Support of Intelligent Integration of Information Using an Asynchronous Trigger Processor
Abstract
This effort has resulted in the development of an architecture for and prototype implementation of a new type of database rule system: the scalable, asynchronous trigger processor. A scalable, asynchronous trigger processor enhances existing database management systems and other information sources with the ability to trigger actions when new information meets criteria specified in rule conditions. Rule actions can update the database, call database procedures, or send alerts to people or running application programs. The prototype developed as part of this project, called TriggerMan, has demonstrated the capability to process tens of thousands of rules efficiently, and has the potential to scale to millions. The techniques implemented in TriggerMan to achieve scalability include a novel rule indexing technique and multi-threaded parallelism on a shared-memory multiprocessor. Investigations include easy-to-use techniques for processing temporal triggers (triggers with conditions based on how data changes over time); efficient methods for inferring updates done to passive data sources that do not support triggers; a query-decomposition based approach to testing join conditions in rules; and methods for testing trigger conditions efficiently when they contain expensive predicates.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 01, 2000
- Accession Number
- ADA379345
Entities
People
- Eric N. Hanson
Organizations
- University of Florida