A Transparent Distributed Shared Memory for Clustered Symmetric Multiprocessors
| Autor: | Jyh-Biau Chang, 張志標 |
|---|---|
| Rok vydání: | 2005 |
| Druh dokumentu: | 學位論文 ; thesis |
| Popis: | 93 A transparent distributed shared memory (DSM) system must achieve complete transparency in data distribution, workload distribution, and reconfiguration respectively. The transparency of data distribution allows programmers to be able to access and allocate shared data using the same user interface as is used in shared-memory systems. The workload distribution transparency can optimize the parallelism at both the user-level and the kernel-level. The achievement of the reconfiguration transparency can prompt system-wide jobs’ throughput rather than DSM programs’ performance and resolve the reconfiguration problem of software distributed shared memory (DSM) systems in non-dedicated clusters. In this dissertation, a transparent DSM system referred to as Teamster is proposed and is implemented for clustered symmetric multiprocessors. Teamster provides a Global Memory Image (GMI), whose purpose is to accomplish transparency of data distribution. With the GMI, the address space of each processor is precisely identical. Programmers are able to access and allocate shared data in the same way as they do in single SMP computers. Teamster uses the hybrid thread architecture to achieve the transparency of workload distribution. This thread architecture optimizes the parallelism at both the user-level and the kernel-level, and also improves the efficiency of run-time reconfiguration. We also invent a novel approach called Progressive Multilayer Reconfiguration (PMR) for DSM systems. As named, reconfiguration is divided into three different layers, i.e., processor, application, and node in this approach. According to the state transfer of the workload, the three different layer reconfigurations are progressively and respectively performed during the execution of DSM applications. The preliminary results show that the GMI and the hybrid thread architecture of Teamster not only provide the transparency to users, but also promise the performance of the DSM applications. Meanwhile, the PMR can enable Teamster not only to effectively utilize abundant CPU cycles available in non-dedicated clusters for DSM applications but also to effectively minimize the slowdown of local jobs caused by the disturb from DSM applications. Therefore, PMR is shown able to prompt the job throughput of the whole cluster effectively. With the transparency provided by Teamster, programmers can exploit all the computing power of the clustered SMP nodes in a transparent way as they do in single SMP computer. Compared with the results of previous researches, Teamster can realize the transparency of cluster computing and obtain satisfactory system performance. |
| Databáze: | Networked Digital Library of Theses & Dissertations |
| Externí odkaz: |
|