期刊文献+

面向拟态存储系统的高效同步方法研究 被引量:2

Research on efficient synchronization method for mimic storage system
下载PDF
导出
摘要 针对现有拟态存储架构中数据同步方法时延增速过快,导致系统安全调度时的性能下降问题,提出了一种预同步模型,让备用执行体在异构池中利用检查点进行数据预同步工作,从而减少执行体上线时间。进一步地,根据预同步模型的同步特点和切换调度情况,提出一种执行周期最大有效率的检查点放置(execution cycle maximum efficiency checkpointing,CMEC)方法。通过最大化每个执行周期的有效工作率求得最佳的检查点间隔,较好地平衡了检查点开销和回滚开销。实验证明,与现有的全量同步策略相比,该方法缩短了执行体上线过程中的同步时间开销,提高了同步效率,保障了系统在业务量不断增加场景下的服务稳定性和连续性。 Aiming at solving the problem that rapid increase of time delay in data synchronization within existing mimic storage architecture,which leads to the performance degradation during system security scheduling,this paper proposed a pre-synchronization model that enabled standby executors to utilize checkpoints in heterogeneous pools for data pre-synchronization,thereby reducing the time required for the executor to go online.Furthermore,according to the synchronization characteristics of the pre-synchronization model and the switching scheduling situation,this paper designed CMEC method.This method could determine the optimal checkpoint interval by maximizing the effective work rate of each execution cycle,balancing checkpoint overhead and rollback overhead.The experiments show that,compared with the existing full-volume synchronization strategy,this method shortens the synchronization time required for the executor to go online,improves the synchronization efficiency and ensures the stability and continuity of the system under increasingly heavy business loads.
作者 李婉桃 张帆 陈鑫 朱进 黄潇 Li Wantao;Zhang Fan;Chen Xin;Zhu Jin;Huang Xiao(School of Cyber Science&Engineering,Southeast University,Nanjing 210000,China;Endogenous Security Research Center,Purple Mountain Laboratory,Nanjing 210000,China;National Digital Switching System Engineering&Technological R&D Center,Zhengzhou 450000,China)
出处 《计算机应用研究》 CSCD 北大核心 2023年第9期2792-2797,共6页 Application Research of Computers
基金 国家重点研发计划专项资助项目(2022YFB3104300)。
关键词 拟态存储系统 预同步模型 检查点 服务稳定性和连续性 mimic storage system pre-synchronization model checkpoint service stability and continuity
  • 相关文献

参考文献10

二级参考文献60

  • 1刘厚贵,邢晶,霍志刚,安学军.一种支持海量数据备份的可扩展分布式重复数据删除系统[J].计算机研究与发展,2013,50(S2):64-70. 被引量:5
  • 2HANDIGOL N, SEETHARAMAN S, FLAJSLIK M, et al. Plugin-serve: load-balancing Web traffic using OpenFlow[J]. ACM SIGCOMM Demo, 2009, 4(5): 6. 被引量:1
  • 3WANG R, BUTNARIU D, REXFORD J. OpanFlow-based server load balancing gone wild[C]//USENIX HotlCE. c2011:12. 被引量:1
  • 4LI H, LI P, GUO S, et al. Byzantine-resilient secure software-defined networks with multiple controllers in cloud[J]. IEEE Transactions on Cloud Computing, 2014, 2(4): 436-447. 被引量:1
  • 5DIXIT A, HAO F, MUKHERJEE S, et al. Towards an elastic distrib- uted SDN controller[C]//ACM SIOCOMM Computer Communication Review, 2013, 43(4): 7-12. 被引量:1
  • 6KOPONEN T, CASADO M, GUDE N, et al. Onix: a distributed control platform for large-scale production networks[C]//OSDl. c2010: 1-6. 被引量:1
  • 7LEV1N D, WUNDSAM A, HELLER B, et al. Logically centralized: state distribution trade-offs in software defined networks[C]//The First Workshop on Hot Topics in Software Defined Networks. ACM, e2012: 1-6. 被引量:1
  • 8STRAUB J. Control-plane consensus in software-defined networking: distributed controller synchronization using the ISIS2 toolkit[J/OL]. http://elib.uni-stuttgart.de/handle/1162/357/. 被引量:1
  • 9TOOTOONCHIAN A, GANJALI Y. HyperFlow: a distributed control plane for OpenFlow[C]//The 2010 Internet Network Management Conference on Research on Enterprise Networking. USENIX Associa- tion, c2010: 3. 被引量:1
  • 10LUO M, WU X, ZENG Y, et al. Multi-dimensional hashing for fast network inlbrmation processing in SDN[C]//Complex, Intelligent, and Software Intensive Systems (CISIS), 2015 Ninth International Con- ference. IEEE, c2015: 140-147. 被引量:1

共引文献210

同被引文献23

引证文献2

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