Agenda What is Grid Computing Goals of Grid Computing Pros and Cons Examples Of Grid Systems Grid Softwares Application Suitability Example Researches Conclusion
What is Grid Computing
Emerging computing model; that provides the ability to perform higher throughput computing by taking advantage of many networked computers to model a virtual computer architecture that is able to distribute process execution accross a parallel infra-structure.
Parts of Grid Computing
Infrastructure Grid: Hardware resources such as
processors, storage and network Applications Grid: Packaged or custom applications used
for business logic and process flow within application software
Goals Of Grid Computing
Pervasive, uniform, and reliable access to data, storage capacity and computation power
Resource utilization Delivery of computer power as a
utility. (City Water, electricity systems)
Pros And Cons
PROS CONS
Time Saving Resource Management(Who is Prior?)
Resource Saving Security problem(Data is remote)
Space Saving Schedule Problem(Who , when?)
Money Saving
Examples Of Grid Systems
Seti@Home is the most popular(1st). Climateprediction, Einstein@Home
etc. All uses BOINC software Over 18 projects working using BOINC
Blade Servers
For power, heat and space savings Includes processors, memory,
network cards and storage
Globus Grid Management Software
Product for managing grids Supported by many large vendors
GRAM Grid Resource Allocation Manager “gatekeeper” process that controls execution of
job managers MDS Monitoring and Directory Service
GRIS (Grid Resource Information Service) collects local info
GIIS (Grid Index Information Service) collects GRIS info
GASS (Global Access to Secondary Storage)
Java Technology RMI, Web Services and JNDI RMI : Remote Method Invocation Web Services : Software As Services,
Soap Packages over HTTP Orchestration of Services E-commerce, Online Community
Servers J2EE Application Servers > Load
Balancing, Fail Over mechanisms
Application Suitability
For a grid system to perform applications must have some criteria.
Best Suitable > Large Batch jobs that requires minimum inter process communication and synchronization
Academic Researches In 2004, Genaud : Algorithm for load
balancing. Allock Et. Al. : Data transfer problem
in high performance grid environments. (Replica management Service that uses an FTP-like protocol to make replicas of data elements accross the grid to improve performance)
Conclusion In summary, Grid computing is;
cooperation of different computers, for a specific task, so that the user acquires better performance for that specific task.
“The Load” is on remote machines rather than clients.
Like: Transfer of CPU power as electricity power.
References en.wikipedia.org/wiki/Grid_computing www.oracle.com/technologies/grid/index.html Web Services and Service Oriented Architectures: A
Savvy Manager’s Guide, Douglas K Barry, Elsevier Science, 2003.
Loosely Coupled: The Missing Pieces of Web Services, Doug Kaye, RDS Press, 2003.
Intel White Paper, Changing the IT Paradigm: Blade Server Computing, 2003
Allock B. Et al. Data management and transfer in high performance computational grid environments. Parallel Computing 28, 2002, pp 749-771