-
公开(公告)号:KR1020130074953A
公开(公告)日:2013-07-05
申请号:KR1020110143097
申请日:2011-12-27
Applicant: 한국과학기술정보연구원 , 한국과학기술원
CPC classification number: G06F9/45504 , G06F9/50 , G06F15/161 , G06F15/167
Abstract: PURPOSE: An apparatus for dynamic virtual machine placement and a method thereof are provided to increase resource efficiency of a whole cluster system, and to improve performance degradation of a virtual machine (VM) due to shared resource collision in a cloud computing system. CONSTITUTION: A cache miss rate collection part (71) collects cache miss rate of virtual machines included in nodes included in a cluster, and a VM rearrangement part (73) rearranges the VM according to the cache miss rate. A global phase VM rearrangement part of the VM rearrangement part includes a node selection part, a VM selection part and anode VM rearrangement part. The node selection part receives the cache miss rate information, and selects anode having maximum and minimum miss rate. The VM selection part selects a VM having maximum and minimum cache miss rate in the node having maximum and minimum cache miss rate. The node VM rearrangement part rearranges by exchanging the selected VM mutually. [Reference numerals] (71) Cache miss rate collection part; (73) VM rearrangement part
Abstract translation: 目的:提供一种用于动态虚拟机放置的装置及其方法,以增加整个集群系统的资源效率,并且由于云计算系统中的共享资源冲突而改善虚拟机(VM)的性能下降。 构成:高速缓存未命中收集部分(71)收集包括在群集中的节点中包括的虚拟机的高速缓存未命中率,并且VM重新排列部分(73)根据高速缓存未命中率重新排列VM。 VM重排部分的全局相位VM重排部分包括节点选择部分,VM选择部分和阳极VM重新排列部分。 节点选择部分接收高速缓存未命中率信息,并选择具有最大和最小错失率的阳极。 VM选择部分选择具有最大和最小高速缓存未命中率的节点中具有最大和最小高速缓存未命中率的VM。 通过相互交换所选择的VM来重新排列节点VM重排部分。 (附图标记)(71)缓存错误率收集部分; (73)VM重排部分
-
公开(公告)号:KR1020130074952A
公开(公告)日:2013-07-05
申请号:KR1020110143096
申请日:2011-12-27
Applicant: 한국과학기술정보연구원 , 한국과학기술원
CPC classification number: G06F9/45504 , G06F9/50 , G06F15/161 , G06F15/167
Abstract: PURPOSE: An apparatus for providing cloud computing system guide and a method thereof are provided to optimize computing time and usage charge of a cloud computing system, as providing optimum virtual machine (VM) selection information to a user. CONSTITUTION: A data input part (20) receives input data, execution time restriction condition and cost restriction condition of an execution application. An available virtual machine (VM) assembly information generation part (22) generates information for the assembly of VMs available in computing of the application. A time-memory usage amount estimation part (24) estimates execution time and memory usage amount of the application for the VM combinations, by using an artificial neutral network model on the basis of the received data and the restriction conditions. A final VM combination information providing part (26) guides an optimum VM combination, by using an estimated value. [Reference numerals] (20) Data input unit; (22) Available virtual machine assembly information generation part; (24) Time-memory usage amount estimation part; (26) Final VM combination information providing part
Abstract translation: 目的:提供一种用于提供云计算系统指南的装置及其方法,以优化云计算系统的计算时间和使用费用,从而为用户提供最佳的虚拟机(VM)选择信息。 构成:数据输入部(20)接收执行应用的输入数据,执行时间限制条件和成本限制条件。 可用的虚拟机(VM)组装信息生成部分(22)生成用于在计算应用程序时可用的VM的组装的信息。 时间存储器使用量估计部分(24)通过基于所接收的数据和限制条件使用人工中性网络模型来估计VM组合的应用的执行时间和存储器使用量。 最终的VM组合信息提供部分(26)通过使用估计值来引导最佳VM组合。 (附图标记)(20)数据输入单元; (22)可用虚拟机装配信息生成部分; (24)时间存储器使用量估计部; (26)最终的VM组合信息提供部分
-