期刊文献+

多核下基于远程直接内存访问的高效远程过程调用研究

Study on RDMA-based high performance RPC in multi-core systems
下载PDF
导出
摘要 针对目前基于远程直接内存访问(RDMA)的远程过程调用(RPC)设计在多核背景下存在的性能和扩展性问题,提出了一种新的基于RDMAWrite的高效RPC设计方法并对网络文件系统(NFS)提供了支持。重点就设计中的安全性问题和内存管理机制进行了分析,并针对内存有效利用提出了被动和主动两种回收机制。测试结果表明,这种新的RPC设计方法能降低RPC延迟13%~20%。对内存文件系统及真实磁盘文件系统的测试表明,此设计能有效提高多核系统中RPC的性能和扩展性。 In consideration of the performance and scalability problems of current designs of remote procedure call (RPC) over remote direct memory access (RDMA) in multi-core systems, this paper proposes a novel design of high performance RPC based on RDMA Write and provides support to the network file system (NFS), The paper mainly focuses on the security concerns and the RDMA memory management mechanism of the design. In terms of the utilization of the RDMA memory, this paper proposes the passive reclaim strategy and the active reclaim strategy. The evaluations show that the new design proposed in this paper decreases the RPC latency by 13% - 20%. The design can highly improve the RPC performance and scalability in multi-core systems.
出处 《高技术通讯》 CAS CSCD 北大核心 2011年第7期681-686,共6页 Chinese High Technology Letters
基金 863计划(2009AA01A129)和国家自然科学基金(60633040)资助项目.
关键词 远程直接内存访问(RDMA) 远程过程调用(RPC) 多核 网络文件系统(NFS) remote direct memory access (RDMA), remote procedure call (RPC), multi-core, network file system (NFS)
  • 相关文献

参考文献8

  • 1Sandberg R. The sun network filesystem: design, implementa- tion, and experience. In: Proceedings of the Distributed Com- puting Systems: Concepts and Structures. Los Alamos, USA, 1992. 300-316.
  • 2Ko M, Chadalapaka M, Hufferd J, et al. iSCSI Extensions for RDMA. http://tools, ieff. org/html/rfc5046: IETF, 2007.
  • 3Magoutis K, Addetia S, Fedorova A, et al. Making the most out of direct-access network attached storage. In: Proceedings of 2nd USENIX Conference on File and Storage Technologies, San Francisco, USA, 2003. 159-202.
  • 4Goglin B, Prylli L. Performance analysis of remote file systemaccess over a high-speed local network. In: Proceedings of the Workshop on Communication Architecture for Clusters, in Con- junction with International Parallel and Distributed Processing Symposium '04, Santa Fe, New Mexico, USA, 2004. 185- 192.
  • 5Callaghan B, Lingutla-Raj T, Chiu A, et al. NTS over RD- MA. In: Proceedings of the ACM SIGCOMM workshop on Net- work-I/O convergence: Experience, Lessons, Implications. New York, 2003. 196-208.
  • 6Noronha R, Chai L, Talpey T, et al. Designing NFS with RD- MA for Security, Performance and Sealability. In: Proceedings of the 2007 International Conference on Parallel Processing, Xi'an, China, 2007. 49-56.
  • 7Noronha R, Chai L, Shepler S, et al. Enhancing the Perfor- mance of NFSv4 with RDMA. In: Proceedings of the Interna- tional Workshop on Storage Network Architecture and Parallel I/Os, San Diego, USA, 2007. 90-96.
  • 8Yu W K, Nageswara S V R, Wyckoff P, et al. Performance of RDMA-capable storage protocols on wide-area network. In: Proceedings of the 2008 Petascale Data Storage Workshop, Austin, USA, 2008. 1-5.

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部