一种提高N-Code可扩展性的数据重组方案  被引量:2

Data Reorganization Approach to Improve the Scalability of N-Code

在线阅读下载全文

作  者:刘靖宇 李萧言 李浩鹏 李娟 武优西 LIU Jing-yu;LI Xiao-yan;LI Hao-peng;LI Juan;WU You-xi(School of Artificial Intelligence,Hebei University of Technology,Tianjin 300401,China)

机构地区:[1]河北工业大学人工智能与数据科学学院,天津300401

出  处:《小型微型计算机系统》2023年第2期442-448,共7页Journal of Chinese Computer Systems

基  金:国家自然科学基金项目(61976240)资助.

摘  要:在原有大规模存储系统中增加磁盘可以提高I/O带宽和存储容量,为了恢复原磁盘阵列的编码布局,数据需要在所有磁盘上重新分布.RAID6在充分利用存储空间的基础上提高了容错能力,但由于其编码的复杂性给存储系统扩容带来困难.本文提出了一种加速N-Code RAID6扩容的新型数据重组方案Cross-Scale,方案中定义了扩容阈值,根据扩容阈值与磁盘阵列剩余空间的关系提出两种阵列标准化和数据迁移策略,使扩容后阵列维持原编码布局,并有效减少了奇偶校验更新的开销,提高了扩容效率.实验表明,与现有的方案相比最多可降低97.40%的数据迁移率,I/O操作数可减少6.74%~73.37%,XOR操作次数可降低6.13%~100%,节省6.81%~73.39%的扩容总时间.Adding disks to the original large-scale storage system can increase I/O parallelism and storage capacity.Data need to be redistributed evenly among all disks including the old and the new,so as to restore the layout of the original disk array.RAID6 improves the fault tolerance on the basis of making full use of storage space.However,due to its coding complexity,it makes storage system scaling difficult.In order to accelerate N-Code RAID6 scaling,this paper proposes a data reorganization approach Cross-Scale.According to the characteristic of N-code,it defines the capacity threshold,and proposes two kinds of array normalizing operation and data migration strategies to maintain the original code layout based on the relationship between the capacity threshold and the remaining space of the disk array.It can greatly reduce the overhead of parity update and improve the scaling efficiency.Simulation experiments indicate that,compared with the existing scheme,the data migration rate is reduced by 97.40%at most,the number of I/O operations and the number of XOR operations are reduced by 6.74%~73.37%and 6.13%~100%,respectively.Finally,the total scaling time is saved by 6.81%~73.39%.

关 键 词:RAID6 存储系统扩容 扩容阈值 条带拼接 数据迁移 

分 类 号:TP368[自动化与计算机技术—计算机系统结构]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象