GcpSensor: a CPU performance tool for grid environments

摘要

In grid computing environments, CPU performance discovery and allocation is a serious issue. Before they run their applications, grid users need to choose grid nodes based on available CPUs performance. Hence a tool is needed for measuring CPUs performance. To address this need, GcpSensor (grid CPU performance sensor) was developed. GcpSensor runs a typical benchmark program in every grid node to measure CPU performance and returns a metric named WMFLOPS which can be used to compare different CPUs performance directly. An algorithm, which can choose a moderate computation cost dynamically, is designed to ensure that every test is comprehensive and the added overhead is comparatively small. Our experiments have proved that GcpSensor can indicate the CPU performance sensitively and that users can use the GcpSensor data to compare different CPUs performance straightforward, thus the accuracy and efficiency for selecting computing resources are increased.