Status Reports

[ Status Reports] [ Research Agenda ]

 

Status Reports

---

Year 1: Project started May 2003, 2-page Executive Summary, Fourth quarter annual report is here

Year 2:

Year 3:

 

Research Agenda (DOE)

---

Year 1:

Develop algorithms for mapping an AGS onto one or more ARPs (i.e. resource allocation)

Develop ARP selection algorithms
Extend algorithms to allow a request to span multiple ARPs

Identify ARP feaures and attributes in support of adaptive resource management
Develop interfaces for AGSs and ARPs
Establish Grid prototype consisting of multiple Linux clusters

Each cluster managed by an ARP running Condor/Globus

Select several high-end DOE applications as candidates

Year 2:

Develop algorithms for workload-sensitive dynamic replication of an AGS
Develop and implement reliability techniques for service requests

Examine tradeoffs between replicated execution, fail-restart, checkpoint-restart
Extend ARP/AGS interfaces in support of reliability

Deploy DOE-relevant services/applications on Grid testbed
Develop and deploy OGSA/OGSI-compliant implementations of ARPs and AGSs
Evaluate performance of PP, RM, AS modules and dynamic replication
Examine performance/reliability tradeoff

Year 3:

Develop algorithms/software for robustness of infrastructure itself, e.g. AGS
Integrate best practice solutions for AAA into service model
Middleware packaging and documentation release
Evaluate middleware and DOE-relevant applications and sercices on Grid testbed

 

[Contact Us] [U of M DCSG] [DOE National Collaboratories]

---

Distributed Computing Systems Group
Dept. of Computer Science & Engineering
University of Minnesota
Last updated : July 5. 2003