QoS-Based Data Access on Grid

A wide variety of applications require access to multiple heterogeneous, distributed data sources on grid. However, the quality of data access is still low such as unpredictable response time, irrelevant results, because existing query optimization techniques do not consider the dynamic properties of data resource in grid environments. In this paper, we present a QoS (quality of service)-based data access on grid, this approach integrates QoS requirements into distributed query processing and considers also the dynamic properties of the system. Segmented pipe tree based on pipelined parallelism is also proposed to optimize the multi-join queries, which can not only improve the parallelism but also return the partial results to user as soon as possible. The experimental results confirm that the proposed solution is effective for providing different levels data access for different users.