Past Meeting - February
17, 2005
Hive Computing: Scalability and Fault Tolerance from Commodity
Hardware
Thursday, February 17, 2005
Summary
Mills Ripley and Jason Wyrick gave a fascinating talk on how to weave
commodity computing platforms (i.e., Windows PCs) into a mesh so as
to achieve a reliable computing base. The secret is in the executive,
which keeps track of all computing resources and deploys them to meet
the task at hand. Hive computing is similar to autonomic computing,
in that it treats the available resources as entities that could fulfill
a transaction, and if it fails, the transaction is rolled back and tried
elsewhere. 15 people shared the muchies and great material.
Abstract
Hive computing is a new computing model that directly addresses the
problems of spiraling costs, escalating complexity, and ever-increasing
service level requirements. It does this by providing a middleware fabric
that converts a collection of dedicated, commodity computers (x86 running
Windows or Linux) into a fault tolerant, scalable and dependable execution
environment for transactional and service-oriented applications. Hive
computing allows this network of commodity machines to be programmed
and managed as if they were one system. The net result is an extremely
simple, reliable, and scalable execution environment. For this meeting,
an overview of hive computing concepts will be followed by a hive demonstration
and technical discussions on hive programming and the hive application
execution environment.
Presenter Bio
Mills Ripley is a Regional Manager for Appistry
(Formerly Tsunami Research, Inc.) He has more than twenty years experience
in the computer industry. Before joining Appistry, he spent 7 years
at Rational Software (and subsequently IBM) as a technical representative,
technical lead, and manager. Prior to that, he held numerous programming,
consulting, and management positions at Xerox, Perot Systems, Digital
Equipment Corporation, and Diebold, Inc.
Jason Wyrick is a Regional Technical Lead for Appistry.
Prior to joining Appistry, he held technical positions at IBM, Rational
Software, Lockheed Martin, and General Electric.