Tuesday Sept 25th, 14.30 - 16.00, Auditorium Maximum
[slides]
Abstract: In two previous papers some of us predicted the end of "one size fits all" as a commercial relational DBMS paradigm. These papers presented reasons and experimental evidence that showed that the major RDBMS vendors can be outperformed by 1-2 orders of magnitude by specialized engines in the data warehouse, stream processing, text, and scientific data base markets.
Assuming that specialized engines dominate these markets over time, the current relational DBMS code lines will be left with the business data processing (OLTP) market and hybrid markets where more than one kind of capability is required. In this paper we show that current RDBMSs can be beaten by nearly two orders of magnitude in the OLTP market as well. The experimental evidence comes from comparing a new OLTP prototype, H-Store, which we have built at M.I.T. to one of the popular RDBMSs on the standard transactional benchmark, TPC-C.
We conclude that the current RDBMS code lines, while attempting to be a "one size fits all" solution, in fact, excel at nothing. Hence, they are 25 year old legacy code lines that should be retired in favor of a collection of "from scratch" specialized engines. The DBMS vendors (and the research community) should start with a clean sheet of paper and design systems for tomorrow's requirements, not continue to push code lines and architectures designed for the yesterday's requirements.
Bio: Dr. Stonebraker has been a pioneer of data base research and technology for more than a quarter of a century. He was the main architect of the INGRES relational DBMS, and the object-relational DBMS, POSTGRES. These prototypes were developed at the University of California at Berkeley where Stonebraker was a Professor of Computer Science for twenty five years. More recently at M.I.T. he was a co-architect of the Aurora/Borealis stream processing engine as well as the C-Store column-oriented DBMS. He is the founder of four venture-capital backed startups, which commercialized these prototypes. Presently he serves as Chief Technology Officer of StreamBase Systems, Inc., which is commercializing Aurora/Borealis and Vertica, which is commercializing C-Store.
Professor Stonebraker is the author of scores of research papers on data base technology, operating systems and the architecture of system software services. He was awarded the ACM System Software Award in 1992, for his work on INGRES. Additionally, he was awarded the first annual Innovation award by the ACM SIGMOD special interest group in 1994, and was elected to the National Academy of Engineering in 1997. He was awarded the IEEE John Von Neumann award in 2005, and is presently an Adjunct Professor of Computer Science at M.I.T., where he is working on a variety of future-generation data-oriented projects.
Tuesday Sept 25th, 14.30 - 16.00, Auditorium Maximum
[slides] [animated slides (high quality, 38MB)] [animated slides (low quality, 9MB)]
Abstract: Data management, one of the most successful software technologies, is the bedrock of almost all business, government, and scientific activities, worldwide. Data management continues to grow, more than doubling in data and transaction volumes every two years with a growth in deployments to fuel a $15 billion market. A continuous stream of innovations in capabilities, robustness, and features lead to new, previously infeasible data-intensive applications. Yet forty years of DBMS innovation are pushing DBMSs beyond the complexity barrier where one-size-fits-all DBMSs does not meet the requirements of emerging applications. While data management growth will continue based primarily on the relational data model and conventional DBMSs, a much larger and more challenging data management world is emerging. In the 1990's data under DBMS management reached 10% of the world's data. The six-fold growth of non-relational data in the period 2006-2010 will reduce that number to well below 5%.
We are entering the next generation of computing with a fundamentally different computing model and paradigm characterized technologically by multi-core architectures, virtualization, service-oriented computing, and the semantic web. Computer Science 2.0 will mark the end of the Computing Era with its focus on technology and the beginning of the Problem Solving Era with its focus on higher levels of abstraction and automation (i.e., intelligent) tools for real world (i.e., imprecise) domains in which approximate and ever-changing answers are the norm. This confluence of limitations of conventional DBMSs, the explosive growth of previously unimagined applications and data, and the genuine need for problem solving will result in a new world of data management.
The data management world should embrace these opportunities and provide leadership for data management in Computer Science 2.0. Two emerging areas that lack such guidance are service-oriented computing and the semantic web. While concepts and standards are evolving for data management in service-oriented architectures, data services or data virtualization has not been a focus of the DBMS research or products communities. Missing this opportunity will be worse than missing the Internet. The semantic web will become the means by which information is accessed and managed with modest projections of 40 billion pages with hundreds of triples per page - the largest distributed system in the world - the only one. Tim Berners-Lee and the World Wide Web Consortium view databases are being nodes that need to be turned inside out. What does the database community think? Semantic web services will constitute a programming model of Computer Science 2.0. How does data management fit into this semantically rich environment?
Computer Science 2.0 offers the data management community one of the biggest challenges in its forty-year history and opens up a new world of data management. Key to success in this new world will be collaboration with other disciplines whether at the technical level - partnering with the semantic technologies community to augment their reasoning capabilities with systems support - or at the problem solving level - partnering with real world domains as proposed by the new discipline of Web Science.
Bio: Dr. Michael L. Brodie is the Chief Scientist of Verizon Services Operations in Verizon Communications, one of the world's leading providers of communications services. Dr. Brodie works on large-scale strategic Information Technology (IT) opportunities and challenges for Verizon Communications senior executives. His primary interest is in delivering business value from advanced and emerging technologies and practices to enable business objectives while optimizing and transforming IT. In addition to Computer Science he addresses business and economic issues such as computing-communications-entertainment-business convergence and the economic impact of IT.
His research focus is on advanced computational models and architectures and the large-scale information systems that they must support. He is concerned with the Big Picture, business and technical contexts, core technologies, and integration within a large scale, operational telecommunications environment. Dr. Brodie holds a PhD in Databases from the University of Toronto.