May 2006 - saga-rg - lists.ogf.org

Minutes from 1. session of SAGA-RG at GGF17
by Thilo Kielmann 17 May '06

17 May '06

Meeting of SAGA-RG at GGF17 1st meeting, may 10, 1:45 - 3:15 Session chaired by Shantenu Jha Notes taken by Andre Merzky, Thilo Kielmann Agenda: - agenda bashing - group progress - requirements document Shantenu opens the session, circulates the sign-up sheet, referring to GGF's IPR pollicy. agenda bashing: --------------- the agenda is accepted "as is" by the participants group progress: --------------- SAGA-RG's Use case document has been published as GFD.70 SAGA-RG's Requirements documents is now in public comments period requirements document: ---------------------- Shantenu is summarizing the findings of the requirements document, asking for comments and proposals to approach the missing pieces of the intended scope of SAGA 1.0. - monitoring: - comment: globus is giving insufficient info on job status - answer: is backend problem, not API problem - comment: what do you mean by steering - answer: track and change parameters of running applications - RPC: - not covered, but needed, in work -> have a design team - bulk: - comment: what means bulk - more than 2? - answer: bulk is just optimization, and API should not hinder optimized implementations - comment: replace 'saga version 2' with 'future saga versions' - resource discovery/information - comment (Michael Russell): Res. Discovery crucial for portal centered use cases, and probably for interactive usage in general - comment (Dieter, others): why handle security at all, if no general paradigm can be defined/found? - answer: why indeed :-) That is why we limit ourself to minimal approach (context), and use only paradigms as available and acknowledged (ACL) - CPR: - no resources to advance docs/API - want/need feedback from API - take offline time in Tokyo and draft API - circulate docs as are on the SAGA list - prioritize - 5 items identified, with no clear priority - RPC (rpc-wg) - CPR (Derek, Thilo, Shantenu) - InfoService (Andre) - Message (Andrei, Andre) - resource discovery (simple) At 3:15 the session is closed.

1 0

Fwd (dk@gup.jku.at): REMINDER: AppsAreaWorkshop @ GGF17 [TOMORROW, Thursday, May 11, 09:00 G409]
by Andre Merzky 10 May '06

10 May '06

FYI, Andre. ----- Forwarded message from Dieter Kranzlmueller <dk(a)gup.jku.at> ----- > From: Dieter Kranzlmueller <dk(a)gup.jku.at> > To: Steven Newhouse <s.newhouse(a)omii.ac.uk>, Andre Merzky <merzky(a)cs.vu.nl>, > hide-nakada(a)aist.go.jp, Thilo Kielmann <kielmann(a)cs.vu.nl>, > moore(a)sdsc.edu, aoyagi(a)cc.kyushu-u.ac.jp, csmith(a)platform.com > Subject: REMINDER: AppsAreaWorkshop @ GGF17 [TOMORROW, Thursday, May 11, 09:00 G409] > Date: Wed, 10 May 2006 10:05:47 +0200 > > > Dear Speakers, > > First of all, thanks for agreeing to give a presentation at the > Applications Workshop during GGF17 tomorrow. > > Please find attached some last minute information including the below. > Please help us distribute it to the respective interested parties. > > Looking forward to an interesting workshop, > Dieter > > ---- > > GGF-17 Applications Workshop > (Tokyo, Japan) > > "Bridging the Divide: > Community Application Requirements Driving Standards Development" > > Session 1: "Current Status - Development of Standard APIs" > Chair: Steven Newhouse > > Thursday, May 11, 2006, 09:00-10:30 G409 > > * Welcome, Motivation & Goals (Steven Newhouse) [10 min] > * DRMAA - Distributed Resource Mgnt App API (Andre Merzky) [20 min] > * GridRPC - Grid Remote Procedure Call (Hidemoto Nakada) [20 min] > * SAGA - Simple API for Grid Applications (Thilo Kielmann) [20 min] > * Discussion [20 min] > > Session 2: "Consuming Standards - Current Uses and Future Requirements" > Chair: Dieter Kranzlmüller > > Thursday, May 11, 2006, 11:00-12:30 G409 > > * Summary of Session 1 and Intro to Session 2 (Dieter Kranzlmüller) [10 min] > * Astronomical Grid Community (Reagan Moore) [20 min] > * NAREGI - National Research Grid Initiative (Mutsumi Aoyagi) [20 min] > * Grid Application Integration (Chris Smith) [20 min] > * Discussion [20 min] > > Session Summary: > > This workshop will provide a showcase of standards and community activity > within GGF by bringing together groups in three different areas (Standards > Applications, Community Industry Applications & Community Research > Applications). Through a series of presentations we will show the current > state of the standards activity and critically the current barriers to > pervasive adoption as perceived by the consuming user community. Finally, a > panel session will help to define areas of collaboration and potentially new > areas for standards development. > > Session Description: > > GGF has an active standards and community areas focusing on engagement with > application developers and tool builders in research and industry. For these > separate areas to work effectively we need to build a relationship where the > APIs being developed within the standards area are relevant to the research > and industry consumers, and in turn the application and tool builders are > communicating to the standards area their requirements in terms of stable > defined interfaces. > > With the greater emphasis on GGF's redefined standards and community > functions, and the factorization of the standards area into having an > 'umbrella' research group to drive the coordinated development of a set of > consistent APIs across several working groups, there is a need to establish > the communication links between the standards and community areas for > Applications. > > This workshop, hopefully the first of many, will start to establish the > communication links and provide a showcase as to the current state of the > standards development and the current priorities for application and tool > development in industry and research. In addition, it will provide a general > dissemination mechanism of activity to developers and users (both critical > consumers in terms of achieving pervasive grid adoption) and the wider GGF > community as to activity in the standards and community areas. > > Session Leaders: > Dieter Kranzlmueller, Steven Newhouse, Thilo Kielmann, Craig Lee ----- End forwarded message ----- -- "So much time, so little to do..." -- Garfield

1 0

Re: ISSUE 131
by Andre Merzky 08 May '06

08 May '06

Hi Graeme, Pascal, forgot CC to list - A. Quoting [Andre Merzky] (May 08 2006): > > I checked all usages of NULL in the spec, and all but two > can be safely removed, and replaced by empty strings etc. > > The two places are: > > 1) wait on a stream_server returns a stream if > successfull, and NULL otherwise. I don't see any > nice way to get rid of that NULL - I think that > should be dealt with in the language bindings. > > 2) remove_callback on a metric returns the callback > removed. If that cb was removed earlier (or removed > itself after execution), the method returns NULL. > Same here I think: check in the language bindings. > > Does that sound ok? > > Cheers, Andre. > > > Quoting [Andre Merzky] (Apr 27 2006): > > Date: Thu, 27 Apr 2006 21:52:23 +0200 > > From: Andre Merzky <andre(a)merzky.net> > > To: G.E.POUND(a)soton.ac.uk > > Cc: Andre Merzky <andre(a)merzky.net> > > Subject: Re: ISSUE 131 > > > > Ah, ok - I retract the Java argument then :-) > > > > Thanks, Andre. > > > > > > Quoting [G.E.POUND(a)soton.ac.uk] (Apr 27 2006): > > > > > > Andre > > > > > > 'null' is a valid value in Java (in most instances). It is not a valid value > > > in FORTRAN. > > > > > > Graeme > > > > > > > > > > > > Quoting Andre Merzky <andre(a)merzky.net>: > > > > > > > Quoting [Andre Merzky] (Apr 20 2006): > > > > > > > > > > 131) clearify what a NULL string/object is we sometimes return > > > > > - OPEN, URGENT > > > > > > > > > > iObvious again. Opinions? Should that even go into the > > > > > API spec, or into the language bindings? What should we say > > > > > in the spec? > > > > > > > > No opinion. > > > > > > > > I think NULL does not make sense for Java and Fortran (Is > > > > that true?) so I guess it should stay out of the general > > > > spec. > > > > > > > > I changed all occurences to either throw, or empty string > > > > returns etc. There are only a couple of places where we > > > > used NULL anyway. > > > > > > > > > Volonteers? > > > > > > > > > > Cheers, Andre > > > > > > > > Cheers, Andre. > > > > > > > > -- > > > > "So much time, so little to do..." -- Garfield -- "So much time, so little to do..." -- Garfield

1 0

ISSUE 130
by Andre Merzky 08 May '06

08 May '06

130) clearify the error handler interface for async calls - OPEN, URGENT I am unsure how the error handler interface applies to asynchroneous calls. Opinions? Tom, did you consider this? Cheers, Andre. -- "So much time, so little to do..." -- Garfield

1 1

ISSUE 88
by Andre Merzky 08 May '06

08 May '06

88) for job: add clean_up (can also be called after termination) - OPEN, URGENT we got that request out of bound, to allow to remove old output files from past jobs. The request comes from medical guys IIRC, where data privacy is very important. Does anybody know how backend systems handle data cleanup? Should we add that method? Cheers, Andre. -- "So much time, so little to do..." -- Garfield

1 1

Extensible state model for jobs
by Christopher Smith 04 May '06

04 May '06

Hi all, As Marvin mentioned in his email on extensibility and composition of protocols, I'd like to present this short document describing, in more detail, a model for extending the state model of jobs from a "base profile". The document is based on some discussions that Marvin and I have been having about scheduler interoperability, and is very close to the model that has been proposed within the SAGA working group. Please review Marvin's email about extensibility to get a feel for the "context" of the model. It is my intention to show how the ESI and BES state models can be represented using this approach. I'll try to get that done before Tokyo, but might just need to present some slides at GGF17, assuming I'm given a few minutes to do so. This is very much a work in progress, so I'm looking forward to feedback.... A note for the SAGA group .... I'd still like to see "pending" in the base job state diagram ... don't know why it disappeared. ;-) -- Chris

3 8

Satoshi M.'s Japan FAQ
by Thilo Kielmann 04 May '06

04 May '06

Upon request from today's telecon, I'll attach Satoshi Matsuoka's Japan FAQ from GGF7. Please be aware that it is 3 years old and that the local references to the meeting place (different part of Tokyo) don't match. Nevertheless, it has a tremendous lot of handy information for "the rest of us" trying to find our way in Tokyo... See you next week, Thilo -- Thilo Kielmann http://www.cs.vu.nl/~kielmann/

1 0

tokyo...
by Andre Merzky 04 May '06

04 May '06

Hi all, here are the most important notes from todays phone call: - present: Stephan, Thilo, Andre, Shantenu - agreed to meet Wednesday 8:30 at reception for last minute planning etc - power outlets in Tokyo SHOULD be compatible to US plugs - the agenda in compressed form can be found on http://www.ggf.org/gf/session_request/index.php?event=4&browse&order=12 (nice for printout) - session agenda items should be redistributed - discuss on mailing list Cheers, Andre. -- "So much time, so little to do..." -- Garfield

1 0

phone con details
by Andre Merzky 04 May '06

04 May '06

Hi all, Here are the details for tomorrows call: Date: 04 May 2006 Time: 2:00 pm London 3:00 pm Amsterdam 8:00 am Baton Rouge 10:00 pm Tokyo Number: 0844.888.8888 (UK) 01805.004.102 (Germany) +44.870.088.5706 (Others) Code 808044 Agenda: - spec editing - spec issues - ggf17 planning - AOB Cheers, Andre. -- "So much time, so little to do..." -- Garfield

1 0

Fwd (matsu@is.titech.ac.jp): [gf-chairs] Getting around TIF (was: [gfsg] Chair Appreciation event at GGF17)
by Andre Merzky 03 May '06

03 May '06

FYI, Andre. ----- Forwarded message from Satoshi Matsuoka <matsu(a)is.titech.ac.jp> ----- > Folks, > > Satoshi (M) here. Here is the guide on navigating around the conference > site. Scott may have more detailed update on my older document for GGF7, > much of which is still useful but does not cover the regions around TIF. > > Note that some of the information are more applicable to many of the > GFSG members who are staying in the Shiodome (Eastern Shinbashi) area, > esp. the hotel Villa Fontaine. However, feel free to distribute it to > your colleagues as you feel fit. > > Every morning except Friday, I shall be picking up folks at the lobby of > Villa Fontaine Shiodome around 8AM. > > *** General Directions: > > When you arrive at the NARITA airport, NEVER TAKE A TAXI, as it will > cost over $200 US. The definitive way to get to Tokyo - Yuurakucho - > Shinbashi area is to use the JR Narita Express, and get off at Tokyo > Station, and then taka a local train, a taxi, or simply walk. I also > recommend NOT to take a shuttle bus, since NARITA express is much more > convenience and wont get you stuck in the rushhour traffic. > > As it was in my older guide, Narita Express is one of the two train > company lines (JR and Keisei co.) that leave adjacent to each other from > the underground stations at the Narita airport, both Terminals 1 and 2. > The other line is Keisei Skyliner, which is cheaper and will make more > sense if you are going to North of Tokyo, but not in this cae. There are > various local, non-express trains that are much cheaper but unless you > have been to Japan several times and know the Tokyo geography let's > leave them out unless you are adventurous. > > The first stop for the Narita Express is Tokyo Station, about a > 50-55 minute ride. It arrives at the underground platform of Tokyo > Station, 4 levels deep (very deep). BTW, most of the trains at Tokyo St., > including the Shinkansen, leave from above ground-level platforms, so in > most cases you have to find your way up to make your connection. > > For details see: > http://www.jreast.co.jp/e/estation/e_tokyo.html > Narita express arrives at the Sobu line platforms shown at the bottom. > > For JR train maps > http://www.jreast.co.jp/e/info/map_a4ol.pdf > > Once you get off, if you are staying at Yaesu Fujiya, find your way > towards the surface and walk to the hotel (outside of the top Yaesu > Central Exit). It will be a fairly long walk from the platform (about > 15-20 min), but you cant shortcut it much with a taxi since you will be > making your way cutting through a huge station, and getting to a taxi > stand is already a 5-10 min walk unless you know your way. > > If you are staying near Yuurakucho/Ginza, take the option 2 below. > > If you are staying near Shinbashi/Shiodome, there are two options: > > Option 1 (recommended). The Southbound JR Yokosuka line leaves from the > same level as the > Narita express, and it's next stop is Shinbashi. One caveat is that > there are two platforms / 4 tracks, and sometimes it leaves from a track > on the other one so you have to climb up one level and come down again. > This is a good strategy in any case since the # of lines of display on > signboard on the platform is fairly limited, whereas if you go up one > level there is a more comprehensive one. Or, you can ask a station > master walking around in their uniforms. > > At Shinbashi a few minutes later you will again arrive at a similar > underground platform. > Go up the escalators, and find your way towards Shiodome; there will be > bunch of signs there. If it is not raining it may be easier to head to > the surface and look for a Skyscraper cluster and what seemingly is an > elevator walkway leading to them. Also, you could stick to the > underground corridor which also is fairly easy as there will be lots of > maps once you get to the corridors. > > Option 2. From the Narita Express platform climb up a series of > escalators to arrive at one of the three main concourses. The Southbound > Yamanote and Keihin-Tohoku lines are platforms 5 and 6 (towards > Shinagawa). It will take about 5 mins to reach there. Hop on either of > them and the next stop about 800meter away is Yuurakucho, and Shinbashi > is a stop after that. During midday around 10:00-15:00 Keihin Tohoku > becomes an express and will not stop at Yuurakucho or Shinbashi. Look at > the electronic signboard and also the side marking on the train and if > it has the red express mark dont ride on that. In case you end up in > Hamamatsucho I have already outlined below for Villa Fontaine. > > At Shinbashi, if you are heading towards Shiodome, there are two > staircases. Climb down the staircase with a train connection sign to > pink (Asakusa-line) subway and Yuri-Kamome (Skytrain) and Shiodome. If > you see an orange (Ginza line) it is the wrong one. Again, if it is > not raining, it may be easier to stay above ground by re-climing the > escalators to the Yuri-Kamome (skytrain) corridors then going > underground and possibly losing the sense of your directions (although > there are lots of signs, and huge number of people walking around so > you could ask). > > For Dai-ichi Hotel the situation is opposite. It is fairly a short walk > from the Shinbashi station thru a busy shopping/nightlife district, > which will seem like a different world c.f. Shiodome which is clean and > futuristic. > > > *** Villa Fontainte > > To Akihabara from Villa Fontaine Shiodome: > (Notice that there are several Villa Fontaines in Tokyo. THe one where > people are staying at is the one in Shiodome (shee-oh-doh-meh)). > > From the hotel exit walk to JR Shinbashi station (not raining then take > the escalator up and the Yuri-kamome skytrain corridors (passing the > Subway sandwich stand on the way) above ground - 6 min, if raining then > take the escalator down and the underground passageway - 8 min). JR > Yamanote (olive encircling commuter) or the Keihin-Tohoku (blue > North-South commuterline), Northbound (towards Tokyo St.), 4 stops > Shinbashi-> Yuurakucho->Tokyo->Kanda->Akihabara. Akihabara Dai- > Building is right in front of the Akihabara St. The journey is 20 > minutes tops door-to-door. > > Both Yamanote and Keihin-Tohoku lines arrive at the same > platform, on both sides, and thus collectively they arrive every 1-2 > mins. There is a Starbucks both above and below ground. > > Hiro indicatetes that they have a new Starbucks in Akihabara in a > building next to the Dai Bldg. (they have > the competitor Excelisor Coffee in the Dai bldg as well as others). > > During midday around 10AM-3PM only the Yamanote line stops at > Shinbashi. So, if you get on the Keihin Tohoku line at Akihabara, and it does > not stop at Shinbashi, dont panic. Simply get off at Hamamatsucho > (the next stop after Shinbashi), head towards the rear of the train and > you see a descending staircase. Climb down and get out of the gate > there, cross > the street right in front of you under the train tracks, turn > right then immediately left, ending up in the street that parallels the > train tracks to the East (train tracks to your left). Simply walk north > towards the skyscraper cluster parallel to the tracks for about 6-7 > mins and you will arrive at the Villa Fontaine Shiodome. > > Attached is the picture taken from South i.e., in the event that > happens (BTW, Shinbashi is further North). Villa Fountaine Shiodome is > the bottom half of the black building you see in the center. The > triangular building to the left is the Park Hotel, while the light-brown > one is the Royal Park Hotel. The picture BTW was taken from a skytrain > you will also see in the area (besides the JR, the subways, the monorail > etc.) > > ---- > > Walking to Tokyo Int'l Forum from Villa Fontaine > > Although people should take the train from Shinbashi (130 yens) in which > the directions will be identical to above but just one stop, or Taxi > directly from Villa Fontaine (about 800 yens) if they are in a > hurry, you might walk through the plush shopping area of Ginza if you > have time. > > Take the escalator and go below ground, walk the corridor towards > JR Shinbashi past the Shiodome subway station. In about 200m you > arrive at a underground crossing which is like a big underground square > surrounded by the basins of several skyscrapers. Ahead and left > you will see the large revolving door of the Shiodome Center Building > (which by the way 1/2 of which is Fujitsu HQ)---there are chairs, a > bakery "La vie de France". If you see McDonalds or Kinkos they are > wrong buidings. Enter the building at the entrace where nail salon is to > the right and ANA travel agency is to the left. Go up the escalators > for a couple of floors until you see a > Starbucks to your left. Exit the main building gate and you will see a > small railroad museium (Shinbashi is where Japan's railway system > initiated). Take the lefthand corridor with the Center building to your > left and musieum to your righ for about 100ms passing an English pub > until you see a Porche showroom to your left. > Go a little past and you see a big road crossing. Up to here should be > less than 5 mins. > > The crossing you see is the leftmost-center big crossing on the > map (that is bit chopped off) with Ginza-Dori, and you are at the > left-bottom corner: > > http://www.ginza.jp/eng/mapindex-e.html > > You will want to proceed North thru Ginza-Dori, and enjoy the most > affluent shopping section in Tokyo. Eventually arbitrarily head East > Towards the North-East corner of Yuurakucho St. where the TIF is. > Once you are in mid-Ginza there is no way to miss TIF as it is so > visible. > > *** Chairs Appreciation night location info. (Just walk from TIF) > > BTW, Gon-pachi for the Chairs Appreciation night is under the highway > along the Ginza Sakuramachi Dori you see on the rightmost area of the > map, a block up from the X sign (police box) Kyobashi Br. As such it is > only a few hundred meters from TIF so there is no use in riding a subway > if you are already at TIF. > > ----------------------------------------- > > Comments and suggestions are welcome. > > Satoshi (M) ----- End forwarded message ----- -- "So much time, so little to do..." -- Garfield

1 0