Abstract
The coordination and scheduling of affiliated tasks to be run at different sites is a challenging problem, specifically in the domain of network performance monitoring. This paper presents a software implementation of the Probes Coordination Protocol (PCP) which provides a solution to this problem. The PCP allows tasks to be executed regularly on a multitude of sites without the need for repeated user or administrator intervention. In addition it provides a robust mechanism for handling site failures, and a VOMS based security model to ensure appropriate usage. The paper provides the original motivation for the PCP, describes the PCP itself and discusses the new software implementation and its application.