内存网格调度方法分析  

Research on Memory Distributing and Sharing of RAM Grid

作　　者：卢俊文[1] 崔建峰[1] 肖蕾[1] 庄蔚蔚[1] 谢小竹[1] 

机构地区：[1]厦门理工学院计算机与信息工程学院,福建厦门361024

出　　处：《厦门理工学院学报》2015年第3期75-79,共5页Journal of Xiamen University of Technology

摘　　要：内存网格研究目的是在物理内存不足之时,各个节点之间共享物理内存.在分析内存访问速度、U-disk访问速度、网络带宽的基础上,提出了一种网格内存调度方法及内存管理框架.在某个网格节点内存不足之时,网格节点可以共享各个节点之间的内存,从而提高内存容量.通过分析其他节点内存读取时间、U-disk读取时间、与本地硬盘读取时间的差异,决定是否采用其他网格节点的内存或者U-disk内存.实验表明,提出的内存网格扩展了内存容量,减少了内存替换概率,提高了性能.Based on the analysis of the speed of RAM and U-disk and network widths, we proposed an approach of physical memory sharing of grid nodes and RAM grid managing. Memory can be shared among grid nodes so that their storage capacity is enhanced. Through an analysis of the difference in access time between different grid nodes, the U-disk and the local hard disk, we can decide whether it is efficient to get free physical memory from other grid nodes or U-disk. Simulation test proves that the proposed grid extends RAM, reduces its rate of replacement, and enhances its performance.

关 键 词：内存 内存网格 调度方法 网格数据库 时间延迟 

分 类 号：TP333[自动化与计算机技术—计算机系统结构]