Administration and Monitoring Service for Storage Virtualization in Grid Environments

Abstract

A grid is collection of computers and storage resources maintained to serve the needs of some community (Foster et al., 2002). It addresses collaboration, data sharing, cycle sharing and other patterns of interaction that involve distributed resources. Actually, the de facto building block is for high performance storage systems. In grid context, the scale and the reliability have key issues as many independently failing and unreliable components need to be continuously accounted for and managed over time (Porter & Katz, 2006). Manageability also becomes of paramount importance, since nowadays the grid commonly consists of hundred or even thousands of storage and computing nodes (Foster et al., 2002). One of the key challenges faced by high performance storage system is scalable administration and monitoring of system state. A monitoring system captures subset of interactions amongst the myriad of computational nodes, links, and storage devices. These interactions are interpreted in order to improve performance in grid environment. On a grid scale basis, ViSaGe aims at providing to the grid users a transparent, reliable and powerful storage system underpinned by a storage virtualization layer. ViSaGe is based on three services, namely as: administration and monitoring service, storage virtualization service and distributed file system. The virtualization service incorporates storage resources distributed on the grid in virtual spaces. These virtual spaces will be attributed by the distributed file system various qualities of service and data placement policies. In this paper, we present our scalable distributed system: Admon. Admon consists of an administration module that manages virtual storage resources according to their workloads based on the information collected by a monitoring module. It is known that the performance of a system depends deeply on the characteristics of its workload. Usually, the node's workload is associated to the service response time of a storage application. As the workload increases, the service response time becomes longer. However, the utilization percentage of system resources (CPU load, the Disk load and Network load) must be taken into more consideration. Therefore, Admon traces ViSaGe's applications and collects system resources (CPU, Disk, Network...) percentage of utilization. It is characterized by an automatic instrumentation. It is an auto-manager monitoring system. In this paper, we demonstrate Admon efficiency due to nodes workload and user constraint like CPU usage workload. 7