A Centralized, Extensible, Multilayer Monitoring System for Distributed Infrastructures: The Atlas Tier2-Naples Experience

Abstract

Services and Infrastructures monitoring is a crucial issue in a distributed computing environment that involves all the Grid architecture layers, from the network to the single application service. In this paper we propose a multi layer model that aims to integrate monitoring applications of different subsystems and to offer management and resource presentation services to the end-user and to system administrator, through a set of middleware components. After a discussion about the principal requirements, in the paper we identify a set of different access levels, then we design a general and extensible environment, which provides a single entry point, with different views depending on the user privileges. As a case of study, we describe the monitoring system implemented for the distributed Atlas Tier2 at Naples, the monitored subsystem, the functionality and the technologies used to map the requirements in a real use case.