Performance Analysis of A Grid Monitoring System - Ganglia

Monitoring is critical for providing robust, high performance Grid environment. Without monitoring mechanisms, it will be impossible to determine the status or health of environment. And so it will be difficult to use it effectively & efficiently. Monitoring data can be used for performance analysis, performance diagnosis, performance tuning, fault detection and scheduling. Due to the complexity and dynamics of grid, a strong monitoring system is needed which should have features like Grid Monitoring Architecture compliance, Scalability, Resource Performance Analysis and Forecasting, Directory service for events subscription & notification etc. In this paper, we have reviewed the grid monitoring system – ‘Ganglia’. ‘Ganglia’ is a distributive monitoring system for high performance computing systems such as clusters and the Grid. It is based on a hierarchical design targeted at federations of clusters. This paper discusses the various performance parameters and criteria of the grid monitoring system – ‘Ganglia’ and provides some useful information as far as Grid Monitoring Architecture is concerned. Keywords— Ganglia Monitoring System, gmond daemon, gmetad daemon, Scalability, CPU overhead, Bandwidth, Memory Overhead