Los Alamos National Laboratory

Science >  LANL Institutes

National Security Education Center


ISTI Seminar: “Automated Index Tuning in Database Systems”

December 3, 2009
Time: 2:00 - 3:00 PM
Location: National Security Education Center (NSEC), Suite 101A, Access Grid Conference Room (LARP, 1st Floor)

Hosted by the Information Science and Technology Institute (ISTI) and the Institute for Scalable Scientific Data Management (ISSDM)

Speaker:  Neoklis Polyzotis
University of California-Santa Cruz

Abstract:  Database systems rely heavily on indexing in order to achieve good performance. Selecting the right indices is a difficult task for database administrators, since the selection has to balance the benefit of indexing, the cost of index creation and maintenance, and the usage of storage resources. This has led to the development of algorithms that generate index recommendations based on an analysis of the database workload. This paradigm of automated index tuning has been gaining in popularity and is currently integrated (in some form) in most modern database systems.

This talk will summarize our research efforts in the area of automated index tuning. The first part will describe the COLT framework for online index tuning. COLT monitors continuously the performance of query processing and automatically generates index recommendations that match the latest traits of the workload. An important feature of COLT is that it self-regulates its own performance, lowering its overhead when the system is well tuned and being more aggressive when the workload shifts and it becomes necessary to re-tune. The second part of the talk will present our recent work on analyzing interactions among the recommended indices. Interactions affect crucially the benefit of indices and are thus important in selecting which indices to materialize. Our work proposes the first systematic framework for computing efficiently such interactions. Moreover, we show how interactions can be applied in the development of novel administration tools for index tuning.

Biography:  Neoklis Polyzotis is currently an associate professor at UC Santa Cruz. His research focuses on database systems, and in particular on approximate query answering, on-line database tuning, and P2P databases. He is the recipient of an NSF CAREER award in 2004 and of an IBM Faculty Award in 2005 and 2006. He has also received the runner-up for best paper in VLDB 2007 and the best newcomer paper award in PODS 2008. He received his PhD from the University of Wisconsin at Madison in 2003.

Contact:  Josephine Olivas, jojo@lanl.gov, 505-663-5725 (office).

<< Back to calendar
Operated by Los Alamos National Security, LLC for the U.S. Department of Energy's NNSA
Inside | © Copyright 2008-09 Los Alamos National Security, LLC All rights reserved | Disclaimer/Privacy | Web Contact