2012 Scientific Collaborations at Extreme-Scale
Title: CILogon-HA: Higher Assurance Federated Identities for DOE ScienceProject URL: http://www.cilogon.org/ha
Abstract
PI Jim Basney – UIUC
Summary: Enhancements to the CILogon project to support High Assurance capabilities to the tool. This work separates the act of issuing user ID's from consuming them. The work deliverables are weak (white papers) and it is mostly an advanced deployment project.
Title: Tigres: Template Interfaces for Agile Parallel Data-Intensive Science
Project URL: http://tigres.lbl.gov/
Abstract
PI: Deb Agarwal (LBNL), Dan Gunter (LBNL), Lavanya Ramakrishnan (LBNL), Keith Jackson (LBNL)
Summary: Develop a new workflow model based on defining a few 'primitives' that can be used to create workflow models. Users then use scripting languages instead of GUI's to construct workflows.
Title: International Collaboration Framework for Extreme Scale Experiments (ICEE)
Project URL: http://sdm.lbl.gov/icee/
Abstract
PI: Kesheng (John) Wu (LBNL), Scott Klasky (ORNL), CS Chang (PPPL), Yeong-Kook Oh (Kstar)
Summary: Framework for real-time workflows with KSTAR as the initial test case. Project leverages ADIOS and FastBIT technologies to enhance distributed collaboration environments.
Title: Next Generation Workload Management and Analysis System for Big Data
Project URL: http://www.usatlas.bnl.gov/twiki/bin/view/PanDA/WebHome
Abstract
PI: Alexei Klimentov (BNL), Kaushik De (UTA)
Summary: Enhance PanDA Workflow Management System (WMS) (software development and engineering work) to 1) modularize the code to encourage use by other science communities; 2) create domain specific extensions; 3) add in support for dynamic circuits using VNOD or something similar; and 4) add real-time monitoring functions including perfSONAR based network monitoring. The goal is to make PanDA attractive to other science communities. A 2nd goal is to look at ways to run jobs on leadership class machines.
BNL Press release http://www.bnl.gov/atlas/news/news.asp?a=3336&t=today
UTA Press release http://www.uta.edu/news/releases/2012/09/BigData-release.php
Title: Infrastructure Strategy to Support Collaboration
Project URL: http://press3.mcs.anl.gov/irmo/
Abstract
PI: Katarzyna Keahey (ANL), Lavanya Ramakrishnan (LBNL)
Summary: Develop models and reference implementations for VM based computing ecosystems. These systems would allow scientific users to easily use a wide range of compute resources (laptop to supercomptuer).
Title: An Integrated Middleware Framework to Enable Extreme Collaborative Science
Project URL: http://aimes-project.org
Abstract
PI: Shantenu Jha (Rutgers), Daniel S. Katz(UC), Jon Weissman(UMN)
Summary: Develop a middleware framework that extends the PilotJob paradigm to storage and abstracts the physical resource from the application or job. The team will develop both algorithms needed to find or run-time link these application to physical resources and generate reference implementations of these algorithms. AIMES will co-design middleware from an application and infrastructure perspective by integrating abstractions for compute, data, and network across multiple levels to provide an interoperable, extensible and scalable middleware stack for extreme scale science.
Title: dV/dt – Accelerating the rate of progress towards extreme scale collaborative science
Project URL: https://sites.google.com/site/acceleratingexascale/
Abstract
PI: Miron Livny(UWisc), Frank Wuerthwein (UCSD), Ewa Deelman (USC), Douglas Thain (ND), William Allcock (ANL)
Summary: Develop a series of algorithms and maybe mechanisms, required to create and maintain scientific collaborations. Elements include, finding/allocating CPU, GPU, storage, and networking. The team will also evaluate trust relationship issues. The focus is on using glideinWMS technologies to create a computer science VO to do this research.
Title: Automated Metadata, Provenance Cataloging and Navigable Interfaces: Ensuring the Usefulness of Extreme-Scale Data
Project URL: http://fusion.gat.com/mpo
Abstract
PI: David Schissel (GA), Arie Shoshani (LBNL), Martin Greenwald (MIT)
Summary: Define and develop mechanisms and tools that enhance the scientific discovery process. 4 specific tasks are called out, automatic provenance generation, storing and searching this providence data, graphical and textual display methods, and deployment/testing via the FES partners. The stated goal is to create generic solutions while using FES to test ideas and concepts.
Title: Advanced Identity Management for Extreme-Scale Scientific Collaboration
Project URL: http://cacr.iu.edu/collab-idm
PR: http://newsinfo.iu.edu/news/page/normal/23058.html
Abstract
PI: Von Welch
Summary: Develop a comprehensive model for Identity Management in large scale collaborations. The focus is on first defining the potential trust relationships and then developing some software to validate this model.
Title: Resource Discovery for Extreme Scale Collaboration
Project URL: http://tw.rpi.edu/web/project/rdesc
Abstract
PI: Karen Schuchardt (PNNL), Peter Fox (RPI)
Summary: Develop and implement an architecture to support query and discovery mechanisms based on semantic web ideas. The goal is to handle complex queries made by scientists.