Provided by: Association for Computing Machinery
Topic: Big Data
Date Added: Jun 2013
Modern data appliances such as IBM's Netezza and Oracle's Exadata are moving toward architectures with widespread query off-loading to different elements of the system. The goal is both to speed up query processing and to reduce data movement since the network bandwidth is still one of the dominant bottlenecks in cluster-based appliances. An important component in these systems is an intelligent storage engine, i.e., a storage engine that already filters data before sending it to the query processor, by executing queries in part or whole on behalf of the processing nodes.