Priority Mechanisms for OLTP and Transactional Web Applications

  • Authors:
  • David T. McWherter;Bianca Schroeder;Anastassia Ailamaki;Mor Harchol-Balter

  • Affiliations:
  • -;-;-;-

  • Venue:
  • ICDE '04 Proceedings of the 20th International Conference on Data Engineering
  • Year:
  • 2004

Quantified Score

Hi-index 0.00

Visualization

Abstract

Transactional workloads are a hallmark of modernOLTP and Web applications, ranging from electronic commerceand banking to online shopping. Often, the databaseat the core of these applications is the performance bottleneck.Given the limited resources available to the database,transaction execution times can vary wildly as they competeand wait for critical resources. As the competitor is "only aclick away," valuable (high-priority) users must be ensuredconsistently good performance via QoS and transaction prioritization.This paper analyzes and proposes prioritization fortransactional workloads in traditional database systems(DBMS). This work first performs a detailed bottleneckanalysis of resource usage by transactional workloads oncommercial and noncommercial DBMS (IBM DB2, PostgreSQL,Shore) under a range of configurations. Second,this work implements and evaluates the performance of severalpreemptive and non-preemptive DBMS prioritizationpolicies in PostgreSQL and Shore. The primary contributionsof this work include (i) understanding the bottleneckresources in transactional DBMS workloads and (ii) ademonstration that prioritization in traditional DBMS canprovide 2x-5x improvement for high-priority transactionsusing simple scheduling policies, without expense to low-prioritytransactions.