WO2016082546A1 - Decision-making method and apparatus for virtual machine migration - Google Patents

Decision-making method and apparatus for virtual machine migration Download PDF

Info

Publication number
WO2016082546A1
WO2016082546A1 PCT/CN2015/083192 CN2015083192W WO2016082546A1 WO 2016082546 A1 WO2016082546 A1 WO 2016082546A1 CN 2015083192 W CN2015083192 W CN 2015083192W WO 2016082546 A1 WO2016082546 A1 WO 2016082546A1
Authority
WO
WIPO (PCT)
Prior art keywords
virtual machine
migrated
information
migration
physical node
Prior art date
Application number
PCT/CN2015/083192
Other languages
French (fr)
Chinese (zh)
Inventor
郑晓航
Original Assignee
中兴通讯股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 中兴通讯股份有限公司 filed Critical 中兴通讯股份有限公司
Publication of WO2016082546A1 publication Critical patent/WO2016082546A1/en

Links

Images

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/34Signalling channels for network management communication
    • H04L41/342Signalling channels for network management communication between virtual entities, e.g. orchestrators, SDN or NFV entities

Definitions

  • This paper deals with the field of cloud computing and network virtualization, and in particular relates to a virtual machine migration decision method and device.
  • OpenStack no Chinese technical terminology
  • OpenStack provides a set of operating platforms or tools to deploy the cloud to provide solutions for cloud computing infrastructure services.
  • OpenStack is open source, interoperable, portable, and easy to use. It also provides an API (Application Programming Interface) for secondary development and easy integration with other systems. Win more and more attention in the industry.
  • the migration of a virtual machine refers to moving the virtual machine operating system and applications running on the source physical host to the target physical host and running normally on the target host.
  • the migration function of OpenStack's own virtual machine allows a running virtual machine instance to be migrated from one compute node to another, but the migration of the virtual machine requires the operator to manually operate the interface because the operator is in the virtual machine migration. It is impossible to understand the network utilization of the entire network. It is not clear how to migrate to achieve the optimal effect. The network may become more congested after migration. Therefore, the current implementation cannot achieve the purpose of dynamic monitoring, dynamic decision-making, and dynamic migration.
  • the embodiment of the invention provides a virtual machine migration decision method and device, which solves the problem that the network is blocked due to virtual machine migration and the dynamic decision cannot be made.
  • the embodiment of the invention provides a virtual machine migration decision method, which includes:
  • the difference between the network traffic occupation information and the traffic occupation information of each virtual machine is determined according to the preset traffic occupation information of each virtual machine;
  • the network traffic statistics information includes a current actual physical network card bandwidth occupation of the physical node
  • Determining whether the network traffic statistics information of the physical node meets the threshold information, where the threshold information is that the traffic information of the physical node under normal working includes:
  • performing the migration to the virtual machine according to the difference includes:
  • the network traffic occupation information includes a current actual egress bandwidth occupation of the virtual machine
  • Determining, according to the traffic occupation information preset by each virtual machine, the difference between the network traffic occupation information and the traffic occupation information of each virtual machine includes:
  • performing the migration on the at least one virtual machine whose positive value is positive includes:
  • the at least one virtual machine to be migrated as a virtual machine to be migrated, and the virtual machine to be migrated
  • the machine is scheduled to perform the migration.
  • the determining, in the list of the virtual machines to be migrated, the at least one virtual machine that needs to be migrated, so that the network traffic statistics of the physical node after the migration of the at least one virtual machine that needs to be migrated meets the threshold information include:
  • the method before the migrating the virtual machine to be migrated, the method further includes:
  • All the virtual machines in the virtual machine list to be migrated are arranged in descending order according to the difference;
  • the method further includes:
  • the traffic restriction on the unmigrated virtual machine is cancelled.
  • the method before the performing the migration on the virtual machine according to the difference, the method further includes:
  • the physical node that meets the virtual machine migration condition so that the network traffic statistics information of the physical node that is moved into the virtual machine still meets the threshold information, and the selected physical The node performs the migration as the target physical node of the at least one virtual machine.
  • the embodiment of the invention further provides a virtual machine migration decision device, comprising:
  • Obtaining a module configured to obtain network traffic statistics information of the physical node, and network traffic occupation information of each virtual machine in the physical node;
  • the determining module is configured to determine whether the network traffic statistics information of the physical node acquired by the acquiring module meets threshold information, where the threshold information is traffic information under normal working of the physical node;
  • a determining module configured to determine, according to the traffic occupation information preset by each virtual machine, a difference between the network traffic occupation information and the traffic occupation information of each virtual machine, respectively, if the determining module determines that the content is not satisfied; as well as
  • the migration module is configured to perform migration on the virtual machine according to the difference determined by the determining module.
  • the network traffic statistics information includes a current actual physical network card bandwidth occupation of the physical node
  • the determining module is further configured to determine, according to a maximum actual physical NIC bandwidth occupancy of the physical node, a bandwidth occupied by the physical NIC under normal working of the physical node;
  • the determining module is configured to determine whether the current actual physical network card bandwidth occupancy of the physical node meets the physical network card bandwidth occupancy of the physical node under normal working.
  • the determining module is further configured to determine whether the difference determined by the determining module is positive;
  • the migration module is configured to perform migration on the at least one virtual machine whose difference is positive.
  • the network traffic occupation information includes a current actual egress bandwidth occupation of the virtual machine
  • the determining module includes:
  • the calculating sub-module is configured to calculate, according to the maximum amount of the egress bandwidth preset by each virtual machine, a difference between the current actual egress bandwidth occupancy of each virtual machine and the maximum egress bandwidth.
  • the device further includes:
  • Add a module and set to add the virtual machine whose positive value is positive to the virtual machine list to be migrated;
  • the determining module is further configured to: in the list of the virtual machines to be migrated, determine at least one virtual machine that needs to be migrated, so that the network of the physical node after the at least one virtual machine that needs to be migrated is migrated The traffic statistics information meets the threshold information;
  • the migration module is configured to perform the migration on the virtual machine to be migrated by using the at least one virtual machine to be migrated as a virtual machine to be migrated.
  • the determining module is further configured to determine, in the list of the virtual machines to be migrated, whether the network traffic statistics information of the physical node after the virtual machine having the largest difference value is migrated meets the threshold information;
  • the determining module is configured to determine, if the determining module is satisfied, the virtual machine with the largest difference as the virtual machine that needs to be migrated, and if not, the determining module is further configured to determine that the migration is performed. Whether the network traffic statistics of the physical node having the largest difference and the virtual node having the second largest difference satisfy the threshold information, and the determining module is configured to: if the determining module determines that the content is satisfied, the The virtual machine with the largest difference and the virtual machine with the largest difference are determined to be virtual machines that need to be migrated. If not, the judgment module is further configured to continue to judge until the network traffic of the physical node after the virtual machine is migrated is determined.
  • the statistical information meets the threshold information
  • the determining module is configured to determine a virtual machine that meets the condition determined by the determining module as a virtual machine that needs to be migrated, where the condition is a network traffic of a physical node after the virtual machine is migrated.
  • the statistical information satisfies the threshold information.
  • the device further includes:
  • the arranging module is configured to arrange all the virtual machines in the virtual machine list to be migrated according to the difference from large to small;
  • a limiting module configured to limit traffic of the at least one virtual machine that is sequentially arranged by the arranging module
  • the module is canceled, and if the at least one virtual machine includes the unmigrated virtual machine, the traffic restriction on the unmigrated virtual machine is cancelled.
  • the acquiring module is further configured to obtain network traffic statistics information of other physical nodes;
  • the device further includes: a selecting module, configured to select, according to network traffic statistics information of other physical nodes acquired by the acquiring module, a physical node that satisfies a virtual machine moving in condition, and a network of the physical node that moves into the virtual machine The traffic statistics information still satisfies the threshold information, and the selected physical node is used as the target physical node of the at least one virtual machine to perform migration.
  • a selecting module configured to select, according to network traffic statistics information of other physical nodes acquired by the acquiring module, a physical node that satisfies a virtual machine moving in condition, and a network of the physical node that moves into the virtual machine The traffic statistics information still satisfies the threshold information, and the selected physical node is used as the target physical node of the at least one virtual machine to perform migration.
  • the embodiment of the invention further provides a computer readable storage medium storing program instructions, which can be implemented when the program instructions are executed.
  • the embodiment of the invention provides a virtual machine migration decision method and device, and combines the services provided by the OpenStack framework and the OpenStack to determine whether it is needed according to the network traffic statistics of the physical node and the network traffic occupation information of each virtual machine in the physical node.
  • the migration of virtual machines in physical nodes overcomes the shortcomings of OpenStack's existing migration methods. It can dynamically monitor the network traffic usage of virtual machines, dynamically determine the migration of virtual machines, and reduce the manual participation of system maintenance. By obtaining and storing network traffic statistics, the network traffic usage of the virtual machine is analyzed to determine whether to trigger the dynamic migration of the virtual machine and the dynamic control of the virtual machine network traffic.
  • the original OpenStack host selection can be modified. The algorithm increases the filtering conditions for the network and optimizes the host selection algorithm when the virtual machine is migrated.
  • FIG. 1 is a flowchart of a virtual machine migration decision method according to Embodiment 1 of the present invention
  • FIG. 2 is a schematic structural diagram of a virtual machine migration decision apparatus according to Embodiment 2 of the present invention.
  • FIG. 3 is a schematic structural diagram of a virtual machine migration decision apparatus according to Embodiment 3 of the present invention.
  • Embodiment 1 is a diagrammatic representation of Embodiment 1:
  • FIG. 1 is a flowchart of a virtual machine migration decision method according to Embodiment 1 of the present invention. As shown in FIG. 1 , the virtual machine migration decision method includes:
  • S101 Obtain network traffic statistics information of the physical node, and network traffic occupation information of each virtual machine in the physical node;
  • S102 Determine whether the network traffic statistics information of the physical node meets the threshold information, where the threshold information is traffic information under normal working of the physical node;
  • S104 Perform migration on the virtual machine according to the difference.
  • the network traffic statistics of the physical node and the network traffic occupation information of each virtual machine in the physical node can be obtained, and the obtaining action can be obtained in real time. It can also be used for timing acquisition. According to the network traffic statistics of the physical node and the network traffic occupation information of each virtual machine in the physical node, it is determined whether the virtual machine in the physical node needs to be migrated, thereby reducing the manual participation of the system maintenance. .
  • the physical node may be an OpenStack computing node, that is, a physical host, and whether the OpenStack computing node needs to migrate the virtual machine therein, and the OpenStack control node may make a decision on the OpenStack computing node, that is, the OpenStack control node reports the network traffic reported by the node according to the OpenStack. Statistics to determine whether to migrate virtual machines in OpenStack compute nodes.
  • the network traffic statistics of the physical node are used to determine whether the virtual machine in the physical node needs to be migrated, so that the network traffic statistics of the physical node are lower than the threshold information, and the threshold information is the physical node.
  • the traffic information under normal work makes it meet the actual application requirements.
  • the network traffic statistics include, but are not limited to, the physical host's identification information, the current actual physical network card bandwidth usage, and the like.
  • the current actual physical network card bandwidth usage includes, but is not limited to, the physical port of the physical node, the data packet rate, and the actual Processing network throughput, etc.
  • the bandwidth of the physical network card under normal working is determined according to the maximum actual physical NIC bandwidth usage of the physical node (MaxVMPhyNet), that is, the physical node is working normally.
  • the traffic information needs to be described.
  • the maximum physical NIC bandwidth usage if the system does not limit the physical NIC bandwidth usage of the physical node, the maximum physical NIC bandwidth occupancy supported by the physical node may be used as the traffic information. The maximum actual physical NIC bandwidth usage, and then determine the traffic information under normal working.
  • the maximum physical NIC bandwidth occupied by the system for the physical node is taken as its maximum actual The physical NIC bandwidth usage, and then determine the traffic information under normal work.
  • the physical network card bandwidth usage of the physical node under normal working can be set to be lower than the maximum actual. 80% of the physical NIC bandwidth usage, that is, the physical NIC bandwidth usage that is less than 80% of the maximum actual physical NIC bandwidth usage is the traffic information under normal working of the physical node, and the actual physical NIC bandwidth of the physical node is determined.
  • the current physical network card bandwidth usage of the physical node is lower than 80% of the maximum actual physical NIC bandwidth of the physical node, if the current physical physics of a physical node If the NIC bandwidth usage is less than 80% of the maximum actual physical NIC bandwidth, you need to migrate the running virtual machine so that the physical node meets the actual network application requirements. If it is not lower, you do not need to The running virtual machine is migrated, so that the physical node continues to work normally and continues to monitor other physical nodes.
  • the virtual machine in the physical node needs to be migrated according to the network traffic occupation information of each virtual machine in the physical node.
  • the network traffic statistics information of the physical node meets the threshold information, for example, according to the traffic occupation information preset by each virtual machine, respectively, the difference between the network traffic occupation information and the traffic occupation information of each virtual machine is determined, that is, separately calculated.
  • the difference between the network traffic usage information and the traffic occupation information of each virtual machine is determined, and the difference is positive.
  • the difference indicates that the network traffic usage information of the virtual machine is greater than the preset traffic occupation information.
  • the network traffic occupies at least one virtual machine whose information is greater than its preset traffic occupation information, and performs migration on the at least one virtual machine.
  • the network traffic occupation information includes, but is not limited to, the unique identification information of the virtual machine and the current actual egress bandwidth occupancy
  • the current actual egress bandwidth occupancy includes but is not limited to the transmission packet rate of the virtual machine, Receive packet rate, etc.
  • the current actual physical NIC bandwidth usage in the network traffic usage information of a virtual machine is preset, and the maximum export bandwidth of the virtual machine is preset (MaxVMPhyNetUse). That is, the maximum value of the data transmitted by the virtual machine using the physical network card is defined. It should be noted that multiple virtual machines can be created in the same physical node.
  • the virtual machine whose positive value is positive is added to the virtual machine to be migrated, and the virtual machine to be migrated In the machine list, the at least one virtual machine that needs to be migrated is determined.
  • the principle of the determination is that the network traffic statistics of the physical node after the migration of the at least one virtual machine that needs to be migrated meets the threshold information, so that the physical node can continue to work normally.
  • the virtual machine to be migrated is used as the virtual machine to be migrated, and the migration is performed on the virtual machine to be migrated.
  • the embodiment in order to make the network traffic statistics information of the physical node after the virtual machine is migrated meet the threshold information, provides a method for determining the virtual machine to be migrated, and of course, other determination manners are not excluded.
  • the virtual machine list to be migrated it is determined whether the network traffic statistics information of the physical node after the virtual machine having the largest difference value meets the threshold information, that is, the virtual machine with the largest difference is calculated to migrate from the physical node to another physical node. Then, the network traffic statistics of the physical node meets the threshold information. If yes, the virtual machine with the largest difference is determined as the virtual machine to be migrated, and the virtual machine to be migrated is used as the virtual machine to be migrated.
  • the virtual machine to be migrated performs migration so that the network traffic statistics of the network node after migrating the virtual machine with the largest difference meets the threshold information for normal operation; if not, the virtual machine with the largest migration difference and the difference are determined. Whether the network traffic statistics of the physical node after the virtual machine with the second largest value meets the threshold information. That is, if it is determined that the network traffic statistics of the physical node after the virtual machine having the largest migration difference still does not satisfy the threshold information, the virtual machine with the largest difference and the virtual machine with the largest difference are migrated from the physical node to the other. After the physical node, the network traffic statistics of the physical node meet the threshold information.
  • the virtual machine with the largest difference and the virtual machine with the largest difference are determined as virtual machines to be migrated.
  • the virtual machine is configured as a virtual machine to be migrated, and the virtual machine to be migrated is migrated, so that the network traffic statistics of the network node after the virtual machine with the largest difference and the virtual machine with the largest difference meet the threshold information, so as to be normal. Work; if not, then so on, continue to follow the above calculations, the difference between the virtual machine order, from large to small, Then, the network traffic statistics threshold information of the physical node after the virtual machine is migrated is determined, and then the virtual machine to be migrated is determined, and the virtual machine to be migrated is used as the virtual machine to be migrated, and the virtual machine to be migrated is executed. migrate.
  • the virtual machine communication in the virtual machine list to be migrated is restricted, such as All the virtual machines in the migrating virtual machine list are arranged in descending order according to the above-mentioned differences.
  • the traffic of the at least one virtual machine in the sequence is limited.
  • the virtual machine to be migrated can be restricted.
  • the migration is performed on the at least one virtual machine to be migrated, so that the target physical node of the at least one virtual machine to be migrated needs to be determined, and network condition filtering is added.
  • the network traffic statistics of the physical node that is to be migrated to the at least one virtual machine to be migrated do not meet the threshold information, so the network traffic of other physical nodes is acquired before the migration is performed.
  • the statistics information is selected according to the network traffic statistics of the other physical nodes, and the physical node that meets the virtual machine migration condition is selected, so that the network traffic statistics of the physical node that is moved into the virtual machine still meet the threshold information, that is, the calculation moves into the at least one Whether the network traffic statistics of the physical node after the virtual machine is to be migrated meets the threshold information. If the physical node is lower than the target, the physical node can be used as the target physical node for performing the migration of the at least one virtual machine to be migrated, and the dynamic migration process of the OpenStack virtual machine is started. , but need to be dynamic on the virtual machine before starting Target physical node selection algorithm shift the process to optimize the migration of virtual machines can be moved to the correct physical node, after the end of the optimization, you can start the migration process to treat migrate virtual machines to perform the migration.
  • the OpenStack framework and OpenStack based on the network traffic statistics of the physical nodes and the network traffic occupation information of each virtual machine in the physical node, it is determined whether the virtual machines in the physical nodes need to be migrated, and the virtual machine can be dynamically monitored.
  • Network traffic usage Dynamically decide the migration of virtual machines and reduce the manual participation of system maintenance.
  • the OpenStack host selection algorithm can be modified. Increase the filtering conditions for the network and optimize the host selection algorithm when the virtual machine is migrated.
  • Embodiment 2 is a diagrammatic representation of Embodiment 1:
  • the virtual machine migration decision apparatus includes:
  • the obtaining module 1 is configured to obtain network traffic statistics information of the physical node, and network traffic occupation information of each virtual machine in the physical node;
  • the determining module 2 is configured to determine whether the network traffic statistics information of the physical node acquired by the acquiring module 1 meets the threshold information, and the threshold information is the traffic information under normal working of the physical node;
  • the determining module 3 is configured to determine, according to the traffic occupation information preset by each virtual machine, the difference between the network traffic occupation information and the traffic occupation information of each virtual machine, respectively, if the determining module 2 determines that the content is not satisfied;
  • the migration module 4 is configured to perform migration on the virtual machine according to the difference determined by the determination module 3.
  • the network traffic statistics information includes the current actual physical network card bandwidth usage of the physical node
  • the determining module 3 is further configured to determine the physical network card bandwidth usage of the physical node under normal working according to the maximum actual physical network card bandwidth usage of the physical node.
  • the judging module 2 is configured to determine whether the current actual physical network card bandwidth occupancy of the physical node satisfies the physical network card bandwidth occupancy of the physical node under normal working.
  • the determining module 2 is further configured to determine whether the difference determined by the determining module 3 is positive, and the migration module 4 is configured to perform migration on the at least one virtual machine whose positive value is positive.
  • the network traffic occupation information includes a current actual outlet bandwidth usage of the virtual machine;
  • the determining module 3 includes a setting submodule 31 and a computing submodule 32, wherein: the setting submodule 31 is configured to preset the maximum of each virtual machine respectively. The amount of the egress bandwidth; the calculation sub-module 32 is configured to calculate the current actual egress bandwidth usage of each virtual machine according to the maximum amount of the egress bandwidth preset by each virtual machine. The difference from the maximum amount of exit bandwidth.
  • the device further includes an adding module 5, the adding module 5 is configured to add a virtual machine with a positive difference to the virtual machine to be migrated; the determining module 3 is further configured to determine at least the virtual machine to be migrated, determining at least A virtual machine that needs to be migrated, so that the network traffic statistics of the physical node after the migration of at least one virtual machine that needs to be migrated meets the threshold information; the migration module 4 is configured to treat at least one virtual machine to be migrated as the virtual machine to be migrated, and treat Migrate the virtual machine to perform the migration.
  • the adding module 5 is configured to add a virtual machine with a positive difference to the virtual machine to be migrated
  • the determining module 3 is further configured to determine at least the virtual machine to be migrated, determining at least A virtual machine that needs to be migrated, so that the network traffic statistics of the physical node after the migration of at least one virtual machine that needs to be migrated meets the threshold information
  • the migration module 4 is configured to treat at least one virtual machine to be migrated as the virtual machine
  • the determining module 2 is further configured to determine, in the list of the virtual machines to be migrated, whether the network traffic statistics information of the physical node after the virtual machine having the largest difference value meets the threshold information, and the determining module 3 is configured to determine the module 2 If the determination is satisfied, the virtual machine with the largest difference is determined as the virtual machine that needs to be migrated. If not, the determination module 2 is further configured to determine the virtual machine with the largest migration difference and the physical entity after the virtual machine with the largest difference.
  • the determining module 3 is configured to determine, if the determining module 2 is satisfied, the virtual machine with the largest difference and the virtual machine with the largest difference as the virtual machine to be migrated, if If the method is not satisfied, the determining module 2 is further configured to continue to determine until the network traffic statistics of the physical node after the virtual machine is migrated meets the threshold information; the determining module 3 is configured to satisfy the determining by the determining module 2
  • the virtual machine of the condition is determined to be a virtual machine that needs to be migrated, and the condition is that the network traffic statistics of the physical node after the virtual machine is migrated are satisfied. Said threshold information.
  • the device further includes an arranging module 6, a limiting module 7, and a canceling module 8, wherein: the arranging module 6 is configured to arrange all the virtual machines in the virtual machine list to be migrated according to the difference from large to small.
  • the restriction module 7 is configured to limit the traffic of the at least one virtual machine after the arrangement module 6 is sequentially arranged; the cancellation module 8 is configured to cancel the unmigrated virtual machine if at least one of the virtual machines includes the unmigrated virtual machine Traffic restrictions.
  • the obtaining module 1 is further configured to obtain network traffic statistics information of other physical nodes
  • the device further includes a selecting module 9 configured to select, according to network traffic statistics information of other physical nodes acquired by the acquiring module 1, The physical node that meets the conditions for the virtual machine to move in, the network traffic statistics of the physical node that is moved into the virtual machine still meets the threshold information, and the selected physical node is the target physical node that performs the migration as at least one virtual machine.
  • Embodiment 3 is a diagrammatic representation of Embodiment 3
  • FIG. 3 is a schematic structural diagram of a virtual machine migration decision apparatus according to Embodiment 3 of the present invention.
  • this embodiment includes an OpenStack control node A and multiple OpenStack compute nodes Bn.
  • the OpenStack controller node A the network monitoring module 11 and the plug-in module 12 are added, and the network component module 13 and the migration module 15 are modified.
  • the data proxy module 22 is added, and the component module is modified. twenty three.
  • S1 The switch module running on the OpenStack compute node B1 needs to configure the sflow (a network monitoring technology) statistics collection function, and report the statistics to the proxy module 22 resident in the OpenStack compute node, and the proxy module of each OpenStack compute node.
  • sflow a network monitoring technology
  • the proxy module 22 uploads the processed statistical data to the network component module 13 in the OpenStack control node A by means of a message queue;
  • the network monitoring module 11 monitors network utilization on each OpenStack computing node in real time, and obtains statistical data through the rest API interface provided by the plug-in module 12;
  • the network monitoring module 11 determines whether there is a virtual machine to be migrated according to the statistics returned by the plug-in module 12. For example, the virtual machine migration decision method provided in the first embodiment; if present, the network monitoring module can invoke the rest API. Calling the network component module 13 to limit the flow of the virtual machine;
  • the network component module 13 controls the component module 23 on the OpenStack computing node to implement the current limiting of the virtual machine by means of message communication;
  • the component module 23 implements current limiting by means of an Openflow meter by invoking a command of the switch module 21;
  • the migration module 15 needs to make a selection according to the current network condition when selecting the target migration host, and needs to invoke the network component module 13 to acquire network data by means of the rest API;
  • the network component module 13 acquires statistical data from the database module 14;
  • the network component module 13 controls the component module 23 on the OpenStack computing node by means of message communication to implement the current limiting of the virtual machine.
  • the network traffic statistics information of the physical node and the network traffic occupation information of each virtual machine in the physical node it is determined whether the virtual machine needs to be migrated, and the network traffic occupancy of the virtual machine can be dynamically monitored, and the dynamic decision virtual Machine migration reduces manual participation in system maintenance.
  • the network traffic usage of the virtual machine is analyzed to determine whether to trigger dynamic migration of the virtual machine and dynamic control of the virtual machine network traffic.

Abstract

A decision-making method and apparatus for virtual machine migration. The method comprises: acquiring network traffic statistics information of a physical node and network traffic occupancy information of each virtual machine in the physical node; determining whether the network traffic statistics information of the physical node meets threshold information, the threshold information being traffic information of the physical node in a normal working state; if the network traffic statistics information of the physical node does not meet the threshold information, determining, according to preset traffic occupancy information of each virtual machine, a difference between the network traffic occupancy information and the preset traffic occupancy information of each virtual machine; and implementing virtual machine migration according to the difference.

Description

一种虚拟机迁移决策方法及装置Virtual machine migration decision method and device 技术领域Technical field
本文涉及云计算和网络虚拟化领域,尤其涉及一种虚拟机迁移决策方法及装置。This paper deals with the field of cloud computing and network virtualization, and in particular relates to a virtual machine migration decision method and device.
背景技术Background technique
随着云计算技术和网络虚拟化技术的发展,OpenStack(无中文技术术语)提供了一套部署云的操作平台或工具集,为云计算基础设施服务提供解决方案。OpenStack具有开源、互操作性、可移植性、易用性的特点,而且还提供开发的API(Application Programming Interface,应用程序编程接口),以便进行二次开发,能够方便地与其他系统进行集成,赢得业界越来越多的关注。With the development of cloud computing technology and network virtualization technology, OpenStack (no Chinese technical terminology) provides a set of operating platforms or tools to deploy the cloud to provide solutions for cloud computing infrastructure services. OpenStack is open source, interoperable, portable, and easy to use. It also provides an API (Application Programming Interface) for secondary development and easy integration with other systems. Win more and more attention in the industry.
虚拟机的迁移是指将源物理主机上运行的虚拟机操作系统及应用程序移动到目标物理主机上,并且在目标主机上能够正常运行。OpenStack自带虚拟机的迁移功能,允许一个正在运行的虚拟机实例从一个计算节点迁移到另一个计算节点,但是虚拟机的迁移需要操作人员手动的在界面操作,因操作人员在虚拟机迁移时无法了解全网的网络利用情况,不明确如何迁移才能达到最优的效果,可能会出现迁移后网络出现更加拥塞的局面,所以目前的实现无法实现动态监测、动态决策、动态迁移的目的。The migration of a virtual machine refers to moving the virtual machine operating system and applications running on the source physical host to the target physical host and running normally on the target host. The migration function of OpenStack's own virtual machine allows a running virtual machine instance to be migrated from one compute node to another, but the migration of the virtual machine requires the operator to manually operate the interface because the operator is in the virtual machine migration. It is impossible to understand the network utilization of the entire network. It is not clear how to migrate to achieve the optimal effect. The network may become more congested after migration. Therefore, the current implementation cannot achieve the purpose of dynamic monitoring, dynamic decision-making, and dynamic migration.
发明内容Summary of the invention
本发明实施例提供了一种虚拟机迁移决策方法及装置,解决了相关技术中因虚拟机迁移导致网络出现阻塞,且无法进行动态决策的问题。The embodiment of the invention provides a virtual machine migration decision method and device, which solves the problem that the network is blocked due to virtual machine migration and the dynamic decision cannot be made.
本发明实施例提供了一种虚拟机迁移决策方法,包括:The embodiment of the invention provides a virtual machine migration decision method, which includes:
获取物理节点的网络流量统计信息,以及所述物理节点内每个虚拟机的网络流量占用信息;Obtaining network traffic statistics information of the physical node, and network traffic occupation information of each virtual machine in the physical node;
判断所述物理节点的网络流量统计信息是否满足阈值信息,所述阈值信息为所述物理节点正常工作下的流量信息; Determining whether the network traffic statistics information of the physical node meets threshold information, where the threshold information is traffic information under normal working of the physical node;
若不满足,根据所述每个虚拟机预设的流量占用信息,分别确定所述每个虚拟机的网络流量占用信息与流量占用信息的差值;If not, the difference between the network traffic occupation information and the traffic occupation information of each virtual machine is determined according to the preset traffic occupation information of each virtual machine;
根据所述差值,对虚拟机执行迁移。Perform migration on the virtual machine based on the difference.
可选地,所述网络流量统计信息包括所述物理节点的当前实际物理网卡带宽占用量;Optionally, the network traffic statistics information includes a current actual physical network card bandwidth occupation of the physical node;
所述判断所述物理节点的网络流量统计信息是否满足阈值信息,所述阈值信息为所述物理节点正常工作下的流量信息包括:Determining whether the network traffic statistics information of the physical node meets the threshold information, where the threshold information is that the traffic information of the physical node under normal working includes:
根据所述物理节点的最大实际物理网卡带宽占用量,确定所述物理节点正常工作下的物理网卡带宽占用量;Determining, according to the maximum actual physical NIC bandwidth usage of the physical node, the bandwidth occupied by the physical NIC under normal working of the physical node;
判断所述物理节点的当前实际物理网卡带宽占用量是否满足所述物理节点正常工作下的物理网卡带宽占用量。Determining whether the current actual physical NIC bandwidth usage of the physical node meets the physical NIC bandwidth usage of the physical node under normal working.
可选地,所述根据所述差值,对虚拟机执行迁移包括:Optionally, performing the migration to the virtual machine according to the difference includes:
判断所述差值是否为正;Determining whether the difference is positive;
对所述差值为正的至少一个虚拟机执行迁移。Performing a migration on at least one virtual machine whose difference is positive.
可选地,所述网络流量占用信息包括所述虚拟机的当前实际出口带宽占用量;Optionally, the network traffic occupation information includes a current actual egress bandwidth occupation of the virtual machine;
所述根据所述每个虚拟机预设的流量占用信息,分别确定所述每个虚拟机的网络流量占用信息与流量占用信息的差值包括:Determining, according to the traffic occupation information preset by each virtual machine, the difference between the network traffic occupation information and the traffic occupation information of each virtual machine includes:
分别预设所述每个虚拟机的最大出口带宽量;Presetting the maximum amount of outlet bandwidth of each virtual machine respectively;
根据所述每个虚拟机预设的最大出口带宽量,分别计算所述每个虚拟机的当前实际出口带宽占用量与最大出口带宽量的差值。And calculating, according to the maximum amount of the egress bandwidth preset by each virtual machine, a difference between a current actual egress bandwidth occupancy of each virtual machine and a maximum egress bandwidth amount.
可选地,所述对所述差值为正的至少一个虚拟机执行迁移包括:Optionally, performing the migration on the at least one virtual machine whose positive value is positive includes:
将所述差值为正的虚拟机添加至待迁移虚拟机列表;Add the virtual machine whose positive value is positive to the virtual machine list to be migrated;
在所述待迁移虚拟机列表中,确定至少一个需要迁移的虚拟机,使迁移所述至少一个需要迁移的虚拟机后的物理节点的网络流量统计信息满足所述阈值信息;Determining at least one virtual machine to be migrated in the virtual machine list to be migrated, so that network traffic statistics of the physical node after the migration of the at least one virtual machine to be migrated meets the threshold information;
将所述至少一个需要迁移的虚拟机作为待迁移虚拟机,对所述待迁移虚 拟机执行迁移。Using the at least one virtual machine to be migrated as a virtual machine to be migrated, and the virtual machine to be migrated The machine is scheduled to perform the migration.
可选地,所述在所述待迁移虚拟机列表中,确定至少一个需要迁移的虚拟机,使迁移所述至少一个需要迁移的虚拟机后的物理节点的网络流量统计信息满足所述阈值信息包括:Optionally, the determining, in the list of the virtual machines to be migrated, the at least one virtual machine that needs to be migrated, so that the network traffic statistics of the physical node after the migration of the at least one virtual machine that needs to be migrated meets the threshold information. include:
在所述待迁移虚拟机列表中,判断迁移所述差值最大的虚拟机后的物理节点的网络流量统计信息是否满足所述阈值信息;Determining, in the list of the virtual machines to be migrated, whether the network traffic statistics of the physical node after the virtual machine having the largest difference is migrated meets the threshold information;
若满足,则将所述差值最大的虚拟机确定为需要迁移的虚拟机,若不满足,则判断迁移所述差值最大的虚拟机以及差值次大的虚拟机后的物理节点的网络流量统计信息是否满足所述阈值信息,若满足,则将所述差值最大的虚拟机以及差值次大的虚拟机确定为需要迁移的虚拟机,若不满足,则直到判断出迁移虚拟机后的物理节点的网络流量统计信息满足所述阈值信息,将所述虚拟机确定为需要迁移的虚拟机。If yes, determine the virtual machine with the largest difference as the virtual machine to be migrated. If not, determine the network of the physical node after the virtual machine with the largest difference and the virtual machine with the largest difference. Whether the traffic statistics information meets the threshold information, and if yes, the virtual machine with the largest difference and the virtual machine with the largest difference are determined as virtual machines to be migrated, and if not, the virtual machine is determined to be migrated. The network traffic statistics of the subsequent physical node satisfy the threshold information, and the virtual machine is determined to be a virtual machine that needs to be migrated.
可选地,在所述将所述至少一个需要迁移的虚拟机作为待迁移虚拟机,对所述待迁移虚拟机执行迁移之前,所述方法还包括:Optionally, before the migrating the virtual machine to be migrated, the method further includes:
将所述待迁移虚拟机列表中的所有虚拟机按照所述差值由大到小依次排列;All the virtual machines in the virtual machine list to be migrated are arranged in descending order according to the difference;
对所述依次排列后的前至少一个虚拟机的流量进行限制;Limiting the traffic of the at least one virtual machine in the prior arrangement;
在所述将所述至少一个需要迁移的虚拟机作为待迁移虚拟机,对所述待迁移虚拟机执行迁移之后,所述方法还包括:After the performing, by using the virtual machine to be migrated as the virtual machine to be migrated, performing the migration on the virtual machine to be migrated, the method further includes:
若所述前至少一个虚拟机中包括未迁移的虚拟机,则取消对所述未迁移虚拟机的流量限制。If the at least one virtual machine includes the unmigrated virtual machine, the traffic restriction on the unmigrated virtual machine is cancelled.
可选地,在所述根据所述差值,对虚拟机执行迁移之前,所述方法还包括:Optionally, before the performing the migration on the virtual machine according to the difference, the method further includes:
获取其他物理节点的网络流量统计信息;Obtain network traffic statistics of other physical nodes;
根据所述其他物理节点的网络流量统计信息,选择满足虚拟机迁入条件的物理节点,使迁入虚拟机后的物理节点的网络流量统计信息仍满足所述阈值信息,将所述选择的物理节点作为所述至少一个虚拟机执行迁移的目标物理节点。 And selecting, according to the network traffic statistics information of the other physical nodes, the physical node that meets the virtual machine migration condition, so that the network traffic statistics information of the physical node that is moved into the virtual machine still meets the threshold information, and the selected physical The node performs the migration as the target physical node of the at least one virtual machine.
本发明实施例还提供一种虚拟机迁移决策装置,包括:The embodiment of the invention further provides a virtual machine migration decision device, comprising:
获取模块,设置为获取物理节点的网络流量统计信息,以及所述物理节点内每个虚拟机的网络流量占用信息;Obtaining a module, configured to obtain network traffic statistics information of the physical node, and network traffic occupation information of each virtual machine in the physical node;
判断模块,设置为判断所述获取模块获取的物理节点的网络流量统计信息是否满足阈值信息,所述阈值信息为所述物理节点正常工作下的流量信息;The determining module is configured to determine whether the network traffic statistics information of the physical node acquired by the acquiring module meets threshold information, where the threshold information is traffic information under normal working of the physical node;
确定模块,设置为若所述判断模块判断为不满足,根据所述每个虚拟机预设的流量占用信息,分别确定所述每个虚拟机的网络流量占用信息与流量占用信息的差值;以及a determining module, configured to determine, according to the traffic occupation information preset by each virtual machine, a difference between the network traffic occupation information and the traffic occupation information of each virtual machine, respectively, if the determining module determines that the content is not satisfied; as well as
迁移模块,设置为根据所述确定模块确定的差值,对虚拟机执行迁移。The migration module is configured to perform migration on the virtual machine according to the difference determined by the determining module.
可选地,所述网络流量统计信息包括所述物理节点的当前实际物理网卡带宽占用量;Optionally, the network traffic statistics information includes a current actual physical network card bandwidth occupation of the physical node;
所述确定模块还设置为根据所述物理节点的最大实际物理网卡带宽占用量,确定所述物理节点正常工作下的物理网卡带宽占用量;The determining module is further configured to determine, according to a maximum actual physical NIC bandwidth occupancy of the physical node, a bandwidth occupied by the physical NIC under normal working of the physical node;
所述判断模块是设置为判断所述物理节点的当前实际物理网卡带宽占用量是否满足所述物理节点正常工作下的物理网卡带宽占用量。The determining module is configured to determine whether the current actual physical network card bandwidth occupancy of the physical node meets the physical network card bandwidth occupancy of the physical node under normal working.
可选地,所述判断模块还设置为判断所述确定模块确定的差值是否为正;Optionally, the determining module is further configured to determine whether the difference determined by the determining module is positive;
所述迁移模块是设置为对所述差值为正的至少一个虚拟机执行迁移。The migration module is configured to perform migration on the at least one virtual machine whose difference is positive.
可选地,所述网络流量占用信息包括所述虚拟机的当前实际出口带宽占用量;Optionally, the network traffic occupation information includes a current actual egress bandwidth occupation of the virtual machine;
所述确定模块包括:The determining module includes:
设置子模块,设置为分别预设所述每个虚拟机的最大出口带宽量;以及Setting a sub-module, which is set to respectively preset a maximum amount of exit bandwidth of each of the virtual machines;
计算子模块,设置为根据所述每个虚拟机预设的最大出口带宽量,分别计算所述每个虚拟机的当前实际出口带宽占用量与最大出口带宽量的差值。The calculating sub-module is configured to calculate, according to the maximum amount of the egress bandwidth preset by each virtual machine, a difference between the current actual egress bandwidth occupancy of each virtual machine and the maximum egress bandwidth.
可选地,所述装置还包括:Optionally, the device further includes:
添加模块,设置为将所述差值为正的虚拟机添加至待迁移虚拟机列表;Add a module, and set to add the virtual machine whose positive value is positive to the virtual machine list to be migrated;
所述确定模块还设置为在所述待迁移虚拟机列表中,确定至少一个需要迁移的虚拟机,使迁移所述至少一个需要迁移的虚拟机后的物理节点的网络 流量统计信息满足所述阈值信息;The determining module is further configured to: in the list of the virtual machines to be migrated, determine at least one virtual machine that needs to be migrated, so that the network of the physical node after the at least one virtual machine that needs to be migrated is migrated The traffic statistics information meets the threshold information;
所述迁移模块是设置为将所述至少一个需要迁移的虚拟机作为待迁移虚拟机,对所述待迁移虚拟机执行迁移。The migration module is configured to perform the migration on the virtual machine to be migrated by using the at least one virtual machine to be migrated as a virtual machine to be migrated.
可选地,所述判断模块还设置为在所述待迁移虚拟机列表中,判断迁移所述差值最大的虚拟机后的物理节点的网络流量统计信息是否满足所述阈值信息;Optionally, the determining module is further configured to determine, in the list of the virtual machines to be migrated, whether the network traffic statistics information of the physical node after the virtual machine having the largest difference value is migrated meets the threshold information;
所述确定模块是设置为若所述判断模块判断为满足,则将所述差值最大的虚拟机确定为需要迁移的虚拟机,若不满足,则所述判断模块还设置为判断迁移所述差值最大的虚拟机以及差值次大的虚拟机后的物理节点的网络流量统计信息是否满足所述阈值信息,所述确定模块是设置为若所述判断模块判断为满足,则将所述差值最大的虚拟机以及差值次大的虚拟机确定为需要迁移的虚拟机,若不满足,则所述判断模块还设置为继续判断,直到判断出迁移虚拟机后的物理节点的网络流量统计信息满足所述阈值信息,所述确定模块是设置为将所述判断模块判断出的满足条件的虚拟机确定为需要迁移的虚拟机,所述条件为迁移虚拟机后的物理节点的网络流量统计信息满足所述阈值信息。The determining module is configured to determine, if the determining module is satisfied, the virtual machine with the largest difference as the virtual machine that needs to be migrated, and if not, the determining module is further configured to determine that the migration is performed. Whether the network traffic statistics of the physical node having the largest difference and the virtual node having the second largest difference satisfy the threshold information, and the determining module is configured to: if the determining module determines that the content is satisfied, the The virtual machine with the largest difference and the virtual machine with the largest difference are determined to be virtual machines that need to be migrated. If not, the judgment module is further configured to continue to judge until the network traffic of the physical node after the virtual machine is migrated is determined. The statistical information meets the threshold information, and the determining module is configured to determine a virtual machine that meets the condition determined by the determining module as a virtual machine that needs to be migrated, where the condition is a network traffic of a physical node after the virtual machine is migrated. The statistical information satisfies the threshold information.
可选地,所述装置还包括:Optionally, the device further includes:
排列模块,设置为将所述待迁移虚拟机列表中的所有虚拟机按照所述差值由大到小依次排列;The arranging module is configured to arrange all the virtual machines in the virtual machine list to be migrated according to the difference from large to small;
限制模块,设置为对所述排列模块依次排列后的前至少一个虚拟机的流量进行限制;以及a limiting module, configured to limit traffic of the at least one virtual machine that is sequentially arranged by the arranging module; and
取消模块,设置为若所述前至少一个虚拟机中包括未迁移的虚拟机,则取消对所述未迁移虚拟机的流量限制。The module is canceled, and if the at least one virtual machine includes the unmigrated virtual machine, the traffic restriction on the unmigrated virtual machine is cancelled.
可选地,所述获取模块还设置为获取其他物理节点的网络流量统计信息;Optionally, the acquiring module is further configured to obtain network traffic statistics information of other physical nodes;
所述装置还包括:选择模块,其设置为根据所述获取模块获取的其他物理节点的网络流量统计信息,选择满足虚拟机迁入条件的物理节点,使迁入虚拟机后的物理节点的网络流量统计信息仍满足所述阈值信息,将所述选择的物理节点作为所述至少一个虚拟机执行迁移的目标物理节点。 The device further includes: a selecting module, configured to select, according to network traffic statistics information of other physical nodes acquired by the acquiring module, a physical node that satisfies a virtual machine moving in condition, and a network of the physical node that moves into the virtual machine The traffic statistics information still satisfies the threshold information, and the selected physical node is used as the target physical node of the at least one virtual machine to perform migration.
本发明实施例还提供一种计算机可读存储介质,存储有程序指令,当该程序指令被执行时可实现上述方法。The embodiment of the invention further provides a computer readable storage medium storing program instructions, which can be implemented when the program instructions are executed.
本发明实施例提供一种虚拟机迁移决策方法及装置,结合OpenStack框架和OpenStack提供的服务,根据物理节点的网络流量统计信息,以及物理节点内每个虚拟机的网络流量占用信息,判断是否需要对物理节点内的虚拟机进行迁移,克服了OpenStack已有迁移方式的不足,能动态监测虚拟机的网络流量占用情况,动态决策虚拟机的迁移,减少系统维护的人工参与度。通过获取并存储网络流量统计信息,实现对虚拟机网络流量占用情况的分析来决策是否触发虚拟机的动态迁移,以及对虚拟机网络流量的动态管控,此外,还可通过修改原有OpenStack主机选择算法,增加对网络的过滤条件,优化了虚拟机迁移时主机选择算法。The embodiment of the invention provides a virtual machine migration decision method and device, and combines the services provided by the OpenStack framework and the OpenStack to determine whether it is needed according to the network traffic statistics of the physical node and the network traffic occupation information of each virtual machine in the physical node. The migration of virtual machines in physical nodes overcomes the shortcomings of OpenStack's existing migration methods. It can dynamically monitor the network traffic usage of virtual machines, dynamically determine the migration of virtual machines, and reduce the manual participation of system maintenance. By obtaining and storing network traffic statistics, the network traffic usage of the virtual machine is analyzed to determine whether to trigger the dynamic migration of the virtual machine and the dynamic control of the virtual machine network traffic. In addition, the original OpenStack host selection can be modified. The algorithm increases the filtering conditions for the network and optimizes the host selection algorithm when the virtual machine is migrated.
附图概述BRIEF abstract
图1为本发明实施例一提供的虚拟机迁移决策方法的流程图;FIG. 1 is a flowchart of a virtual machine migration decision method according to Embodiment 1 of the present invention;
图2为本发明实施例二提供的虚拟机迁移决策装置的结构示意图;2 is a schematic structural diagram of a virtual machine migration decision apparatus according to Embodiment 2 of the present invention;
图3为本发明实施例三提供的虚拟机迁移决策装置的结构示意图。FIG. 3 is a schematic structural diagram of a virtual machine migration decision apparatus according to Embodiment 3 of the present invention.
本发明的实施方式Embodiments of the invention
下面将结合本发明实施例中的附图,对本发明实施例中的技术方案进行描述。The technical solutions in the embodiments of the present invention will be described below with reference to the accompanying drawings in the embodiments of the present invention.
实施例一:Embodiment 1:
图1为本发明实施例一提供的虚拟机迁移决策方法的流程图,如图1所示,该虚拟机迁移决策方法包括:FIG. 1 is a flowchart of a virtual machine migration decision method according to Embodiment 1 of the present invention. As shown in FIG. 1 , the virtual machine migration decision method includes:
S101:获取物理节点的网络流量统计信息,以及物理节点内每个虚拟机的网络流量占用信息;S101: Obtain network traffic statistics information of the physical node, and network traffic occupation information of each virtual machine in the physical node;
S102:判断物理节点的网络流量统计信息是否满足阈值信息,阈值信息为物理节点正常工作下的流量信息; S102: Determine whether the network traffic statistics information of the physical node meets the threshold information, where the threshold information is traffic information under normal working of the physical node;
S103:若不满足,根据每个虚拟机预设的流量占用信息,分别确定每个虚拟机的网络流量占用信息与流量占用信息的差值;S103: If not, according to the traffic occupation information preset by each virtual machine, determine the difference between the network traffic occupation information and the traffic occupation information of each virtual machine respectively;
S104:根据差值,对虚拟机执行迁移。S104: Perform migration on the virtual machine according to the difference.
为了动态监测虚拟机的网络流量占用情况,动态决策虚拟机的迁移,可以获取物理节点的网络流量统计信息,以及物理节点内每个虚拟机的网络流量占用信息,该获取动作可以为实时获取,也可以为定时获取,根据物理节点的网络流量统计信息,以及物理节点内每个虚拟机的网络流量占用信息,判断是否需要对物理节点内的虚拟机进行迁移,从而减少系统维护的人工参与度。该物理节点可以为OpenStack计算节点,即物理主机,针对OpenStack计算节点是否需要迁移其内的虚拟机,可以由OpenStack控制节点对OpenStack计算节点进行决策,即OpenStack控制节点根据OpenStack计算节点上报的网络流量统计信息,判断是否对OpenStack计算节点中的虚拟机进行迁移。In order to dynamically monitor the network traffic usage of the virtual machine and dynamically determine the migration of the virtual machine, the network traffic statistics of the physical node and the network traffic occupation information of each virtual machine in the physical node can be obtained, and the obtaining action can be obtained in real time. It can also be used for timing acquisition. According to the network traffic statistics of the physical node and the network traffic occupation information of each virtual machine in the physical node, it is determined whether the virtual machine in the physical node needs to be migrated, thereby reducing the manual participation of the system maintenance. . The physical node may be an OpenStack computing node, that is, a physical host, and whether the OpenStack computing node needs to migrate the virtual machine therein, and the OpenStack control node may make a decision on the OpenStack computing node, that is, the OpenStack control node reports the network traffic reported by the node according to the OpenStack. Statistics to determine whether to migrate virtual machines in OpenStack compute nodes.
在本实施例中,根据物理节点的网络流量统计信息,判断是否需要对该物理节点内的虚拟机进行迁移,以使得物理节点的网络流量统计信息低于阈值信息,该阈值信息为该物理节点正常工作下的流量信息,使其满足实际应用要求。该网络流量统计信息包括但不局限于物理主机的标识信息、当前实际物理网卡带宽占用量等,该当前实际物理网卡带宽占用量包括但不局限于物理节点的物理端口发送数据包速率、实际可处理的网络吞吐量等。在判断是否需要对物理节点内的虚拟机执行迁移时,根据该物理节点的最大实际物理网卡带宽占用量(MaxVMPhyNet),确定其正常工作下的物理网卡带宽占用量,即该物理节点正常工作下的流量信息,需要说明的是,针对该最大实际物理网卡带宽占用量,若系统没有限制该物理节点的物理网卡带宽占用量,则可以将该物理节点所支持的最大物理网卡带宽占用量作为其最大实际物理网卡带宽占用量,然后确定正常工作下的流量信息,若系统限制了该物理节点的物理网卡带宽占用量,则将系统为该物理节点设置的最大物理网卡带宽占用量作为其最大实际物理网卡带宽占用量,然后确定正常工作下的流量信息。在实际应用中,为了使物理节点满足对网络流量占用情况的要求,可选地,可以将物理节点正常工作下的物理网卡带宽占用量设置为低于最大实际 物理网卡带宽占用量的80%,即低于最大实际物理网卡带宽占用量的80%的物理网卡带宽占用量均为该物理节点正常工作下的流量信息,判断该物理节点的当前实际物理网卡带宽占用量(CurrentPhyNetUse)是否满足其正常工作下的物理网卡带宽占用量,即物理节点的当前实际物理网卡带宽占用量低于其最大实际物理网卡带宽量的80%,若一物理节点的当前实际物理网卡带宽占用量低于其最大实际物理网卡带宽量的80%,则需要对其内运行的虚拟机进行迁移,以使得该物理节点满足实际网络应用需求,如果不低于,则不需要对其内运行的虚拟机进行迁移,使该物理节点继续正常工作,继续监测其他物理节点。In this embodiment, the network traffic statistics of the physical node are used to determine whether the virtual machine in the physical node needs to be migrated, so that the network traffic statistics of the physical node are lower than the threshold information, and the threshold information is the physical node. The traffic information under normal work makes it meet the actual application requirements. The network traffic statistics include, but are not limited to, the physical host's identification information, the current actual physical network card bandwidth usage, and the like. The current actual physical network card bandwidth usage includes, but is not limited to, the physical port of the physical node, the data packet rate, and the actual Processing network throughput, etc. When determining whether the virtual machine in the physical node needs to be migrated, the bandwidth of the physical network card under normal working is determined according to the maximum actual physical NIC bandwidth usage of the physical node (MaxVMPhyNet), that is, the physical node is working normally. The traffic information needs to be described. For the maximum actual physical NIC bandwidth usage, if the system does not limit the physical NIC bandwidth usage of the physical node, the maximum physical NIC bandwidth occupancy supported by the physical node may be used as the traffic information. The maximum actual physical NIC bandwidth usage, and then determine the traffic information under normal working. If the system limits the physical NIC bandwidth usage of the physical node, the maximum physical NIC bandwidth occupied by the system for the physical node is taken as its maximum actual The physical NIC bandwidth usage, and then determine the traffic information under normal work. In practical applications, in order to make the physical node meet the requirements for network traffic occupation, the physical network card bandwidth usage of the physical node under normal working can be set to be lower than the maximum actual. 80% of the physical NIC bandwidth usage, that is, the physical NIC bandwidth usage that is less than 80% of the maximum actual physical NIC bandwidth usage is the traffic information under normal working of the physical node, and the actual physical NIC bandwidth of the physical node is determined. Whether the current physical network card bandwidth usage of the physical node is lower than 80% of the maximum actual physical NIC bandwidth of the physical node, if the current physical physics of a physical node If the NIC bandwidth usage is less than 80% of the maximum actual physical NIC bandwidth, you need to migrate the running virtual machine so that the physical node meets the actual network application requirements. If it is not lower, you do not need to The running virtual machine is migrated, so that the physical node continues to work normally and continues to monitor other physical nodes.
在本实施例中,若判断出一物理节点的网络流量统计信息不满足阈值信息,则需要根据该物理节点内的每个虚拟机的网络流量占用信息,对该物理节点内的虚拟机进行迁移,以使得该物理节点的网络流量统计信息满足阈值信息,如根据每个虚拟机预设的流量占用信息,分别确定每个虚拟机的网络流量占用信息与流量占用信息的差值,即分别计算每个虚拟机的网络流量占用信息与流量占用信息的差值,依次判断该差值是否为正,该差值为正表示该虚拟机的网络流量占用信息大于其预设的流量占用信息,得到网络流量占用信息大于其预设的流量占用信息的至少一个虚拟机,对该至少一个虚拟机执行迁移。In this embodiment, if it is determined that the network traffic statistics information of a physical node does not satisfy the threshold information, the virtual machine in the physical node needs to be migrated according to the network traffic occupation information of each virtual machine in the physical node. The network traffic statistics information of the physical node meets the threshold information, for example, according to the traffic occupation information preset by each virtual machine, respectively, the difference between the network traffic occupation information and the traffic occupation information of each virtual machine is determined, that is, separately calculated. The difference between the network traffic usage information and the traffic occupation information of each virtual machine is determined, and the difference is positive. The difference indicates that the network traffic usage information of the virtual machine is greater than the preset traffic occupation information. The network traffic occupies at least one virtual machine whose information is greater than its preset traffic occupation information, and performs migration on the at least one virtual machine.
在上述技术方案中,该网络流量占用信息包括但不局限于虚拟机的唯一标识信息、当前实际出口带宽占用量,该当前实际出口带宽占用量包括但不局限于虚拟机的发送数据包速率、接收数据包速率等。针对物理节点内的每个虚拟机,在通过OpenStack创建时,针对一虚拟机的网络流量占用信息中的当前实际物理网卡带宽占用量,分别预设该虚拟机的最大出口带宽量(MaxVMPhyNetUse),即定义该虚拟机利用物理网卡传输数据的最大值,需要说明的是,在同一物理节点内可以创建多个虚拟机,在定义每个虚拟机的最大出口带宽量时,需要保证同一物理节点内的所有虚拟机的最大出口带宽量之和不能超过该物理节点的物理带宽阈值。在实际应用中,需要保证该虚拟机的当前实际出口带宽占用量(CurrentVMPhyNetUse)不得超过其定义的最大出口带宽量,当判断出一物理节点的网络流量统计信息不满足阈值信 息时,根据每个虚拟机预设的最大出口带宽量,分别判断该物理节点内的每个虚拟机的网络流量占用信息是否超过其预设的最大出口带宽量,即分别计算每个虚拟机的当前实际出口带宽占用量与最大出口带宽量的差值,根据该差值,对虚拟机执行迁移,以使得该物理节点的网络流量统计信息满足阈值信息。In the foregoing technical solution, the network traffic occupation information includes, but is not limited to, the unique identification information of the virtual machine and the current actual egress bandwidth occupancy, and the current actual egress bandwidth occupancy includes but is not limited to the transmission packet rate of the virtual machine, Receive packet rate, etc. For each virtual machine in a physical node, when creating through OpenStack, the current actual physical NIC bandwidth usage in the network traffic usage information of a virtual machine is preset, and the maximum export bandwidth of the virtual machine is preset (MaxVMPhyNetUse). That is, the maximum value of the data transmitted by the virtual machine using the physical network card is defined. It should be noted that multiple virtual machines can be created in the same physical node. When defining the maximum amount of egress bandwidth of each virtual machine, it is necessary to ensure the same physical node. The sum of the maximum amount of egress bandwidth for all virtual machines cannot exceed the physical bandwidth threshold of the physical node. In practical applications, it is necessary to ensure that the current actual bandwidth consumption of the virtual machine (CurrentVMPhyNetUse) cannot exceed the maximum amount of the exported bandwidth. When it is determined that the network traffic statistics of a physical node do not meet the threshold value. During the interval, according to the maximum amount of the outlet bandwidth preset by each virtual machine, it is determined whether the network traffic occupation information of each virtual machine in the physical node exceeds the preset maximum outlet bandwidth amount, that is, each virtual machine is separately calculated. The difference between the current actual egress bandwidth occupancy and the maximum egress bandwidth is based on the difference, and the virtual machine is migrated so that the network traffic statistics of the physical node meet the threshold information.
在本实施例中,当分别确定了每个虚拟机的网络流量占用信息与流量占用信息的差值后,将其中差值为正的虚拟机添加至待迁移虚拟机列表,在该待迁移虚拟机列表中,确定至少一个需要迁移的虚拟机,该确定的原则是使得迁移该至少一个需要迁移的虚拟机后的物理节点的网络流量统计信息满足阈值信息,使该物理节点能够继续正常工作,将该至少一个需要迁移的虚拟机作为待迁移虚拟机,对该待迁移虚拟机执行迁移。In this embodiment, after the difference between the network traffic occupation information and the traffic occupation information of each virtual machine is determined, the virtual machine whose positive value is positive is added to the virtual machine to be migrated, and the virtual machine to be migrated In the machine list, the at least one virtual machine that needs to be migrated is determined. The principle of the determination is that the network traffic statistics of the physical node after the migration of the at least one virtual machine that needs to be migrated meets the threshold information, so that the physical node can continue to work normally. The virtual machine to be migrated is used as the virtual machine to be migrated, and the migration is performed on the virtual machine to be migrated.
可选地,在上述技术方案中,为了使得迁移虚拟机后的物理节点的网络流量统计信息满足阈值信息,本实施例提供一种待迁移虚拟机的确定方式,当然也不排除其他确定方式,在该待迁移虚拟机列表中,判断迁移差值最大的虚拟机后的物理节点的网络流量统计信息是否满足阈值信息,即计算将该差值最大的虚拟机从该物理节点迁移至其他物理节点后,该物理节点的网络流量统计信息是否满足阈值信息,若满足,则将该差值最大的虚拟机确定为需要迁移的虚拟机,将该需要迁移的虚拟机作为待迁移虚拟机,只对该待迁移虚拟机执行迁移,以使得迁移该差值最大的虚拟机后的网络节点的网络流量统计信息满足阈值信息,以便正常工作;如果不满足,则判断迁移差值最大的虚拟机以及差值次大的虚拟机后的物理节点的网络流量统计信息是否满足阈值信息,即若判断出迁移差值最大的虚拟机后的物理节点的网络流量统计信息仍然不满足阈值信息,则计算将差值最大的虚拟机以及差值次大的虚拟机从该物理节点迁移至其他物理节点后,该物理节点的网络流量统计信息是否满足阈值信息,若满足,则将该差值最大的虚拟机以及差值次大的虚拟机确定为需要迁移的虚拟机,将该需要迁移的虚拟机作为待迁移虚拟机,对该待迁移虚拟机执行迁移,以使得迁移该差值最大的虚拟机以及差值次大的虚拟机后的网络节点的网络流量统计信息满足阈值信息,以便正常工作;如果不满足,则以此类推,继续按照上述计算的差值由大到小的虚拟机顺序, 再进行判断,直到判断出迁移虚拟机后的物理节点的网络流量统计信息阈值信息,进而确定需要迁移的虚拟机,将该需要迁移的虚拟机作为待迁移虚拟机,对该待迁移虚拟机执行迁移。Optionally, in the foregoing technical solution, in order to make the network traffic statistics information of the physical node after the virtual machine is migrated meet the threshold information, the embodiment provides a method for determining the virtual machine to be migrated, and of course, other determination manners are not excluded. In the virtual machine list to be migrated, it is determined whether the network traffic statistics information of the physical node after the virtual machine having the largest difference value meets the threshold information, that is, the virtual machine with the largest difference is calculated to migrate from the physical node to another physical node. Then, the network traffic statistics of the physical node meets the threshold information. If yes, the virtual machine with the largest difference is determined as the virtual machine to be migrated, and the virtual machine to be migrated is used as the virtual machine to be migrated. The virtual machine to be migrated performs migration so that the network traffic statistics of the network node after migrating the virtual machine with the largest difference meets the threshold information for normal operation; if not, the virtual machine with the largest migration difference and the difference are determined. Whether the network traffic statistics of the physical node after the virtual machine with the second largest value meets the threshold information. That is, if it is determined that the network traffic statistics of the physical node after the virtual machine having the largest migration difference still does not satisfy the threshold information, the virtual machine with the largest difference and the virtual machine with the largest difference are migrated from the physical node to the other. After the physical node, the network traffic statistics of the physical node meet the threshold information. If yes, the virtual machine with the largest difference and the virtual machine with the largest difference are determined as virtual machines to be migrated. The virtual machine is configured as a virtual machine to be migrated, and the virtual machine to be migrated is migrated, so that the network traffic statistics of the network node after the virtual machine with the largest difference and the virtual machine with the largest difference meet the threshold information, so as to be normal. Work; if not, then so on, continue to follow the above calculations, the difference between the virtual machine order, from large to small, Then, the network traffic statistics threshold information of the physical node after the virtual machine is migrated is determined, and then the virtual machine to be migrated is determined, and the virtual machine to be migrated is used as the virtual machine to be migrated, and the virtual machine to be migrated is executed. migrate.
可选地,在本实施例中,为了保障虚拟机迁移过程中对网络流量的要求,在对待迁移虚拟机执行迁移之前,对待迁移虚拟机列表中的虚拟机通信进行限流,如将该待迁移虚拟机列表中的所有虚拟机按照上述计算的差值由大到小依次排列,对依次排列后的前至少一个虚拟机的流量进行限制,对于限流措施,如可以限制该待迁移虚拟机列表中依次排列后的前N(N=1,2,3,…)个虚拟机的流量,其目的是使得物理节点在虚拟机迁移过程中的当前实际物理网卡带宽占用量低于其最大物理网卡带宽量的70%。虚拟机迁移结束后,若之前限制流量的前N个虚拟机都被迁移,则不需要取消流量限制,若之前限制流量的前N个虚拟机中包括未迁移的虚拟机,则需要取消对未迁移虚拟机的流量限制。Optionally, in this embodiment, in order to ensure network traffic requirements during the virtual machine migration process, before the migration of the virtual machine to be migrated, the virtual machine communication in the virtual machine list to be migrated is restricted, such as All the virtual machines in the migrating virtual machine list are arranged in descending order according to the above-mentioned differences. The traffic of the at least one virtual machine in the sequence is limited. For the traffic limiting measures, the virtual machine to be migrated can be restricted. The traffic of the first N (N=1, 2, 3, ...) virtual machines in the list is arranged in such a way that the actual physical NIC bandwidth consumption of the physical node during the virtual machine migration process is lower than its maximum physics. 70% of the amount of NIC bandwidth. After the virtual machine migration is complete, if the first N virtual machines that previously restricted traffic are migrated, you do not need to cancel the traffic restriction. If the first N virtual machines that previously restricted traffic include unmigrated virtual machines, you need to cancel the Migrate virtual machine traffic limits.
可选地,在本实施例中,当确定了至少一个待迁移虚拟机后,对该至少一个待迁移虚拟机执行迁移,因此需要确定至少一个待迁移虚拟机的目标物理节点,增加网络条件过滤,防止将该至少一个待迁移虚拟机执行迁移后,使得迁入该至少一个待迁移虚拟机的物理节点的网络流量统计信息不满足阈值信息,故在执行迁移之前,获取其他物理节点的网络流量统计信息,根据其他物理节点的网络流量统计信息,选择满足虚拟机迁入条件的物理节点,使迁入虚拟机后的物理节点的网络流量统计信息仍满足阈值信息,即计算迁入该至少一个待迁移虚拟机后的物理节点的网络流量统计信息是否满足阈值信息,若低于,则可以将该物理节点作为该至少一个待迁移虚拟机执行迁移的目标物理节点,启动OpenStack虚拟机动态迁移流程,但是在启动之前需要对虚拟机动态迁移流程中的目标物理节点的选择算法进行优化,使待迁移虚拟机可以迁入正确的物理节点,优化结束后,即可启动迁移流程对待迁移虚拟机执行迁移。Optionally, in this embodiment, after the at least one virtual machine to be migrated is determined, the migration is performed on the at least one virtual machine to be migrated, so that the target physical node of the at least one virtual machine to be migrated needs to be determined, and network condition filtering is added. After the migration of the at least one virtual machine to be migrated is performed, the network traffic statistics of the physical node that is to be migrated to the at least one virtual machine to be migrated do not meet the threshold information, so the network traffic of other physical nodes is acquired before the migration is performed. The statistics information is selected according to the network traffic statistics of the other physical nodes, and the physical node that meets the virtual machine migration condition is selected, so that the network traffic statistics of the physical node that is moved into the virtual machine still meet the threshold information, that is, the calculation moves into the at least one Whether the network traffic statistics of the physical node after the virtual machine is to be migrated meets the threshold information. If the physical node is lower than the target, the physical node can be used as the target physical node for performing the migration of the at least one virtual machine to be migrated, and the dynamic migration process of the OpenStack virtual machine is started. , but need to be dynamic on the virtual machine before starting Target physical node selection algorithm shift the process to optimize the migration of virtual machines can be moved to the correct physical node, after the end of the optimization, you can start the migration process to treat migrate virtual machines to perform the migration.
结合OpenStack框架和OpenStack提供的服务,根据物理节点的网络流量统计信息,以及物理节点内每个虚拟机的网络流量占用信息,判断是否需要对物理节点内的虚拟机进行迁移,能动态监测虚拟机的网络流量占用情况, 动态决策虚拟机的迁移,减少系统维护的人工参与度。通过获取并存储网络流量统计信息实现,实现对虚拟机网络流量占用情况的分析来决策是否触发虚拟机的动态迁移,以及对虚拟机网络流量的动态管控,此外,还可通过修改OpenStack主机选择算法,增加对网络的过滤条件,优化了虚拟机迁移时主机选择算法。Combining the services provided by the OpenStack framework and OpenStack, based on the network traffic statistics of the physical nodes and the network traffic occupation information of each virtual machine in the physical node, it is determined whether the virtual machines in the physical nodes need to be migrated, and the virtual machine can be dynamically monitored. Network traffic usage, Dynamically decide the migration of virtual machines and reduce the manual participation of system maintenance. By obtaining and storing network traffic statistics, the network traffic usage of the virtual machine is analyzed to determine whether to trigger dynamic migration of the virtual machine and dynamic control of the virtual machine network traffic. In addition, the OpenStack host selection algorithm can be modified. Increase the filtering conditions for the network and optimize the host selection algorithm when the virtual machine is migrated.
实施例二:Embodiment 2:
图2为本发明实施例二提供的虚拟机迁移决策装置的结构示意图,如图2所示,该虚拟机迁移决策装置包括:2 is a schematic structural diagram of a virtual machine migration decision apparatus according to Embodiment 2 of the present invention. As shown in FIG. 2, the virtual machine migration decision apparatus includes:
获取模块1,设置为获取物理节点的网络流量统计信息,以及所述物理节点内每个虚拟机的网络流量占用信息;The obtaining module 1 is configured to obtain network traffic statistics information of the physical node, and network traffic occupation information of each virtual machine in the physical node;
判断模块2,设置为判断获取模块1获取的物理节点的网络流量统计信息是否满足阈值信息,阈值信息为物理节点正常工作下的流量信息;The determining module 2 is configured to determine whether the network traffic statistics information of the physical node acquired by the acquiring module 1 meets the threshold information, and the threshold information is the traffic information under normal working of the physical node;
确定模块3,设置为若判断模块2判断为不满足,根据每个虚拟机预设的流量占用信息,分别确定每个虚拟机的网络流量占用信息与流量占用信息的差值;以及The determining module 3 is configured to determine, according to the traffic occupation information preset by each virtual machine, the difference between the network traffic occupation information and the traffic occupation information of each virtual machine, respectively, if the determining module 2 determines that the content is not satisfied;
迁移模块4,设置为根据确定模块3确定的差值,对虚拟机执行迁移。The migration module 4 is configured to perform migration on the virtual machine according to the difference determined by the determination module 3.
可选地,网络流量统计信息包括物理节点的当前实际物理网卡带宽占用量,确定模块3还设置为根据物理节点的最大实际物理网卡带宽占用量,确定物理节点正常工作下的物理网卡带宽占用量;判断模块2是设置为判断物理节点的当前实际物理网卡带宽占用量是否满足物理节点正常工作下的物理网卡带宽占用量。Optionally, the network traffic statistics information includes the current actual physical network card bandwidth usage of the physical node, and the determining module 3 is further configured to determine the physical network card bandwidth usage of the physical node under normal working according to the maximum actual physical network card bandwidth usage of the physical node. The judging module 2 is configured to determine whether the current actual physical network card bandwidth occupancy of the physical node satisfies the physical network card bandwidth occupancy of the physical node under normal working.
可选地,判断模块2还设置为判断确定模块3确定的差值是否为正,迁移模块4是设置为对差值为正的至少一个虚拟机执行迁移。Optionally, the determining module 2 is further configured to determine whether the difference determined by the determining module 3 is positive, and the migration module 4 is configured to perform migration on the at least one virtual machine whose positive value is positive.
可选地,网络流量占用信息包括虚拟机的当前实际出口带宽占用量;确定模块3包括设置子模块31和计算子模块32,其中:设置子模块31设置为分别预设每个虚拟机的最大出口带宽量;计算子模块32设置为根据每个虚拟机预设的最大出口带宽量,分别计算每个虚拟机的当前实际出口带宽占用量 与最大出口带宽量的差值。Optionally, the network traffic occupation information includes a current actual outlet bandwidth usage of the virtual machine; the determining module 3 includes a setting submodule 31 and a computing submodule 32, wherein: the setting submodule 31 is configured to preset the maximum of each virtual machine respectively. The amount of the egress bandwidth; the calculation sub-module 32 is configured to calculate the current actual egress bandwidth usage of each virtual machine according to the maximum amount of the egress bandwidth preset by each virtual machine. The difference from the maximum amount of exit bandwidth.
可选地,所述装置还包括添加模块5,添加模块5设置为将差值为正的虚拟机添加至待迁移虚拟机列表;确定模块3还设置为在待迁移虚拟机列表中,确定至少一个需要迁移的虚拟机,使迁移至少一个需要迁移的虚拟机后的物理节点的网络流量统计信息满足阈值信息;迁移模块4是设置为将至少一个需要迁移的虚拟机作为待迁移虚拟机,对待迁移虚拟机执行迁移。Optionally, the device further includes an adding module 5, the adding module 5 is configured to add a virtual machine with a positive difference to the virtual machine to be migrated; the determining module 3 is further configured to determine at least the virtual machine to be migrated, determining at least A virtual machine that needs to be migrated, so that the network traffic statistics of the physical node after the migration of at least one virtual machine that needs to be migrated meets the threshold information; the migration module 4 is configured to treat at least one virtual machine to be migrated as the virtual machine to be migrated, and treat Migrate the virtual machine to perform the migration.
可选地,判断模块2还设置为在待迁移虚拟机列表中,判断迁移差值最大的虚拟机后的物理节点的网络流量统计信息是否满足阈值信息,确定模块3是设置为若判断模块2判断为满足,则将差值最大的虚拟机确定为需要迁移的虚拟机,若不满足,则判断模块2还设置为判断迁移差值最大的虚拟机以及差值次大的虚拟机后的物理节点的网络流量统计信息是否满足阈值信息;确定模块3是设置为若判断模块2判断为满足,则将差值最大的虚拟机以及差值次大的虚拟机确定为需要迁移的虚拟机,若不满足,则判断模块2还设置为继续判断,直到判断出迁移虚拟机后的物理节点的网络流量统计信息满足阈值信息;所述确定模块3是设置为将所述判断模块2判断出的满足条件的虚拟机确定为需要迁移的虚拟机,所述条件为迁移虚拟机后的物理节点的网络流量统计信息满足所述阈值信息。Optionally, the determining module 2 is further configured to determine, in the list of the virtual machines to be migrated, whether the network traffic statistics information of the physical node after the virtual machine having the largest difference value meets the threshold information, and the determining module 3 is configured to determine the module 2 If the determination is satisfied, the virtual machine with the largest difference is determined as the virtual machine that needs to be migrated. If not, the determination module 2 is further configured to determine the virtual machine with the largest migration difference and the physical entity after the virtual machine with the largest difference. Whether the network traffic statistics of the node meets the threshold information; the determining module 3 is configured to determine, if the determining module 2 is satisfied, the virtual machine with the largest difference and the virtual machine with the largest difference as the virtual machine to be migrated, if If the method is not satisfied, the determining module 2 is further configured to continue to determine until the network traffic statistics of the physical node after the virtual machine is migrated meets the threshold information; the determining module 3 is configured to satisfy the determining by the determining module 2 The virtual machine of the condition is determined to be a virtual machine that needs to be migrated, and the condition is that the network traffic statistics of the physical node after the virtual machine is migrated are satisfied. Said threshold information.
可选地,所述装置还包括排列模块6、限制模块7和取消模块8,其中:排列模块6设置为将待迁移虚拟机列表中的所有虚拟机按照所述差值由大到小依次排列;限制模块7设置为对排列模块6依次排列后的前至少一个虚拟机的流量进行限制;取消模块8设置为若前至少一个虚拟机中包括未迁移的虚拟机,则取消对未迁移虚拟机的流量限制。Optionally, the device further includes an arranging module 6, a limiting module 7, and a canceling module 8, wherein: the arranging module 6 is configured to arrange all the virtual machines in the virtual machine list to be migrated according to the difference from large to small. The restriction module 7 is configured to limit the traffic of the at least one virtual machine after the arrangement module 6 is sequentially arranged; the cancellation module 8 is configured to cancel the unmigrated virtual machine if at least one of the virtual machines includes the unmigrated virtual machine Traffic restrictions.
可选地,获取模块1还设置为获取其他物理节点的网络流量统计信息,所述装置还包括选择模块9,选择模块9设置为根据获取模块1获取的其他物理节点的网络流量统计信息,选择满足虚拟机迁入条件的物理节点,使迁入虚拟机后的物理节点的网络流量统计信息仍满足阈值信息,将选择的物理节点作为至少一个虚拟机执行迁移的目标物理节点。Optionally, the obtaining module 1 is further configured to obtain network traffic statistics information of other physical nodes, and the device further includes a selecting module 9 configured to select, according to network traffic statistics information of other physical nodes acquired by the acquiring module 1, The physical node that meets the conditions for the virtual machine to move in, the network traffic statistics of the physical node that is moved into the virtual machine still meets the threshold information, and the selected physical node is the target physical node that performs the migration as at least one virtual machine.
实施例三: Embodiment 3:
图3为本发明实施例三提供的虚拟机迁移决策装置的结构示意图,如图3所示,本实施例包含一个OpenStack控制节点A和多个OpenStack计算节点Bn。在该OpenStack控制器节点A中,增加了网络监测模块11、插件模块12,修改了网络组件模块13和迁移模块15,在该OpenStack的计算节点中,增加了数据代理模块22,修改了组件模块23。FIG. 3 is a schematic structural diagram of a virtual machine migration decision apparatus according to Embodiment 3 of the present invention. As shown in FIG. 3, this embodiment includes an OpenStack control node A and multiple OpenStack compute nodes Bn. In the OpenStack controller node A, the network monitoring module 11 and the plug-in module 12 are added, and the network component module 13 and the migration module 15 are modified. In the computing node of the OpenStack, the data proxy module 22 is added, and the component module is modified. twenty three.
S1:OpenStack计算节点B1上运行的交换机模块需要配置sflow(一种网络监测技术)统计收集功能,并将统计数据上报给OpenStack计算节点内部驻留的代理模块22,每个OpenStack计算节点的代理模块通过sflow对数据进行统计处理;S1: The switch module running on the OpenStack compute node B1 needs to configure the sflow (a network monitoring technology) statistics collection function, and report the statistics to the proxy module 22 resident in the OpenStack compute node, and the proxy module of each OpenStack compute node. Statistical processing of data through sflow;
S2:代理模块22通过消息队列的方式上传处理后的统计数据到OpenStack控制节点A中的网络组件模块13;S2: the proxy module 22 uploads the processed statistical data to the network component module 13 in the OpenStack control node A by means of a message queue;
S3:网络组件插件模块13负责将统计数据存储在数据库模块14中;S3: the network component plug-in module 13 is responsible for storing the statistical data in the database module 14;
S4:网络监测模块11实时监测每个OpenStack计算节点上的网络利用情况,通过插件模块12提供的rest API接口获取统计数据;S4: The network monitoring module 11 monitors network utilization on each OpenStack computing node in real time, and obtains statistical data through the rest API interface provided by the plug-in module 12;
S5:插件模块12从数据库模块14中获取统计数据;S5: the plug-in module 12 obtains statistical data from the database module 14;
S6:网络监测模块11根据插件模块12返回的统计数据进行判断是否存在待迁移虚拟机,请参阅实施例一提供的虚拟机迁移决策方法;若存在,网络监测模块可通过rest API调用的方式,调用网络组件模块13对虚拟机的流量进行限流;S6: The network monitoring module 11 determines whether there is a virtual machine to be migrated according to the statistics returned by the plug-in module 12. For example, the virtual machine migration decision method provided in the first embodiment; if present, the network monitoring module can invoke the rest API. Calling the network component module 13 to limit the flow of the virtual machine;
S7:网络组件模块13通过消息通信的方式控制OpenStack计算节点上的组件模块23实现对虚拟机的限流;S7: The network component module 13 controls the component module 23 on the OpenStack computing node to implement the current limiting of the virtual machine by means of message communication;
S8:组件模块23通过调用交换机模块21的命令通过Openflow meter的方式实现限流;S8: the component module 23 implements current limiting by means of an Openflow meter by invoking a command of the switch module 21;
S9:网络监测模块11调用迁移模块15;S9: the network monitoring module 11 calls the migration module 15;
S10:迁移模块15在选择目标迁移主机时需要根据当前的网络状况做出选择,需要通过rest API的方式调用网络组件模块13获取网络数据;S10: The migration module 15 needs to make a selection according to the current network condition when selecting the target migration host, and needs to invoke the network component module 13 to acquire network data by means of the rest API;
S11:网络组件模块13从数据库模块14获取统计数据;S11: The network component module 13 acquires statistical data from the database module 14;
S12:当迁移结束后,网络监测模块11需要调用网络组件模块13取消未 被迁移的虚拟机的流量限制;S12: After the migration ends, the network monitoring module 11 needs to call the network component module 13 to cancel the The traffic limit of the migrated virtual machine;
S13:网络组件模块13通过消息通信的方式控制OpenStack计算节点上的组件模块23实现取消虚拟机的限流;S13: The network component module 13 controls the component module 23 on the OpenStack computing node by means of message communication to implement the current limiting of the virtual machine.
S14:组件模块23通过调用交换机模块21的命令通过Openflow meter的方式实现取消限流。S14: The component module 23 cancels the current limiting by means of the command of the switch module 21 through the Openflow meter.
本实施例中的实现方法,有如下优点:The implementation method in this embodiment has the following advantages:
(1)引入网络数据统计,通过在OpenStack计算节点增加数据代理和在OpenStack控制节点增加网络数据插件,实现网络统计信息的收集和存储;(1) Introducing network data statistics, by adding data agents in OpenStack compute nodes and adding network data plug-ins on OpenStack control nodes to achieve collection and storage of network statistics;
(2)通过在OpenStack控制节点增加网络监测模块,实现对虚拟机网络利用情况的分析来决策是否触发虚拟机的动态迁移,以及对虚拟机网络流量的动态管控;(2) By adding a network monitoring module to the OpenStack control node to analyze the utilization of the virtual machine network to decide whether to trigger the dynamic migration of the virtual machine and the dynamic control of the virtual machine network traffic;
(3)通过修改OpenStack主机选择算法,增加对网络的过滤条件,优化了虚拟机迁移时的主机选择算法。(3) By modifying the OpenStack host selection algorithm, increasing the filtering conditions on the network, and optimizing the host selection algorithm during virtual machine migration.
本领域普通技术人员可以理解上述方法中的全部或部分步骤可通过程序来指令相关硬件完成,上述程序可以存储于计算机可读存储介质中,如只读存储器、磁盘或光盘等。可选地,上述实施例的全部或部分步骤也可以使用一个或多个集成电路来实现。相应地,上述实施例中的各模块/单元可以采用硬件的形式实现,也可以采用软件功能模块的形式实现。本发明实施例不限制于任何特定形式的硬件和软件的结合。One of ordinary skill in the art will appreciate that all or a portion of the above steps may be accomplished by a program that instructs the associated hardware, such as a read-only memory, a magnetic disk, or an optical disk. Alternatively, all or part of the steps of the above embodiments may also be implemented using one or more integrated circuits. Correspondingly, each module/unit in the foregoing embodiment may be implemented in the form of hardware or in the form of a software function module. Embodiments of the invention are not limited to any specific form of combination of hardware and software.
工业实用性Industrial applicability
本发明实施例根据物理节点的网络流量统计信息,以及物理节点内每个虚拟机的网络流量占用信息,判断是否需要对虚拟机进行迁移,能动态监测虚拟机的网络流量占用情况,动态决策虚拟机的迁移,减少系统维护的人工参与度。通过获取并存储网络流量统计信息,实现对虚拟机网络流量占用情况的分析来决策是否触发虚拟机的动态迁移,以及对虚拟机网络流量的动态管控。 In the embodiment of the present invention, according to the network traffic statistics information of the physical node and the network traffic occupation information of each virtual machine in the physical node, it is determined whether the virtual machine needs to be migrated, and the network traffic occupancy of the virtual machine can be dynamically monitored, and the dynamic decision virtual Machine migration reduces manual participation in system maintenance. By obtaining and storing network traffic statistics, the network traffic usage of the virtual machine is analyzed to determine whether to trigger dynamic migration of the virtual machine and dynamic control of the virtual machine network traffic.

Claims (17)

  1. 一种虚拟机迁移决策方法,包括:A virtual machine migration decision method includes:
    获取物理节点的网络流量统计信息,以及所述物理节点内每个虚拟机的网络流量占用信息;Obtaining network traffic statistics information of the physical node, and network traffic occupation information of each virtual machine in the physical node;
    判断所述物理节点的网络流量统计信息是否满足阈值信息,所述阈值信息为所述物理节点正常工作下的流量信息;Determining whether the network traffic statistics information of the physical node meets threshold information, where the threshold information is traffic information under normal working of the physical node;
    若不满足,根据所述每个虚拟机预设的流量占用信息,分别确定所述每个虚拟机的网络流量占用信息与流量占用信息的差值;If not, the difference between the network traffic occupation information and the traffic occupation information of each virtual machine is determined according to the preset traffic occupation information of each virtual machine;
    根据所述差值,对虚拟机执行迁移。Perform migration on the virtual machine based on the difference.
  2. 根据权利要求1所述的虚拟机迁移决策方法,其中,所述网络流量统计信息包括所述物理节点的当前实际物理网卡带宽占用量;The virtual machine migration decision method according to claim 1, wherein the network traffic statistics information includes a current actual physical network card bandwidth occupancy of the physical node;
    所述判断所述物理节点的网络流量统计信息是否满足阈值信息包括:The determining whether the network traffic statistics information of the physical node meets the threshold information includes:
    根据所述物理节点的最大实际物理网卡带宽占用量,确定所述物理节点正常工作下的物理网卡带宽占用量;Determining, according to the maximum actual physical NIC bandwidth usage of the physical node, the bandwidth occupied by the physical NIC under normal working of the physical node;
    判断所述物理节点的当前实际物理网卡带宽占用量是否满足所述物理节点正常工作下的物理网卡带宽占用量。Determining whether the current actual physical NIC bandwidth usage of the physical node meets the physical NIC bandwidth usage of the physical node under normal working.
  3. 根据权利要求1所述的虚拟机迁移决策方法,其中,所述根据所述差值,对虚拟机执行迁移包括:The virtual machine migration decision method according to claim 1, wherein the performing migration to the virtual machine according to the difference includes:
    判断所述差值是否为正;Determining whether the difference is positive;
    对所述差值为正的至少一个虚拟机执行迁移。Performing a migration on at least one virtual machine whose difference is positive.
  4. 根据权利要求3所述的虚拟机迁移决策方法,其中,所述网络流量占用信息包括所述虚拟机的当前实际出口带宽占用量;The virtual machine migration decision method according to claim 3, wherein the network traffic occupation information comprises a current actual outlet bandwidth occupancy of the virtual machine;
    所述根据所述每个虚拟机预设的流量占用信息,分别确定所述每个虚拟机的网络流量占用信息与流量占用信息的差值包括:Determining, according to the traffic occupation information preset by each virtual machine, the difference between the network traffic occupation information and the traffic occupation information of each virtual machine includes:
    分别预设所述每个虚拟机的最大出口带宽量;Presetting the maximum amount of outlet bandwidth of each virtual machine respectively;
    根据所述每个虚拟机预设的最大出口带宽量,分别计算所述每个虚拟机的当前实际出口带宽占用量与最大出口带宽量的差值。 And calculating, according to the maximum amount of the egress bandwidth preset by each virtual machine, a difference between a current actual egress bandwidth occupancy of each virtual machine and a maximum egress bandwidth amount.
  5. 根据权利要求3或4所述的虚拟机迁移决策方法,其中,所述对所述差值为正的至少一个虚拟机执行迁移包括:The virtual machine migration decision method according to claim 3 or 4, wherein the performing migration to the at least one virtual machine whose positive value is positive comprises:
    将所述差值为正的虚拟机添加至待迁移虚拟机列表;Add the virtual machine whose positive value is positive to the virtual machine list to be migrated;
    在所述待迁移虚拟机列表中,确定至少一个需要迁移的虚拟机,使迁移所述至少一个需要迁移的虚拟机后的物理节点的网络流量统计信息满足所述阈值信息;Determining at least one virtual machine to be migrated in the virtual machine list to be migrated, so that network traffic statistics of the physical node after the migration of the at least one virtual machine to be migrated meets the threshold information;
    将所述至少一个需要迁移的虚拟机作为待迁移虚拟机,对所述待迁移虚拟机执行迁移。The at least one virtual machine to be migrated is used as a virtual machine to be migrated, and the migration is performed on the virtual machine to be migrated.
  6. 根据权利要求5所述的虚拟机迁移决策方法,其中,所述在所述待迁移虚拟机列表中,确定至少一个需要迁移的虚拟机,使迁移所述至少一个需要迁移的虚拟机后的物理节点的网络流量统计信息满足所述阈值信息包括:The virtual machine migration decision method according to claim 5, wherein in the list of virtual machines to be migrated, at least one virtual machine that needs to be migrated is determined, so that the physical after the migration of the at least one virtual machine to be migrated is performed. The network traffic statistics of the node satisfy the threshold information, including:
    在所述待迁移虚拟机列表中,判断迁移所述差值最大的虚拟机后的物理节点的网络流量统计信息是否满足所述阈值信息;Determining, in the list of the virtual machines to be migrated, whether the network traffic statistics of the physical node after the virtual machine having the largest difference is migrated meets the threshold information;
    若满足,则将所述差值最大的虚拟机确定为需要迁移的虚拟机,若不满足,则判断迁移所述差值最大的虚拟机以及差值次大的虚拟机后的物理节点的网络流量统计信息是否满足所述阈值信息,若满足,则将所述差值最大的虚拟机以及差值次大的虚拟机确定为需要迁移的虚拟机,若不满足,则直到判断出迁移虚拟机后的物理节点的网络流量统计信息满足所述阈值信息,将所述虚拟机确定为需要迁移的虚拟机。If yes, determine the virtual machine with the largest difference as the virtual machine to be migrated. If not, determine the network of the physical node after the virtual machine with the largest difference and the virtual machine with the largest difference. Whether the traffic statistics information meets the threshold information, and if yes, the virtual machine with the largest difference and the virtual machine with the largest difference are determined as virtual machines to be migrated, and if not, the virtual machine is determined to be migrated. The network traffic statistics of the subsequent physical node satisfy the threshold information, and the virtual machine is determined to be a virtual machine that needs to be migrated.
  7. 根据权利要求5所述的虚拟机迁移决策方法,The virtual machine migration decision method according to claim 5,
    在所述将所述至少一个需要迁移的虚拟机作为待迁移虚拟机,对所述待迁移虚拟机执行迁移之前,所述方法还包括:Before the performing, by using the virtual machine that needs to be migrated as the virtual machine to be migrated, performing the migration on the virtual machine to be migrated, the method further includes:
    将所述待迁移虚拟机列表中的所有虚拟机按照所述差值由大到小依次排列;All the virtual machines in the virtual machine list to be migrated are arranged in descending order according to the difference;
    对所述依次排列后的前至少一个虚拟机的流量进行限制;Limiting the traffic of the at least one virtual machine in the prior arrangement;
    在所述将所述至少一个需要迁移的虚拟机作为待迁移虚拟机,对所述待迁移虚拟机执行迁移之后,所述方法还包括:After the performing, by using the virtual machine to be migrated as the virtual machine to be migrated, performing the migration on the virtual machine to be migrated, the method further includes:
    若所述前至少一个虚拟机中包括未迁移的虚拟机,则取消对所述未迁移 虚拟机的流量限制。If the at least one virtual machine includes an unmigrated virtual machine, cancel the unmigrated The flow limit of the virtual machine.
  8. 根据权利要求1-4任一项所述的虚拟机迁移决策方法,在所述根据所述差值,对虚拟机执行迁移之前,所述方法还包括:The virtual machine migration decision method according to any one of claims 1 to 4, wherein before the performing the migration on the virtual machine according to the difference, the method further includes:
    获取其他物理节点的网络流量统计信息;Obtain network traffic statistics of other physical nodes;
    根据所述其他物理节点的网络流量统计信息,选择满足虚拟机迁入条件的物理节点,使迁入虚拟机后的物理节点的网络流量统计信息仍满足所述阈值信息,将所述选择的物理节点作为所述至少一个虚拟机执行迁移的目标物理节点。And selecting, according to the network traffic statistics information of the other physical nodes, the physical node that meets the virtual machine migration condition, so that the network traffic statistics information of the physical node that is moved into the virtual machine still meets the threshold information, and the selected physical The node performs the migration as the target physical node of the at least one virtual machine.
  9. 一种虚拟机迁移决策装置,包括:A virtual machine migration decision device includes:
    获取模块,设置为获取物理节点的网络流量统计信息,以及所述物理节点内每个虚拟机的网络流量占用信息;Obtaining a module, configured to obtain network traffic statistics information of the physical node, and network traffic occupation information of each virtual machine in the physical node;
    判断模块,设置为判断所述获取模块获取的物理节点的网络流量统计信息是否满足阈值信息,所述阈值信息为所述物理节点正常工作下的流量信息;The determining module is configured to determine whether the network traffic statistics information of the physical node acquired by the acquiring module meets threshold information, where the threshold information is traffic information under normal working of the physical node;
    确定模块,设置为若所述判断模块判断为不满足,根据所述每个虚拟机预设的流量占用信息,分别确定所述每个虚拟机的网络流量占用信息与流量占用信息的差值;以及a determining module, configured to determine, according to the traffic occupation information preset by each virtual machine, a difference between the network traffic occupation information and the traffic occupation information of each virtual machine, respectively, if the determining module determines that the content is not satisfied; as well as
    迁移模块,设置为根据所述确定模块确定的差值,对虚拟机执行迁移。The migration module is configured to perform migration on the virtual machine according to the difference determined by the determining module.
  10. 根据权利要求9所述的虚拟机迁移决策装置,其中,所述网络流量统计信息包括所述物理节点的当前实际物理网卡带宽占用量;The virtual machine migration decision device according to claim 9, wherein the network traffic statistics information includes a current actual physical network card bandwidth occupancy of the physical node;
    所述确定模块还设置为根据所述物理节点的最大实际物理网卡带宽占用量,确定所述物理节点正常工作下的物理网卡带宽占用量;The determining module is further configured to determine, according to a maximum actual physical NIC bandwidth occupancy of the physical node, a bandwidth occupied by the physical NIC under normal working of the physical node;
    所述判断模块是设置为判断所述物理节点的当前实际物理网卡带宽占用量是否满足所述物理节点正常工作下的物理网卡带宽占用量。The determining module is configured to determine whether the current actual physical network card bandwidth occupancy of the physical node meets the physical network card bandwidth occupancy of the physical node under normal working.
  11. 根据权利要求9所述的虚拟机迁移决策装置,The virtual machine migration decision device according to claim 9,
    所述判断模块还设置为判断所述确定模块确定的差值是否为正;The determining module is further configured to determine whether the difference determined by the determining module is positive;
    所述迁移模块是设置为对所述差值为正的至少一个虚拟机执行迁移。The migration module is configured to perform migration on the at least one virtual machine whose difference is positive.
  12. 根据权利要求11所述的虚拟机迁移决策装置,其中,所述网络流量 占用信息包括所述虚拟机的当前实际出口带宽占用量;The virtual machine migration decision device according to claim 11, wherein said network traffic The occupancy information includes the current actual egress bandwidth usage of the virtual machine;
    所述确定模块包括:The determining module includes:
    设置子模块,设置为分别预设所述每个虚拟机的最大出口带宽量;以及Setting a sub-module, which is set to respectively preset a maximum amount of exit bandwidth of each of the virtual machines;
    计算子模块,设置为根据所述每个虚拟机预设的最大出口带宽量,分别计算所述每个虚拟机的当前实际出口带宽占用量与最大出口带宽量的差值。The calculating sub-module is configured to calculate, according to the maximum amount of the egress bandwidth preset by each virtual machine, a difference between the current actual egress bandwidth occupancy of each virtual machine and the maximum egress bandwidth.
  13. 根据权利要求11或12所述的虚拟机迁移决策装置,所述装置还包括:The virtual machine migration decision device according to claim 11 or 12, further comprising:
    添加模块,设置为将所述差值为正的虚拟机添加至待迁移虚拟机列表;Add a module, and set to add the virtual machine whose positive value is positive to the virtual machine list to be migrated;
    所述确定模块还设置为在所述待迁移虚拟机列表中,确定至少一个需要迁移的虚拟机,使迁移所述至少一个需要迁移的虚拟机后的物理节点的网络流量统计信息满足所述阈值信息;The determining module is further configured to: determine, in the list of virtual machines to be migrated, at least one virtual machine that needs to be migrated, so that network traffic statistics of the physical node after the migration of the at least one virtual machine that needs to be migrated meets the threshold information;
    所述迁移模块是设置为将所述至少一个需要迁移的虚拟机作为待迁移虚拟机,对所述待迁移虚拟机执行迁移。The migration module is configured to perform the migration on the virtual machine to be migrated by using the at least one virtual machine to be migrated as a virtual machine to be migrated.
  14. 根据权利要求13所述的虚拟机迁移决策装置,The virtual machine migration decision device according to claim 13,
    所述判断模块还设置为在所述待迁移虚拟机列表中,判断迁移所述差值最大的虚拟机后的物理节点的网络流量统计信息是否满足所述阈值信息;The determining module is further configured to determine, in the list of the virtual machines to be migrated, whether the network traffic statistics information of the physical node after the virtual machine having the largest difference is migrated meets the threshold information;
    所述确定模块是设置为若所述判断模块判断为满足,则将所述差值最大的虚拟机确定为需要迁移的虚拟机,若不满足,则所述判断模块还设置为判断迁移所述差值最大的虚拟机以及差值次大的虚拟机后的物理节点的网络流量统计信息是否满足所述阈值信息,所述确定模块是设置为若所述判断模块判断为满足,则将所述差值最大的虚拟机以及差值次大的虚拟机确定为需要迁移的虚拟机,若不满足,则所述判断模块还设置为继续判断,直到判断出迁移虚拟机后的物理节点的网络流量统计信息满足所述阈值信息,所述确定模块是设置为将所述判断模块判断出的满足条件的虚拟机确定为需要迁移的虚拟机,所述条件为迁移虚拟机后的物理节点的网络流量统计信息满足所述阈值信息。The determining module is configured to determine, if the determining module is satisfied, the virtual machine with the largest difference as the virtual machine that needs to be migrated, and if not, the determining module is further configured to determine that the migration is performed. Whether the network traffic statistics of the physical node having the largest difference and the virtual node having the second largest difference satisfy the threshold information, and the determining module is configured to: if the determining module determines that the content is satisfied, the The virtual machine with the largest difference and the virtual machine with the largest difference are determined to be virtual machines that need to be migrated. If not, the judgment module is further configured to continue to judge until the network traffic of the physical node after the virtual machine is migrated is determined. The statistical information meets the threshold information, and the determining module is configured to determine a virtual machine that meets the condition determined by the determining module as a virtual machine that needs to be migrated, where the condition is a network traffic of a physical node after the virtual machine is migrated. The statistical information satisfies the threshold information.
  15. 根据权利要求13所述的虚拟机迁移决策装置,所述装置还包括:The virtual machine migration decision device of claim 13, the device further comprising:
    排列模块,设置为将所述待迁移虚拟机列表中的所有虚拟机按照所述差 值由大到小依次排列;Arranging a module, configured to follow all the virtual machines in the list of virtual machines to be migrated according to the difference The values are arranged in order from large to small;
    限制模块,设置为对所述排列模块依次排列后的前至少一个虚拟机的流量进行限制;以及a limiting module, configured to limit traffic of the at least one virtual machine that is sequentially arranged by the arranging module; and
    取消模块,设置为若所述前至少一个虚拟机中包括未迁移的虚拟机,则取消对所述未迁移虚拟机的流量限制。The module is canceled, and if the at least one virtual machine includes the unmigrated virtual machine, the traffic restriction on the unmigrated virtual machine is cancelled.
  16. 根据权利要求9-12任一项所述的虚拟机迁移决策装置,A virtual machine migration decision device according to any one of claims 9-12,
    所述获取模块还设置为获取其他物理节点的网络流量统计信息;The obtaining module is further configured to obtain network traffic statistics information of other physical nodes;
    所述装置还包括:选择模块,其设置为根据所述获取模块获取的其他物理节点的网络流量统计信息,选择满足虚拟机迁入条件的物理节点,使迁入虚拟机后的物理节点的网络流量统计信息仍满足所述阈值信息,将所述选择的物理节点作为所述至少一个虚拟机执行迁移的目标物理节点。The device further includes: a selecting module, configured to select, according to network traffic statistics information of other physical nodes acquired by the acquiring module, a physical node that satisfies a virtual machine moving in condition, and a network of the physical node that moves into the virtual machine The traffic statistics information still satisfies the threshold information, and the selected physical node is used as the target physical node of the at least one virtual machine to perform migration.
  17. 一种计算机可读存储介质,存储有程序指令,当该程序指令被执行时可实现权利要求1-8任一项所述的方法。 A computer readable storage medium storing program instructions that, when executed, implement the method of any of claims 1-8.
PCT/CN2015/083192 2014-11-24 2015-07-02 Decision-making method and apparatus for virtual machine migration WO2016082546A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201410680289.6A CN105656969A (en) 2014-11-24 2014-11-24 Virtual machine migration decision method and device
CN201410680289.6 2014-11-24

Publications (1)

Publication Number Publication Date
WO2016082546A1 true WO2016082546A1 (en) 2016-06-02

Family

ID=56073526

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2015/083192 WO2016082546A1 (en) 2014-11-24 2015-07-02 Decision-making method and apparatus for virtual machine migration

Country Status (2)

Country Link
CN (1) CN105656969A (en)
WO (1) WO2016082546A1 (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111143324A (en) * 2019-12-20 2020-05-12 浪潮软件股份有限公司 Kudu database data balancing system based on size and implementation method
CN111478850A (en) * 2020-02-29 2020-07-31 新华三大数据技术有限公司 Gateway adjusting method and device

Families Citing this family (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107562511A (en) * 2016-06-30 2018-01-09 中兴通讯股份有限公司 Moving method, physical server, virtual server and the system of physical to virtual
CN106534318B (en) * 2016-11-15 2019-10-29 浙江大学 A kind of OpenStack cloud platform resource dynamic scheduling system and method based on flow compatibility
CN108255575A (en) * 2017-12-06 2018-07-06 顺丰科技有限公司 A kind of virtual machine migration method and device
CN108614726B (en) * 2018-03-22 2022-06-10 华为云计算技术有限公司 Virtual machine creation method and device
CN113259473B (en) * 2021-06-08 2021-11-05 广东睿江云计算股份有限公司 Self-adaptive cloud data migration method
CN114296868B (en) * 2021-12-17 2022-10-04 中电信数智科技有限公司 Virtual machine automatic migration decision method based on user experience in multi-cloud environment

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20120324073A1 (en) * 2011-06-17 2012-12-20 International Business Machines Corporation Virtual machine load balancing
CN102929715A (en) * 2012-10-31 2013-02-13 曙光云计算技术有限公司 Method and system for scheduling network resources based on virtual machine migration
CN103559084A (en) * 2013-10-17 2014-02-05 电子科技大学 Virtual machine migration method of energy-saving data center
CN103729252A (en) * 2013-12-20 2014-04-16 杭州华为数字技术有限公司 Virtual machine scheduling method and scheduling monitor
CN104008018A (en) * 2014-05-22 2014-08-27 南京邮电大学 Virtual machine online migration method in cloud computing environment

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20120324073A1 (en) * 2011-06-17 2012-12-20 International Business Machines Corporation Virtual machine load balancing
CN102929715A (en) * 2012-10-31 2013-02-13 曙光云计算技术有限公司 Method and system for scheduling network resources based on virtual machine migration
CN103559084A (en) * 2013-10-17 2014-02-05 电子科技大学 Virtual machine migration method of energy-saving data center
CN103729252A (en) * 2013-12-20 2014-04-16 杭州华为数字技术有限公司 Virtual machine scheduling method and scheduling monitor
CN104008018A (en) * 2014-05-22 2014-08-27 南京邮电大学 Virtual machine online migration method in cloud computing environment

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
MAYANK MISHRA ET AL.: "Dynamic Resource Management Using Virtual Machine Migrations", IEEE COMMUNICATIONS MAGAZINE, vol. 50, no. 9, 30 September 2012 (2012-09-30), pages 35 *

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111143324A (en) * 2019-12-20 2020-05-12 浪潮软件股份有限公司 Kudu database data balancing system based on size and implementation method
CN111143324B (en) * 2019-12-20 2023-05-02 浪潮软件股份有限公司 Kudu database data equalization system based on size and implementation method
CN111478850A (en) * 2020-02-29 2020-07-31 新华三大数据技术有限公司 Gateway adjusting method and device

Also Published As

Publication number Publication date
CN105656969A (en) 2016-06-08

Similar Documents

Publication Publication Date Title
WO2016082546A1 (en) Decision-making method and apparatus for virtual machine migration
EP3284215B1 (en) System and method for sla violation monitoring via multi-level thresholds
WO2020149786A1 (en) Dynamic deployment of network applications having performance and reliability guarantees in large computing networks
US10153979B2 (en) Prioritization of network traffic in a distributed processing system
KR20130115553A (en) Two-level resource management method and appratus for dynamic resource management
CN107786351A (en) Service bandwidth self-adapting regulation method, system and SDN controllers
CN110296501B (en) Centralized management and control system and method for air conditioning system
CN108476175B (en) Transfer SDN traffic engineering method and system using dual variables
KR20170120335A (en) Apparatus and method for managing computing resources in network function virtualization system
JP2017037492A (en) Distributed processing program, distributed processing method and distributed processor
CN104199738B (en) A kind of more data processing equipment collaboration working methods and system
CN106059940A (en) Flow control method and device
CN104641606A (en) Method for network resource equalization processing, and virtual network management entity
CN109739634A (en) A kind of atomic task execution method and device
CN105207856A (en) Load balancing system and method based on SDN virtual switch
EP2940600A1 (en) Data scanning method and device
CN103338240A (en) Cloud server automatic monitoring system and method used for monitoring automatic drifting
CN107995199A (en) The port speed constraint method and device of the network equipment
JP2016054451A (en) Virtual network allocating method and device
TWI644540B (en) Flow meter flexible cutting system for virtual network in multi-tenant software-defined network
JP2019159646A (en) Vm performance assurance system and vm performance assurance method
CN106371912B (en) A kind of resource regulating method and device of streaming computing system
WO2016082365A1 (en) Traffic adjusting method and device
CN114565201A (en) Management system, method, device and storage medium of data center
JP5758358B2 (en) Virtual machine system and high-speed live migration method thereof

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 15863534

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 15863534

Country of ref document: EP

Kind code of ref document: A1