Parallel Virtual Memory for Time Shared Environments 3
|
Abstract:
This paper analyses the issues involved in providing virtual distributed shared memory for time-shared parallel machines. We study the performance of two different page management policies, namely, static and dynamic page allocation under two widely accepted scheduling policies: Gang scheduling and independent processor scheduling. The performance of each page management policy is studied under different replacement scopes (local versus global replacement) Results obtained after extensive simulations show that dynamic page allocation performs better throughout all the environments simulated. We also observe a better performance of independent processor over gang scheduling as well as a similar performance between local and global replacement scope. Sum'ario Este artigo analisa os problemas envolvidos em se realizar mem'oria virtual em ambientes distribu'idos de mem'oria logicamente compartilhada (DSM) em m'aquinas paralelas executando em tempo compartilhado. S~ao analisadas as performances de duas pol'iticas de gerenciamento de p'aginas: alocac~ao est'atica e dinamica sob duas pol'iticas de escalonamento amplamente aceitas: grupo e independente. Estuda-se ainda o desempenho dessas duas pol'iticas de gerenciamento sob diferentes estrat'egias de substituic~ao de p'agina: local e global. Os resultados obtidos ap'os v'arias simulac~oes indicam alocac~ao dinamica de p'aginas como a melhor opc~ao. Observamos tamb'em um melhor desempenho do escalonamento independente sobre o de grupo, e um desempenho equivalente entre as duas pol'iticas de substituic~ao de p'aginas estudadas.
Citations
| 705 | SPLASH: Stanford Parallel Applications for Shared Memory – Singh, Weber, et al. - 1992 |
| 64 | Characterizing the caching and Synchronization Performance of a Multiprocessor Operating System – Torrellas, Gupta, et al. - 1992 |
| 30 | The Cray Research Massively Parallel Processor System CRAY T3D. available by anonymous ftp from ftp.cray.com – Oed - 1993 |
| 25 | Munin: Distributed shared memory using multiprotocol release consistency – Bennett, Carter, et al. |
| 24 | The performance of cache-coherent ring-based multiprocessors – Barroso, Dubois - 1993 |
| 15 | Shared Virtual Memory on Loosely-coupled Multiprocessors – Li - 1986 |
| 7 | Enhancing Applications Performance on Intel Paragon through Dynamic Memory Allocation – Saini, Simon - 1993 |
| 2 | A Virtual Memory Model for Parallel Supercomputers – Reis, Scherson - 1996 |
