The aim of this paper is to deal with the problem of balancing load between the workstations in a network. In particular it addresses the main questions which arise in introducing a software package for distributed task allocation in a workstation network. The aim is to evaluate the impact on performance and the possibility of implementation in a real heterogeneous environment. We have considered the performance of some algorithms originally evaluated by Eager for a homogeneous environment. We point out the further complications introduced by the heterogeneity and possible solutions to them by referring to a distributed job allocator we have developed in a heterogeneous Unix environment. The result of the experimental evaluation allows us to extend to heterogeneous distributed systems the results produced by Eager, showing that in reality it is onerous, in terms of computation and communication, to improve the single workstation's knowledge of the entire load, but the use of a small amount of additional information allows a substantial improvement in the performance of the system. (C) 1997 Elsevier Science Limited. All rights reserved.

Assessment of job allocation on heterogeneous computer networks

DI STEFANO, Antonella;LO BELLO, Lucia;MIRABELLA, Orazio
1997-01-01

Abstract

The aim of this paper is to deal with the problem of balancing load between the workstations in a network. In particular it addresses the main questions which arise in introducing a software package for distributed task allocation in a workstation network. The aim is to evaluate the impact on performance and the possibility of implementation in a real heterogeneous environment. We have considered the performance of some algorithms originally evaluated by Eager for a homogeneous environment. We point out the further complications introduced by the heterogeneity and possible solutions to them by referring to a distributed job allocator we have developed in a heterogeneous Unix environment. The result of the experimental evaluation allows us to extend to heterogeneous distributed systems the results produced by Eager, showing that in reality it is onerous, in terms of computation and communication, to improve the single workstation's knowledge of the entire load, but the use of a small amount of additional information allows a substantial improvement in the performance of the system. (C) 1997 Elsevier Science Limited. All rights reserved.
1997
Distributed computer systems; scheduling
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/20.500.11769/32407
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 3
  • ???jsp.display-item.citation.isi??? 0
social impact