内存云架构的磁盘节能策略
发布时间:2018-08-15 15:06
【摘要】:内存云(RAMCloud)的出现改善了在线数据密集型(OLDI)应用的用户体验,但其能耗高于传统的云数据中心。针对该问题,提出一种适用于该架构的磁盘节能策略。首先,引入遗传算法中适应度函数和轮盘赌法,尽量选择更为节能的磁盘进行数据持久化备份;其次,设定合理的服务器内存缓冲区来延长磁盘的平均连续空闲时间,使得部分服务器磁盘在空闲时进入待机状态。仿真实验结果表明,在50台服务器的内存云系统中,该策略能有效节能约12.69%;而缓冲区大小的设定对于节能效果和数据可用性具有双重影响,需权衡考虑。
[Abstract]:The emergence of memory cloud (RAMCloud) improves the user experience of online data-intensive (OLDI) applications, but its energy consumption is higher than that of traditional cloud data centers. To solve this problem, a disk energy saving strategy suitable for this architecture is proposed. Firstly, the fitness function and roulette method in genetic algorithm are introduced to select more energy efficient disks for data persistent backup as far as possible. Secondly, reasonable server memory buffer is set to prolong the average continuous idle time of disk. Make part of the server disk idle into standby state. The simulation results show that the strategy can save energy effectively about 12.69 in 50 server memory cloud systems, and the buffer size has dual influence on energy saving effect and data availability, which needs to be considered.
【作者单位】: 新疆大学信息科学与工程学院;新疆大学软件学院;
【基金】:国家自然科学基金资助项目(61262088,61063042) 新疆维吾尔自治区自然科学基金资助项目(2011211A011)
【分类号】:TP333
本文编号:2184572
[Abstract]:The emergence of memory cloud (RAMCloud) improves the user experience of online data-intensive (OLDI) applications, but its energy consumption is higher than that of traditional cloud data centers. To solve this problem, a disk energy saving strategy suitable for this architecture is proposed. Firstly, the fitness function and roulette method in genetic algorithm are introduced to select more energy efficient disks for data persistent backup as far as possible. Secondly, reasonable server memory buffer is set to prolong the average continuous idle time of disk. Make part of the server disk idle into standby state. The simulation results show that the strategy can save energy effectively about 12.69 in 50 server memory cloud systems, and the buffer size has dual influence on energy saving effect and data availability, which needs to be considered.
【作者单位】: 新疆大学信息科学与工程学院;新疆大学软件学院;
【基金】:国家自然科学基金资助项目(61262088,61063042) 新疆维吾尔自治区自然科学基金资助项目(2011211A011)
【分类号】:TP333
【相似文献】
相关期刊论文 前7条
1 陈朝学;;数据中心机房节能策略[J];科技信息;2008年29期
2 吴甜;刘利祥;虎嵩林;;绿色数据中心的服务器节能机制与策略[J];微电子学与计算机;2011年08期
3 ;节以溯本 能以致用——华硕服务器2008年深耕节能策略[J];通信世界;2008年15期
4 ;能效逻辑和数据中心定量评价指标[J];电源世界;2008年09期
5 ;节省能源和保护环境的绿色计算机[J];电子产品世界;1994年05期
6 朱文忠;;闪速存储器系统的节能方法研究[J];科技创新导报;2008年06期
7 ;削减数据中心能耗经验谈[J];每周电脑报;2008年28期
相关硕士学位论文 前2条
1 王哲;基于Hadoop集群的节能优化技术研究[D];华中科技大学;2013年
2 崔慎杰;集群系统中动态节能调度方法研究[D];华中科技大学;2013年
,本文编号:2184572
本文链接:https://www.wllwen.com/kejilunwen/jisuanjikexuelunwen/2184572.html