Co-Work:基于HDFS的安全云存储数据一致性保持算法
发布时间：2018-12-09 18:43
【摘要】：针对云存储平台中用户隐私和敏感数据的安全保护问题,在前期提出的基于自主可控机制的安全云存储模型ASOM(All Self-Organization Model)基础上,对MDSS端元数据的管理操作、MDSS和DMS之间的通信过程做进一步优化,针对ASOM模型实际场景引入锁思想实现两个节点一致性的Co-Work算法,完成MDSS和DMS节点之间的协同工作和数据的一致性保持。同时,考虑网络带宽对ASOM中读写效率的影响,引入随机表机制,改变DSS上报的时间结点,以提高ASOM整体读写效率。测试结果表明:执行Co-Work算法后的ASOM模型实现了数据的物理与逻辑隔离,保证用户对元数据的自主控制和管理,而且随着数据尺寸增大读写效率明显提高,在数据达到1 GB时读写效率提高了12%。
[Abstract]:Aiming at the security protection of user privacy and sensitive data in cloud storage platform, based on the security cloud storage model (ASOM (All Self-Organization Model) based on autonomous and controllable mechanism, the management operation of MDSS metadata is proposed in this paper. The communication process between MDSS and DMS is further optimized. The lock idea is introduced into the ASOM model to implement the Co-Work algorithm for the consistency of two nodes. The cooperative work between MDSS and DMS nodes and the consistency of data are achieved. At the same time, considering the influence of network bandwidth on the efficiency of reading and writing in ASOM, the random table mechanism is introduced to change the time node reported by DSS, so as to improve the overall reading and writing efficiency of ASOM. The test results show that the ASOM model after the implementation of the Co-Work algorithm realizes the physical and logical isolation of the data, ensures the user's autonomous control and management of the metadata, and improves the efficiency of reading and writing with the increase of the size of the data. When the data reaches 1 GB, the efficiency of reading and writing is increased by 12.
【作者单位】： 广东工业大学计算机学院;
【基金】：国家科技支撑计划课题(No.2015BAK19B03) 广州市科技计划项目(No.2014XYD-007) 教育部留学回国人员科研启动基金(No.14ZK0152)
【分类号】：TP333;TP309
，
本文编号：2369854
[Abstract]:Aiming at the security protection of user privacy and sensitive data in cloud storage platform, based on the security cloud storage model (ASOM (All Self-Organization Model) based on autonomous and controllable mechanism, the management operation of MDSS metadata is proposed in this paper. The communication process between MDSS and DMS is further optimized. The lock idea is introduced into the ASOM model to implement the Co-Work algorithm for the consistency of two nodes. The cooperative work between MDSS and DMS nodes and the consistency of data are achieved. At the same time, considering the influence of network bandwidth on the efficiency of reading and writing in ASOM, the random table mechanism is introduced to change the time node reported by DSS, so as to improve the overall reading and writing efficiency of ASOM. The test results show that the ASOM model after the implementation of the Co-Work algorithm realizes the physical and logical isolation of the data, ensures the user's autonomous control and management of the metadata, and improves the efficiency of reading and writing with the increase of the size of the data. When the data reaches 1 GB, the efficiency of reading and writing is increased by 12.
【作者单位】： 广东工业大学计算机学院;
【基金】：国家科技支撑计划课题(No.2015BAK19B03) 广州市科技计划项目(No.2014XYD-007) 教育部留学回国人员科研启动基金(No.14ZK0152)
【分类号】：TP333;TP309
，
本文编号：2369854
本文链接：https://www.wllwen.com/kejilunwen/jisuanjikexuelunwen/2369854.html
