Fast Compositional Queries in a Relational Grid Information Service

Abstract

A Grid Information service (GIS) stores information about the resources of a distributed computing environment and answers questions about it. We have developed RGIS, a GIS system that, unlike previous efforts, is based on the relational data model. RGIS users can write SQL queries that search for complex compositions of resources that meet collective requirements. Executing these queries can be very expensive, however. In response, we have introduced three query techniques, nondeterminism, scoping, and approximation, that allow the user (and RGIS) to trade off between the query_s running time and the number of results. Herein we describe RGIS, our query techniques, and their implementation. Our evaluation shows that a meaningful tradeoff between query time and results returned is achievable, and that the tradeoff can be used to keep query time largely independent of query complexity. RGIS uses our techniques to bound query execution time. This strongly supports our general case for GIS systems based on the relational data model and RDBMSes.

An Importance-Aware Architecture for Large-Scale Grid Information Services

“…Going beyond the current state-of-the-art, the architecture proposed in this paper aims to provide additional flexibility by using centrally-controlled prefetching, which makes use of a dynamically adjusted frequency of updates for each grid site, depending on the site's importance. The proposals in [14,12] are similar to the one in this paper in that they also collect grid-wide resource information at a central relational database. However, both proposals collect detailed host and network related information (as opposed to the collections per grid site used in our proposal) and do not prescribe how monitoring is performed nor do they cope with varying frequencies of updates.…”