AG-SAG - Full Meeting #1

IoA, Cambridge, 19 June 2003, 10.00-16.00

Supporting Paperwork:

Meeting minutes prepared by NicholasWalton - 20 Aug 2003

Present:

• Carlton Baugh (CB)
• William Chaplin (WC)
• Andrew Jaffe (AJ)
• Daniel Lennon (DL)
• Steve Maddox (SM)
• Richard McMahon (RM)
• Keith Noddle (In attandence) (KN)
• Nial Tanvir (NT)
• Nicholas Walton (NW)
• Jeremy Yates (JY)

Apologies:

• Lyndsay Fletcher
• Shadia Habbal
• Tom Marsh

0. Summary

This kick-off meeting was used to inform the AGSAG of the current status of the project, and how the key AstroGrid science drivers are used to scope the product deliverables from AstroGrid. To this end Nic Walton described the science drivers and how these will be delivered, noting the increasing levels of capabilities that AstroGrid will offer at each successive iteration referenced against each driver. Keith Noddle updated the group on the process that AstroGrid has adopted in delivering the project.

The AGSAG stressed the importance of demonstration products from AstroGrid, to enable science users to gain early exposure to capabilities, useful for helping in carrying out ?real? science. The AGSAG noted that early inclusion of the most scientifically exciting data available to the UK was also important, thus including access to for instance ESO data sets. The next meeting of the AGSAG will be held before the TSP Iteration 4 meeting at the beginning of September 2003.

1. Welcome

NW welcomed the group to the kickoff meeting of the AstroGrid Science Advisory Group (AGSAG). NW noted that the AGSAG would play a vital role in providing additional scientific input into the AstroGrid development process.

2. AGSAG Terms of Reference

The Terms of Reference were read and AGREED.

The decision on the Chair of the group was deferred to the next meeting.

3. AstroGrid Project Overview

NW presented a brief overview of the current status of the project, and how it was responding to scientific priorities, as determined during the science requirements capture process. [slides: pdf]

NT wondered how AstroGrid was integrated into the more Global Grid world. NW responded that AstroGrid was closely integrated into the UK eScience programme (see the news story for AstroGrid's involvement at the upcoming eScience All Hands 2003 meeting). It was also active - through the International Virtual Observatory Alliance - in setting up a Astronomy Research Group at the Global Grid Forum. (GGF is the body which develops and sets standards across the grid world). Thus AstroGrid was seeking to deploy 'grid' components developed elsewhere if possible (for instance underlying access, job monitoring, etc components), and only develop or adapt new components if specifically needed. Further it was engaged with the relevant groups in ensuring requirements from the Astronomy community were being taken into account in the 'grid' development process.

4. AstroGrid Project Organisation.

KN presented the overview of the AstroGrid project and organisation, indicating how the project was following the iterative process. [slides: ppt]

NW noted how the even numbered iterations would aim to meet the science milestones derived from the key ScienceProblems. The odd numbered iterations woul dbe of a more technical nature, aimed at adding to the technical capabilities of the AstroGrid system.

KN stated that AstroGrid would appreciate feedback on which data sources were felt to be of early importance for inclusion through the AstroGrid System. This information could be entered in the Forum or BetaTesting wiki pages.

5. Current Deliverables available through Iteration 2.

KN noted that the AstroGrid deliverables at each release were available from the SoftwareReleases pages. To date the software is all server side, so the software available for download is mainly of interest for system administrators and/or developers who might wish to deploy AstroGrid infrastructure.

NT asked how much of the AstroGrid infrastructure from Itn02 would be of use longer term, and by other projects? KN noted that Itn02 was a foundation release and that in principle future iterations would build and expand on all the software components developed and released in Itn02. Further, it was anticipated that other VO projects would be able to deploy all or part of the AstroGrid components when setting up their VO's. For instance, a VO project might just deploy the MySpace component to allow for individual grid based user space management.

6. Current Deliverables available in conjunction with external partners.

NW noted that AstroGrid was also deploying deliverables made available through partnerships with the AVO and other VO groups.

Currently the main deliverables are:

• the AVO 1st light demonstrator product, this allows access and manipulation of GOODS data. It is available from the AVO software download pages at http://www.euro-vo.org/twiki/bin/view/Avo/SwgDownload
• in partnership with the Aus-VO, AstroGrid developed a grid access to a volume rendering system - see http://astrogrid.ast.cam.ac.uk:8080/aad-portal/index.jsp for access and test.
• Software to handle VOTable format data - e.g. VOPlot - see http://vo.iucaa.ernet.in/~voi/VOPlot1_1UserGuide/VOPlot_UserGuide_Version1_1.htm

NW noted that the NVO was also generating early capability software - see their early demonstrator prototypes at http://www.us-vo.org/prototypes.html

AJ asked how was AstroGrid related to the AVO and the IVOA.

NW replied that AVO was a founding member of the AVO, a major partner and responsible for the 'Technology' work area - see http://www.euro-vo.org/twiki/bin/view/Avo/WorkAreaThree.

The IVOA - see http://www.ivoa.net - is a standards organization with members from the major VO project s worldwide. Again, AstroGrid was a founder member along with the AVO and the NVO. The IVOA is crucial in ensuring that underlying standards for the astronomy community are devloped and agreed and implemented globally. The most recent meeting making progress was held in May 2003 at the IoA - see http://www.ivoa.net/twiki/bin/view/IVOA/InterOpMay2003

AJ asked what were the main differences between Web Services and Grid Services. KN replied that Grid Services can be thought of as Web Services with a number of extra features allowing for state and security for instance. http://www.ws-i.org is a good entry point into issues around web services interoperability, whilst http://www.ggf.org is the home page of the Global Grid Forum.

7. Upcoming deliverables: The Summer 03 Demo Products

NW reported that AstroGrid would be exhibiting at the IAU General Assembly held in Sydney, July 2003. It would also be at the upcoming All hands meeting of the UK eScience Programme in September 2003.

8. The AstroGrid Beta Tester Programme

NW noted that with the successive release of AstroGrid capabilities, with each Iteration, the community would be encouraged to use and comment on the system. The BetaTesting pages will provide information about the testing process. To date a number of researchers have indicated a willingness to participate in the programme. All members of the AGSAG are asked to participate in this as well, and perhaps ask colleagues who might also be interested. It is anticipated that the Beta Test community will increase as the capabilities offered by AstroGrid expand with future iteration releases.

9. AGSAG: Future Priorities

NW indicated that the major priorities of the AGSAG are outlined in the AG SAG terms of references. In the short term the AGSAG could comment on usability of the system, priorities on data set inclusion, directions for future AVO development,

10. Discussion and AOB

AJ asked as to progress with the AstroGrid2 proposal. NW noted that the initial review process was underway with a Project Peer Review Panel Meeting to take place on 14 July 2003. The proposal is available AG2. The AstroGrid2 proposal aimed to ensure the widespread rollout and deployment of AstroGrid system across the UK.

BC asked how does the Solar Systems Virtual Observatory proposal (on line at SSVO) fit in with the AG2 proposal. NW noted that the SSVO proposal aimed to integrate solar, STP and planetary datasets into a VO for those communities, building upon the core infrastructure that AstroGrid was currently delivering. In principle, future infrastructure requirements coming from the SSVO would be fed into the AG2 process via the SSVO representative on the AG2 Technical Support Panel.

NT commented that usable end-to-end capabilities should be made available through AstroGrid to enable at least some group of researchers a worthwhile aid in carrying out there research. It was vital that some research could be shown to benefit from AstroGrid as soon as possible.

11. Date of Next meeting.

It was AGREED that the next meeting would be held in Birmingham, Tuesday, 9 September 2003 from 10-16. BC would be our host.

The date was agreed here because the meeting needed to be before the Iteration 4 Technical Support Panel meeting of September 11 (so as to ensure science input into the planning process for the next 'science focused' iteration). With holidays etc, dates were limited. In future the meetings would be held before even numbered TSP meetings, thus every 6 months. All members of the SAG would be polled before agreeing the date.

-- NicholasWalton - 19 Jun 2003