Identifying accurate resource monitoring tools and techniques

Abstract : Distributed applications concurrently share and compete for resources in heterogeneous systems. The objective of the Management System for Heterogeneous Networks (MSHN) is to use admission control, smart scheduling, and adaptation awareness in applications to successfully cope with the dynamics of resource availability. MSHN therefore requires knowledge of the expected resource utilization of applications that execute within the MSHN environment and the current state of these resources. MSHN relies on the above information to correctly identify resources to be assigned to these applications. This thesis investigates the capabilities of currently available communication resource status monitoring tools for the purpose of identifying those tools that, with low overhead, can provide accurate, end-to-end communication status information in a Windows NT environment. The techniques used by the various tools are described and the methods for determining the accuracy of these tools are specified. Results of the experiments with the various tools show that they add between 2% - 3% overhead in most cases and as much as 10% overhead in the worst case. Finally, none of the existing commercial tools studied gave an accurate assessment of the end-to-end communication throughput and latency for Windows NT 4.0.

[1]  Chris Peterson,et al.  Implementing a Performance Forecasting System for Metacomputing The Network Weather Service , 1997, ACM/IEEE SC 1997 Conference (SC'97).

[2]  Thomas R. Gross,et al.  Adaptive distributed applications on heterogeneous networks , 1999, Proceedings. Eighth Heterogeneous Computing Workshop (HCW'99).

[3]  John P Kresho.  Quality Network Load Information Improves Performance of Adaptive Applications , 1997 .

[4]  Bronis R. de Supinski,et al.  Accurately measuring MPI broadcasts in a computational grid , 1999, Proceedings. The Eighth International Symposium on High Performance Distributed Computing (Cat. No.99TH8469).

[5]  Thomas R. Gross,et al.  Direct queries for discovering network resource properties in a distributed environment , 1999, Proceedings. The Eighth International Symposium on High Performance Distributed Computing (Cat. No.99TH8469).

[6]  Paul F. Carff.  When is a simple model adequate for in scheduling in MSHN , 1999 .

[7]  William E. Johnston,et al.  The NetLogger methodology for high performance distributed systems performance analysis , 1998, Proceedings. The Seventh International Symposium on High Performance Distributed Computing (Cat. No.98TB100244).

[8]  Peter A. Dinda,et al.  An evaluation of linear models for host load prediction , 1999, Proceedings. The Eighth International Symposium on High Performance Distributed Computing (Cat. No.99TH8469).