1. 제 목 : Timely, Fresh Web Data Services
2. 발 표 자 : Kyoung-Don Kang
(Binghamton University Computer Science Department)
3. 일 시 : 2007년 8월 23일(목) 15 : 00 - 17 : 00
4. 장 소 : 경북대 공대 9호관 521호
5. 초청교수 : 한욱신 교수
6. 강사약력 :
University of Virginia Computer Science Ph.D. 2003
University of Virginia Computer Engineering M.S. 2000
Kyungpook National University, South Korea Computer Engineering M.E. 1992
Kyungpook National University, South Korea Computer Science B.E. 1990
Present: Assistant Professor, Department of Computer Science,
State University of New York at Binghamton
7. 내용요약 :
In a number of important applications, e.g., stock trading and transportation management, a database needs to process transactions in a timely fashion using fresh data reflecting the current market or traffic status. Unfortunately, database workloads may considerably vary due to dynamic data/resource contention. Further, transaction timeliness and data freshness requirements may compete for system resources. In this talk, I will present a novel feedback control scheme designed to support the desired data service delay by managing the amount of the backlog in the database. I will also present a new self-adaptive update policy, which can automatically adapt the freshness of temporal data based on temporal data access and update patterns. Unlike most existing work on feedback control of real-time database performance, our work actually implements and evaluates feedback control and database workload adaptation techniques in a real database testbed modeling online stock trades. In our testbed, called Chronos, thousands of client threads concurrently request database services for stock quotes, trades, and portfolio updates in a bursty manner. In these experiments, our database system supports the desired response time bound and data freshness, while processing a significantly larger number of transactions in time compared to the tested baselines.
※ 주최 : BK21 정보기술연구인력양성사업단, 컴퓨터공학과
◀ 문의처 : BK21정보기술연구인력양성사업단 ☎ 950-6613 ▶
2007 .8. 20
|
