Doug Judd, Creator, Hypertable

Doug has over a decade of software engineering experience in the area of distributed computing and information retrieval. He joined Inktomi's Web Search division in 1997 where he held both engineering and management positions. During his five year tenure, he designed and developed large-scale distributed systems, including significant pieces of the crawling and indexing software.

Doug later joined Kosmix, Inc., where he built a distributed web crawler and scaled it to a billion documents. Earlier in his career he worked as an engineer at Verity, helping to develop VDK, the core developer toolkit. Doug currently holds the position of Principal Search Architect at Zvents, Inc., a local search engine.

His primary responsibility is to lead the development of Hypertable, a next generation, high performance, scalable, database. Doug earned a B.S. in Computer Science from U.C. Santa Barbara in 1992 and holds four patents in search technology.

Presentation: "Hypertable - An Open Source, High Performance, Scalable Database"

Track: Data Storage Rethinking: Document Oriented Distributed Databases

Time: Friday 09:15 - 10:15

Location: Stanford

Abstract:

Hypertable is an open source, high performance, distributed database modeled after Google's Bigtable. It differs from traditional database technology in that the design emphasis is on scalability on commodity hardware as opposed to transactions and the relational model. Hypertable supports massive sparse tables sorted by a single primary key, which has proven to be sufficient for many Web 2.0 workloads. A Hypertable database can smoothly and cost-effectively scale to petabytes in size by simply adding machines to the underlying cluster.

Hypertable is designed to run on top of an existing distributed file system, such as the Hadoop DFS. One of the top design objectives for this project has been optimum performance. To that end, the system is written almost entirely in C++.