摘要
IP存储是当前信息技术领域中一个热门的研究方向。iSCSI技术是IP存储的底层存储协议,它的目的是在TCP/IP网络上实现SCSI传输协议。
RAID(磁盘阵列)存储系统是构建大型存储系统的关键设备。现阶段RAID系统的协议接口主要采用的是光纤协议,成本高且管理复杂。随着iSCSI技术的飞速发展,支持iSCSI协议的RAID系统在国外已经发展得比较成熟。在国内,对磁盘阵列及iSCSI协议的研究由于起步较晚,取得的成果还不多,与国外相比仍有相当大的差距。实现支持iSCSI技术的RAID系统具有十分重要的意义。
深入探讨了iSCSI协议的相关标准,详细论述了iSCSI的体系结构及关键技术,从会话管理和连接、命名与寻址、协议数据单元的格式、错误恢复机制等方面对iSCSI协议进行了详细分析。
采用纯软件方式实现了LINUX平台下基于iSCSI协议接口的RAID存储系统。系统分为启动器和目标器两部分,启动器作为客户端发出请求,目标器作为服务器端处理请求及响应。目标器的实现包括前端驱动和中间层的实现,同时需要通过接口函数与内存管理和全局管理等RAID模块交互。
简要介绍了启动器的功能及实现原理,详细分析了iSCSI目标器的功能以及实现细节,详细描述了目标器与启动器之间的交互原理及过程,具体分析了目标器对于读写命令的执行过程。
搭建了一个支持iSCSI协议传输的RAID平台,通过专门的测试工具对该系统进行了测试,测试结果表明,系统在I/O传输率可以达到30MB/S左右,虽然带宽略低于光纤传输,但基本实现了支持iSCSI接口的RAID系统。
IP Storage is a hot area in the current information technology field. iSCSI is a low-level protocol in the IP storage technology. Its aim is to transfer SCSI command through TCP/IP network.
RAID(Redundant Array of Independent Disks) is the key element in a large-scale storage system. Currently most RAID system use Fibre Channel as their interface protocol, which is not only expensive but difficult to configure. With the rapid development of iSCSI technology, RAID systems using iSCSI protocol is becoming mature. In the domestic area, research on iSCSI is not very widely done. Implementing a RAID system supporting iSCSI protocol really means a lot.
Have discussed the standard iSCSi protocol and analyzed the detail of iSCSI protocol including session management,naming and addressing,format of PDU,error recovery. Have implemented a RAID storage system using iSCSI protocol.The whole system consistes of two parts,Initiator and Target.The initiator workes as a client and target is the server.The implementation of target includes FETD and STML,and the functions to communicate with RAID module CM and GC.
Describe the function of Initiator,anaylyse the detail of iSCSI Target,the communication between Initiator and Target,show the process of SCSI READ/WRITE command.
Established a RAID platform supporting iSCSI protocol.Have done some test on the iSCSI storage system. The result shows it perfoms well on data rate.
引文
[1]倪永军,谢长生.网络存储技术现状、存在的问题及对策研究.计算机工程与应用, 2003, 10(3): 159~161
[2]陈凯,白英彩.网络存储技术及发展趋势.电子学报. 2002, 30(12): 1929~1931
[3] Michael Abd-El-Malek, William V. Courtright II, Chuck Cranor, Gregory R. Ganger, et al. Ursa Minor: Versatile Cluster-based Storage. In Proceedings of the 4th USENIX Conference on File and Storage Technology (FAST '05). San Francisco, CA. December 13-16, 2005. USENIX Association. 59~72
[4] J. S. Plank, A. Bassi, M. Beck et al. Managing Data Storage in the Network. IEEE Internet Computing, 2001, 5(5): 50～58
[5]付长冬,舒继武,沈美明等.网络存储体系结构的发展和研究.小型微型计算机系统. 2004, 25(4): 485~489
[6] C. K. Wong. Minimizing Expected Head Movement in One-Dimensional and Two-Dimensional Mass Storage Systems. Computing Surveys, 1980, 12(2): 167~178
[7] Steven W. Schlosser, Jiri Schindler, Stratos Papadomanolakis, Minglong Shao Anastassia Ailamaki, Christos Faloutsos, Gregory R. Ganger. On Multidimensional Data and Modern Disks. In Proceedings of the 4th USENIX Conference on File and Storage Technology (FAST '05). San Francisco, CA. December 13-16, 2005. 225~238
[8]欧阳兴华.计算机系统接口-SCSI.北京:电子工业出版社,1994.10～12
[9] Ann L. Drapeau, Ken W. Shirrif, John H. Hartman, Ethan L. Miller, Srinivasan Seshan, Randy H. Katz, Ken Lutz, David A. Patterson, Edward K. Lee, Peter M. Chen, and Garth A. Gibson. RAID-II: a high-bandwidth network file server. In Proceedings of the 21st Annual International Symposium on Computer Architecture, 1994, 234~244
[10]张江陵,冯丹.海量信息存储.北京:科学出版社, 2003. 95~111
[11] T. E. Anderson, M. D. Dahlin, J. M. Neefe, D. A. Patteson, D. S. Roselli, and R. Y. Wang. Serverless network file systems. In 15th Symposium on Operating Systems Principles, ACM Transactions on Computer Systems, 1995. 109~126
[12]付长冬,舒继武,沈美明等.网络存储体系结构的发展和研究.小型微型计算机系统. 2004, 25(4): 485~489
[13] J. R. Heath, P. J. Yakutis. High speed storage area networks using a fibre channel arbitrated loop interconnect. IEEE Network, 2000, 14(2): 51~56
[14]赵亮.高性能磁盘阵列(RAID)关键技术的研究: [硕士学位论文].长沙:国防科技大学研究生院. 2002
[15] K. Hwang, H. Jin, and R. S. Ho. Orthogonal striping and mirroring in distributed RAID for I/O-centric cluster computing. IEEE Transactions on Parallel and Distributed Systems, 2002, 13(1): 26~44
[16] Hong Tang, Aziz Gulbeden, Jingyu Zhou, Lingkun Chu, and Tao Yang. Sorrento: A Self-Organizing Storage Cluster for Parallel Data-Intensive Applications. Technical Report 2003-30, UCSB, September 2003. 1~14
[17] Paul Massiglia. The RAID Book(Sixth Edition). RAID Advisory Board, 1997
[18]尚涛,施亮,吴智铭. RAID系统的结构化模型.微型电脑应用, 2001, 17(1): 29
[19]刘志君,朱兰娟,吴智铭. RAID系统中管理模块的设计与实现.微型电脑应用, 2003, 19(6): 47~49
[20] E. Riedel, G. Gibson, and C. Faloutsos. Active Storage for Large-Scale Data Mining and Multimedia Applications. In Proceedings of International Conference on Very Large Databases, New York, NY, Aug. 24-27, 1998. 62~73
[21] Erik Riedel and Garth Gibson. Active disks - remote execution for network-attached storage. TR CMU-CS-97-198. December 1997. 1~13
[22] P. M. Chen, E. K. Lee, G. A. Gibson, et al. RAID: high-performance, reliable secondary storage. ACM Computing Surveys, 1994, 26(2): 145~188
[23] Simitci.H,Malakapalli.C,Gunturu.Evaluation of SCSP over TCP/IP and SCSI over fibre channel connections.IEEE Hot Interconnects 9, 2001,87～91
[24] INSIC, Data Storage Devices and Systems (DS2) Roadmap. January 2005, 10~20.
[25] Jiri Schindler, John Linwood Griffin, Christopher R. Lumb, Gregory R. Ganger. Track-aligned Extents: Matching Access Patterns to Disk Drive Characteristics. In Proceedings of the FAST 2002 USENIX Conference on File and Storage Technologies, Monterey, CA, January 2002. 259~274
[26] Friedhelm Schmidt. SCSI Bus&IDE Interface. Harlow:Addison-Wesley ,1997.100～105
[27] R. Golding, et al. Attribute-managed storage. In Workshop on Modeling and Specification of I/O, San Antonio, TX, Oct 1995. 1~4
[28] E. Borowsky, R. Golding, A. Merchant, L. Schreier, E.Shriver, M.Spasojevic, and J. Wilkes. Using attribute-managed storage to achieve QoS. In 5th International Workshop on Quality of Service, Columbia Univ., New York, June 1997. 21~23
[29] E. Shriver. A formalization of the attribute mapping problem. Technical report HPL-SSP-95-10, HP Labs. July 1996. 1~37
[30] K. Voruganti and P. Sarkar. An Anlysis of Three Gigabit Networking Protocols for Storage Area Networks. In 20th IEEE International Performance, Computing, and Communications Conference, April 2001
[31] J. Satran, et al. iSCSI. IETF Work in Progress (IPS group), http://www. ietf.org/html.charters/ips-charter.html, 2001
[32] K. Z. Meth and J. Satran. Design of the iSCSI protocol. In Proceedings of 20th IEEE/11th NASA Goddard Conference, San Diego, California: IEEE, 2003. 116~122
[33] K. Z. Meth and J. Satran. Features of the iscsi protocol. IEEE Communications Magazine, 2003, 41(8): 232~249
[34] P Sarkar and K Voruganti. IP Storage: The Challenge Ahead. In Proceedings of the 19th IEEE Symposium on Mass Storage Systems, College Park, MD, April 2002. 35~42
[35] Peter M. Chen, Edward K. Lee. Striping in a RAID Level 5 Disk Array. In Proceedings of the 1995 ACM SIGMETRICS Conference on Measurement and Modeling of Computer Systems, ACM, May 1995. 136~145
[36] S3.iSCSI-offload.http://www.s3group.com,2003,1~4
[37] E. K. Lee and R. H. Katz. An analytic performance model of disk arrays. In Proceedings of ACM Conf. on Measurement and Modeling of Computer Systems (SIGMETRICS), May 1993. 98~109
[38] Elizabeth Varki, Arif Merchant, Xiaozhou Qiu. An Analytical Performance Model of Disk Arrays under Synchronous I/O Workloads. HP Technical report, 2002. 1~20
[39] Mustafa Uysal, Guillermo A. Alvarez, and Arif Merchant. A modular, analytical throughput model for modern disk arrays. In Proceedings of the Ninth International Symposium on Modeling, Analysis and Simulation of Computer and Telecommunications Systems (MASCOTS-2001), Cincinnati, Ohio. 15-18 August 2001. 183~192
[40] G. R. Ganger and Y. N. Patt. Using system-level models to evaluate I/O subsystem designs. IEEE Transactions on Computers, 1998, 47(6): 667~78
[41] G. Ganger, B. Worthington, and Y. Patt. The DiskSim Simulation Environment Version 3.0 Reference Manual, also published as Carnegie Mellon University Technical Report CMU-CS-03-102, January 2003
[42] William V. Courtright II. A transactional approach to redundant disk array implementation. PhD thesis, Department Electrical Engineering and Computer Science, Carnegie-Mellon University, May 1997. 1~32
