 |
Workers: Proposal |
FusionGrid: An Enhanced Collaborative Environment for Magnetic Fusion Research
Our NFC Project renewal proposal is due January 30, 2004.
- Check if you still need to send me (DS) something
- Schedule for Proposal Preparation
- Dec 5: requirements posted on Web site. Phone call to discuss.
- Dec 19: Tasks lists posted on the web. Phone call to discuss.
- Jan 6: First draft of words for proposal due and posted on web. Phone call to discuss.
- Jan 10: 11am PST Teleconference.
- Jan 12, 13: Meeting in Chicago to finalize proposal technical words
and finalize budgets. People should come to this meeting
having read and commented on other's words posted on January 6.
Meeting information can be found here.
- Jan 15: Bios (2 page maximum) due to Schissel.
- Jan 15: Facility Description due to Schissel.
- Jan 15: 20-page technical narrative distributed as whole
- Jan 15: Draft figures released
- Jan 16: Technical narrative distributed with figures
- Jan 19: Technical narrative close to final form
- Jan 26: Face Pages and Budget Pages due to Schissel
- Jan 28: Letters in by this date to Schissel.
- Jan 28: Electronic Proposal assembled.
- Jan 29: Proposal submitted electronically (one day early for safety).
- Jan 30: Proposal due in D.C.
Writen Proposal
Proposal Outline
User Requirements
Below is a detailed list of User Requirements compiled over the lifetime of the
NFC Project. Goals and Task Lists should address these requirements.
- Authentication
- Authorization, Resource Management, & Enforcement
- Firewalls
- Education & Documentation
- Large Scale Data Management
- New Computational Services
- Collaborative Control Room
- Collaborative Meeting Room
- Advanced Visualization
People Respond to List of Requirements
In bullet form, people respond to the list of requirements
with items they want to work on, rough dates and deliverables,
and a rough estimate of cost/level of effort.
The information below is
being provided as background for when we write the proposal.
Comments of NFC Review Committee
The 3 viewgraph debriefing report is available
for review. At this time the NFC project has not received any further information
on the committee report. A brief summary of those viewgraphs is listed below.
Things you're doing well
- Impressive results; Scope of project is impressive; there is effective leveraging
- Good outreach to Fusion community interest is building
- Grid-based TRANSP is a tremendous success a real impact!
- Many interesting technologies are being developed
- Good fusion user feedback
- Driving policy-based management
- Balanced in your choice of grid technologies
- Driving the Access Grid to more accessiblity (PIG/laptop) is good
- Team is committed
- Leadership role in the real-time remote participation
Concerns
- Subgroups are quite independent - clouding the external view of NFC as a coherent whole
- Project appears to be spread too thin - scope appears larger than budget
- Some goals in terms of level of functionality required are nebulous, leading to inabilities to prioritize efforts
- Planning lacks a sufficiently clear set of milestones, deliverables and accountability
- Balance between ease of use and adding new capabilities
Recommendations
- Prepare a coherent, effective marketing approach for your growing user base
- There are opportunities to evolve MDSplus - international adoption.
- Formalizing planning would be helpful
- Focus on defining integration points, integration demonstrations
- Revisit your priorities and plans to assure maximum impact of the collaboratory mission
- Formalize focus on developing a long-term international vision and path for the Fusion Grid/Collaboratory
- Continue to be a driver for remote collaboration for ITER
- Document the process you used to integrate TRANSP into the grid.
Comments of PSACI PAC Review Committee
The PSACI PAC chaired by William L. Kruer reviewed the SciDAC Fusion Projects
and the NFC Project in June 2003. A letter, dated 6/13/03, was delivered to Robert
Goldston, Director of PPPL, summarizing the committee's findings. Below I summarize
the comments realted to the NFC Project.
Results
- The NFC has continued to deliver impressive technical results.
- FusionGrid has been established.
- Strong outreach to the user community continues.
- New capabilities in 3D visualization and animation have also been facilitated.
- For the ITER Project, a proposal has been developed that the U.S. take primary responsibility
for software for data acquisition and management, and for remote participation.
- The PAC identified Steve Eckstrand as the key point-of-contact for helping to identify
additional codes to be made remotely accessible.
- We also urge a consideration of how the Collaboratory can be used to advance graduate
education in the fusion program by making available simple plasma computational tools
for teaching purposes.
Comments
- As its capabilities become a more established part of the fusion computing infrastructure,
the Collaboratory needs to consider the issue of the most effective approach to development
and support of its software.
- In particular, the Collaboratory should address how build versus buy decisions are made,
both for immediate needs (e.g., visualization software), as well as for the long term.
- The Collaboratory is advised to consider whether corporate partners might at some appropriate
time become interested in contributing to development of the software as its capabilities
become better demonstrated and potentially applicable to a larger application domain.
- The PAC notes that capacity computing resources are clearly needed to accelerate scientific
productivity in the fusion energy sciences program. In view of the encouraging progress in the
development of the FusionGrid by the National Fusion Collaboratory and the successful
deployment of a pilot capacity computing facility at PPPL, we urge OFES to develop a
compelling distributed computing strategy.
