve Fisher wrote: >> >> > >>>>Hi, >>>> >>>>I have heard a number of people complaining this week, at GGF, that >>>>"Information is a hole". It was also on the slides shown by Dave >>>>Snelling on Thursday morning. >>>> >>>>Previously Abdeslem Djaoui liaised between the INFOD-WG and OGSA-WG >>>>but now we have no direct means of communication and I suspect that >>>>some people in OGSA have not completely understood the potential role >>>>of INFOD, nor is INFOD sure of the size or shape of the hole it might >>>>fill within OGSA. >>>> >>>>The INFOD spec has been simplified significantly - or rather we have >>>>decided to concentrate on a base specification before moving on to >>>>more advanced features. We plan to have the base spec going for public >>>>comment in June, but it would obviously be better to understand very >>>>soon how it fits in to the big picture - to avoid having a >>>>specification which does not meet the perceived OGSA needs. How do you >>>>suggest we move forward? >>>> >>>>Steve >>>> >>>> >>>>

1 0

INFO a hole???
by Steve Fisher 20 Feb '06

20 Feb '06

Hi, I have heard a number of people complaining this week, at GGF, that "Information is a hole". It was also on the slides shown by Dave Snelling on Thursday morning. Previously Abdeslem Djaoui liaised between the INFOD-WG and OGSA-WG but now we have no direct means of communication and I suspect that some people in OGSA have not completely understood the potential role of INFOD, nor is INFOD sure of the size or shape of the hole it might fill within OGSA. The INFOD spec has been simplified significantly - or rather we have decided to concentrate on a base specification before moving on to more advanced features. We plan to have the base spec going for public comment in June, but it would obviously be better to understand very soon how it fits in to the big picture - to avoid having a specification which does not meet the perceived OGSA needs. How do you suggest we move forward? Steve

3 2

No monday call on Feb. 20th
by Hiro Kishimoto 19 Feb '06

19 Feb '06

Hi all, Thank you very much for all of you who came to GGF16 and our F2F meeting in Achenes. We have had very lively and productive meetings. Since this Monday is US national holiday, we won't have conf call on Feb. 20th. We will have Wednesday call and Tom and Andreas will chair this call. Thanks again, -- Hiro Kishimoto

1 0

RE: [ogsa-wg] INFO a hole???
by Fred Maciel 17 Feb '06

17 Feb '06

Hi All, > > I have heard a number of people complaining this week, at GGF, that > > "Information is a hole". It was also on the slides shown by Dave > > Snelling on Thursday morning. Hmm, I think that the total "number of people complaining" was me, Dr. Maciel, Frederico, and Mr. Buchholz. (Read the signature in the end of this e-mail or else you won't get the joke :-). I have used the strong expression "black hole" to say that the information services part of OGSA is very important, concrete work on it is more and more needed, but there is nobody working on it. I was not complaining (sorry if I sent the wrong nuance), but I've been quite vocal about it since the lack of work on this important area of OGSA is making me uneasy. >From Hiro: > We believe we do not underestimate role and importance of INFOD. > However, from our OGSA architecture point of view, information > service should also cover higher level services, e.g. discovery, > directory, and registory which will use INFOD for their > transport means. Yes, all this with a *lot* of scalability -- think 10 thousand nodes. I remember Abdeslem saying that there are a number of existing systems and a lot of experience on this area, and the OGSA-WG needs somebody to bring us this experience. (I have tried to follow the INFOD work, but clashes in GGF sessions and lack of bandwidth have prevented me from doing it.) > We are more than happy to talk with INFOD folks on how to fit > your work > in OGSA architecture. We would very much appreciate if you or your > colleage could take care of these higher level information services in > addtion to INFOD? Ditto. Regards, Dr. Frederico Buchholz Maciel Hitachi America R&D

1 0

Friday F2F logistics
by Hiro Kishimoto 16 Feb '06

16 Feb '06

Hi all, The following is the logistics information for OGSA-WG F2F meeting. Location: National Technical University of Athens Zografou Campus, Iroon Polytexneiou 9, 15780 Zografou Multimedia Ampitheatre, the library building ground floor February 17, Friday 9am-5pm Direction from Divani Caravel Hotel: We will meet at hotel lobby 8:30am and take taxi to NTUA. It will take 20-25 minutes depending on traffic. In case you prefer to use public transportation (it will take 30-45 minutes): From “Evangelismos” station near by hotel, take line 3 metro to “Katehaki” station (4th stop). From “Katehaki” station, take bus number 242 to NTUA Administration Building. Bus departs 15 and 45 every hour from Metro station. Library building is the next to the Administration Building. The campus map and information on getting there. http://www.ntua.gr/en_ntua/campus.htm This is an external view of the building where the room is. http://www.ntua.gr/en_ntua/pages/empn02_jpg.htm Proposed meeting agenda is uploaded to the GridForge: https://forge.gridforum.org/projects/ogsa-wg/document/GGF16-B2B-F2F-schedul… Thanks, -- Hiro Kishimoto

1 1

Dial in information change
by Hiro Kishimoto 15 Feb '06

15 Feb '06

Hi all, Since GGF17 meeting room needs special set-up for conference call, we need to change dial-in info for Wednesday call. Time and date are the same, but please use the following dial-in: Toll number: +1 718 3541071 (New York) +1 408 9616509 (San Jose) +81 (0)3 3570 8225 (Tokyo) +44 (0)207 3655269 (London) The pass code: 4371991 Thanks and talk to you soon, ----- Hiro Kishimoto

2 1