HR: 11:20h
AN: IN42A-05 [Abstracts]
TI: Fault Tolerant Distributed Middleware for VLAB
AU: * Lu, Z
EM: zhenyulu@cs.fsu.edu
AF: Florida State University, Dirac Science Library, Tallahassee, FL 32306-4120
United States
AU: Bollig, E F
EM: bollig@msi.umn.edu
AF: University of Minneosta, Minnesota Supercomputing Institute and Dept. of Geology and Geophysics,
Minneapolois, MN 55455
United States
AU: Erlebacher, G
EM: erlebach@csit.fsu.edu
AF: Florida State University, Dirac Science Library, Tallahassee, FL 32306-4120
United States
AU: Gardgil, H
EM: hgadgil@cs.indiana.edu
AF: Indiana University, Community Grids Lab
505 N. Morton St., Bloomington, IN 47404
United States
AU: Yuen, D
EM: davey@krissy.msi.umn.edu
AF: University of Minneosta, Minnesota Supercomputing Institute and Dept. of Geology and Geophysics,
Minneapolois, MN 55455
United States
AU: Pierce, M
EM: mpierce@cs.indiana.edu
AF: Indiana University, Community Grids Lab
505 N. Morton St., Bloomington, IN 47404
United States
AU: Pallickara, S
IN42A-05
AF: Indiana University, Community Grids Lab
505 N. Morton St., Bloomington, IN 47404
United States
AB:
With increasingly large storage media, fast processors and improved data-collecting
instruments, the datasets in scientific fields are growing at an exponential rate. How to
analyze, visualize and manipulate those datasets (geographically distributed in most
cases) easily and efficiently within a collaborative environment is rather challenging.
We address this problem through NaradaBrokering (NB), a unique and flexible middleware
application program interface (API) (http://www.naradabrokering.org, [1]). Topics,
rather than IP addresses and hostnames are used to locate remote services and support
collaboration. In our framework, the underlying hardware, middleware, system load, or
resource availability is transparent to the end users.
Web Services are the key components
within this framework that enable building of loosely coupled applications. Furthermore,
multiple service providers can provide identical services. Our system routes client
requests to the best qualified service provider according to some default or user-defined
conditions. This not only provides a desired Quality of Service (QOS), but also acts as a
load balancing mechanism to better distribute the workload across available services. We
have deployed NB between Florida State University, the University of Minnesota and
Indiana University, and installed multiple instances of a wavelet service. We will
demonstrate fault tolerance with respect to the faulty nodes in the NB network and faulty
Wavelet service providers. Two users sharing an identical view through an applet will
illustrate the collaborative nature of our system.
[1] S. Pallickara and G. Fox, "NaradaBrokering: A Middleware Framework and Architec-
ture for Enabling Durable Peer-to-Peer Grid", in Proceedings of ACM/IFIP/USENIX
International Middleware Conference Middleware-2003. pp 41-61, (2003)
DE: 0520 Data analysis: algorithms and implementation
DE: 0525 Data management
DE: 0530 Data presentation and visualization
SC: Earth and Space Science Informatics [IN]
MN: Fall Meeting 2005