CSE 225
Updated: March 29, 2005
1.
Grid Computing –
Vision and Realities (3 classes)
a.
Lecture 1: Vision
of Wide-Area Shared Resource Environments (Grids)
·
Grids
in Context (Smarr 1999)
·
What is the
Grid? (Foster 2002)
·
What is Grid
Computing? (IBM 2002)
·
The
Grid: Past, Present, Future (Berman, Fox, Hey 2003)
b.
Lecture 2: Real
Production Grids and Experiences
·
SETI@Home
– A Desktop Grid (Anderson 2002) and backup Entropia:
Architecture of a Desktop Grid System (Chien 2002)
·
Open
Science Grid (Avery 2005); and backup The Grid2003
Production Grid:Principles and Practice (Gardner
2004)
·
Planetlab: A Blueprint for Introducing Disruptive
Technology in the Internet (Peterson 2002)
c. Lecture 3: High Performance Distributed Applications
·
Massive
Computation: Genome and Proteomics Analysis Encyclopedia
of Life (Li 2004) , Protein Folding in Folding@home (Pande 2004
video)
·
Online
Simulation:
(1)
Structures
and Earthquakes (Kesselman 2004) (NEESGrid Slides,
Spencer 2004)
(2)
LEAD: Tornadoes
and Hurricanes (Droegemeier 2005)
· Massive Data:
(1)
High
Energy Physics: Atlas
on the Open Science Grid (Gardner 2005)
(2)
Medical Research: Biogrids (Ellisman 2004), and backup Biomedical
Informatics Research Network (Ellisman 2003)
(3)
Astronomy/Astrophysics: National Virtual
Observatory
·
(2)
HP’s Adaptive Enterprise (2004)
2.
Dynamic
Applications are Resource Aware: “Beauty is in the eye of the beholder.” (5 classes)
a.
Lecture 4: Resource Information and Description
·
Metacomputing Directory Service
·
Globus RSL/GRAM and Redline
·
Matchmaking
and Gang Matching
·
Additional: SWORD,
vgDL: Virtual
Grid High Level Resource Abstractions
·
Recipe-Based
Service Composition
b.
Lecture 5: Resource Selection (Nut Taesombut)
·
Same Papers as
Lecture 4, selection discussion
c.
Lecture 6: Dynamic Monitoring and Resource Behavior
·
Desktop Grid Resource
Characterization
·
Statistical
Availability Modeling
·
Impact: Optimizing
Makespan
d.
Lecture 7: Challenges in Supporting Adaptive
Applications: the GrADS Project
e.
Lecture 8: Virtual Grids: Supporting Adaptive
Applications
·
Virtual Grids:
Describing Resource Needs (vgDL)
and Robust Selection and Binding vgES
Resource Selection)
·
Design of the VGrADS System (and Open Research Challenges)
(1)
Scheduling and
Rescheduling (EMAN, Fault tolerance)
(2)
Integration of
Dynamic Information
(3)
Monitoring
3.
Open Resource
Sharing: “You can’t always get what you want.” (4 classes)
a.
Lecture 9:
Resource Models: Asymmetric Resource Sharing in Desktop Grids
·
Condor:
Hunter of Idle Workstations
·
Resource
Management in the Entropia System
·
Bounding
the Resources Savings of Utility Computing Models
b.
Lecture 10:
Resource Models: Traditional Batch Resource Managers and
Co-allocation/Reservation
·
Portable Batch
System (Shareware and product web site)
·
PBS
and Maui Scheduler Comparison
c.
Lecture 11:
Resource Models: Best-effort Slice and Virtualization Systems; Share Scheduling
·
Planetlab Slice and info on Proportional
Scheduling
·
Optional: SHARP:
Resource Exchange and Comparing
Planetlab and Globus
d.
Lecture 12:
Resource Allocation: Economic techniques, Stability
·
Current Resource
Allocation Practice
·
Market-based
Central Allocation and backup Bellagio
·
Network and
Application Uncertainty: Resource
Utilization and Network/Application Uncertainty
4.
Configurable
Networks and Grids: “Communication as a first-class resource.” (4 classes)
a.
Lecture
13: Technology
(OptIPuter), and Large-scale Infrastructure (Starlight)
for configurable optical networks and CANARIE/CALREN/NLR; GLIF,
etc.
b.
Lecture 14: Usage
Models: Intelligent Networks, Super
FTP model, Distributed
Virtual Computer and Application Specification: DVC-ISL, RSL2, ASTDL and DRAC, UCLP and backup (Canarie UCLP),
c.
Lecture 15: Protocols for Lambda Grids: UDT,
Group
Transport Protocol, and Composite
Endpoint Protocol
5.
Advanced
Distributed Applications (1 class)
a. GriPhyN:
the Grid Physics Network
b.
Biomedical
Informatics Research Network and see
the BIRN Web
6.
Course Project
Presentations – Thursday June 2, 1230-330pm, 4882 APM
References
[Gridbook
2003] The Grid: Blueprint for a
New Computing Infrastructure, Second Edition
Morgan Kaufmann (eds. Foster, Kesselman), 2003.
[GridInfrastructure2003] Grid Computing: Making the Global
Infrastructure a Reality, (eds. Fox,
Berman, Hey), Wiley and Sons 2003.
For more information, email Professor Andrew Chien