---------- Forwarded message ----------
Date: Fri, 3 Dec 2004 12:16:05 -0000
From: Dave Berry <[log in to unmask]>
To: [log in to unmask], [log in to unmask],
[log in to unmask]
Subject: [Resend] Seminar: Distributed Monitoring and Information Services
for the Grid
Distributed Monitoring and Information Services for the Grid
Jennifer M. Schopf
National e-Science Centre and Argonne National Laboratory
4pm, Monday January 10th, 2005
e-Science Institute, 15 South College Street, Edinburgh
Monitoring and information services form a key component of any deployed
Grid. As testbeds are achieving basic functionality, more and more
emphasis is being placed on the discovery of available resources and the
monitoring of their status. Current Grid information and monitoring
systems are being used in different and more extensive ways than
originally planned, and current systems are finding difficulty meeting
today's monitoring challenges.
We performed a quantitative study to aid in understanding the
performance limitations, advise in the deployment of the systems, and
help evaluate future development work. Results of some scalability tests
for three monitoring and information services will be presented: the
Globus Toolkit's Monitoring and Discovery Service (MDS2), the European
Data Grid Relational Grid Monitoring Architecture (R-GMA), and Hawkeye,
part of the Condor project. Our study shows that each approach has
different behaviors, often due to their different design goals, but core
principles such as caching, multi-threading and replication are
imperative when deploying a large scale system.
We then give an overview of the design and architecture of the
Monitoring and Discovery System (MDS4), the information system portion
of the Globus Toolkit 4. As part of the WSRF core, every service can
participate as a first class member in this broader MDS by publishing
data about itself, so monitoring in this environment has changed
dramatically. We discuss what monitoring and discovery looks like in a
Grid services world, and what future work is needed to make users happy.
Bio:
Dr. Jennifer M. Schopf is a Scientist at the Distributed Systems Lab,
part of the Mathematics and Computer Science Division at Argonne
National Lab, and is spending the year as a researcher at the National
e-Science Center in Edinburgh, UK. She is a member of the Globus
Alliance, and technology coordinator for the MDS, and spent 5 years
helping to establish the Global Grid Forum standards body. She received
a BA in Computer Science and Mathematics from Vassar College, and MS and
PhD degrees from the University of California, San Diego in Computer
Science and Engineering. Currently, her research interests include
monitoring, performance prediction, and resource scheduling and
selection.
Dave Berry
Research Manager, National e-Science Centre
15 South College St., Edinburgh
Tel: +44 131 6514039
|