CN107544839A - Virtual machine (vm) migration system, method and device - Google Patents
Virtual machine (vm) migration system, method and device Download PDFInfo
- Publication number
- CN107544839A CN107544839A CN201610481831.4A CN201610481831A CN107544839A CN 107544839 A CN107544839 A CN 107544839A CN 201610481831 A CN201610481831 A CN 201610481831A CN 107544839 A CN107544839 A CN 107544839A
- Authority
- CN
- China
- Prior art keywords
- calculate node
- node
- virtual machine
- calculation
- fault
- Prior art date
- Legal status (The legal status 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 status listed.)
- Granted
Links
Landscapes
- Debugging And Monitoring (AREA)
Abstract
The embodiment of the present application provides a kind of virtual machine (vm) migration system, method and device, a kind of virtual machine (vm) migration system that the embodiment of the present application provides, reports the monitoring data of its own respectively by each calculate node；Monitoring data of the collector cluster according to each calculate node, fault detect is carried out to each calculate node respectively, the calculation of fault node to break down is determined, calculation of fault node is reported into cloud controller；Cloud controller determines purpose calculate node, and the virtual machine in calculation of fault node is migrated to the purpose calculate node respectively.So that the virtual machine of calculation of fault node normal operation in purpose calculate node, so that the key business of enterprise and crucial application continue to run with.
Description
Technical field
The invention relates to cloud platform technical field, more particularly relates to virtual machine (vm) migration system, method and device.
Background technology
With the continuous development of cloud computing technology, the complexity of cloud platform in itself is progressively being aggravated, and cloud platform includes cloud control
Device, cluster controller, calculate node controller and calculate node processed.Cloud controller is used to manage cluster information；Clustered control
Device is used for network resource administration information, calculate node information, Virtual Cluster information；Calculate node provide possess hard disk, internal memory,
The physical server of the physical resources such as CPU, calculate node can include one or more virtual machines；Calculate node controller is used for
Manage the virtual machine in calculate node.
With the continuous development of cloud computing technology, the key business of enterprise and crucial application are progressively migrated to cloud platform and fallen into a trap
The virtual machine of operator node.When calculate node breaks down, virtual machine can not be run, and the key business and key for causing enterprise should
With can not run.
The content of the invention
In view of this, the invention provides a kind of virtual machine (vm) migration system, method and device, with overcome in the prior art when
When calculate node breaks down, virtual machine can not be run, the problem of causing the key business of enterprise and crucial application not to run.
To achieve the above object, the present invention provides following technical scheme：
A kind of virtual machine (vm) migration system, including：Collector cluster, cloud controller, storage cluster and multiple calculate nodes,
Wherein：
The collector cluster, the monitoring data reported for receiving each calculate node, the monitoring according to each calculate node
Data carry out fault detect to each calculate node respectively, it is determined that the calculation of fault node to break down, by the calculation of fault section
The information reporting of point is to the cloud controller；
The storage cluster, for storage virtual machine configuration file；
The cloud controller, for receiving the information of the calculation of fault node, do not sent out from the multiple calculate node
In the calculate node of raw failure, purpose calculate node is determined, is sent to the purpose calculate node and obtains virtual machine configuration
Instruction, and it is empty corresponding to calculation of fault node by virtual machine information corresponding to the purpose calculate node of record, adding
Intend machine information；
The purpose calculate node, the acquisition virtual machine configuration instruction sent for receiving the cloud controller, from
The virtual machine configuration is obtained in the storage cluster, and is configured.
A kind of virtual machine migration method, applied to collector cluster, the virtual machine migration method includes：
Receive the monitoring data that each calculate node reports respectively；
Monitoring data according to each calculate node carries out fault detect to each calculate node respectively, it is determined that the event broken down
Hinder calculate node；
By the information reporting of the calculation of fault node to the cloud controller；The information of the calculation of fault node is tactile
Sending out cloud controller described determines purpose calculate node, is sent to the purpose calculate node and obtains virtual machine configuration instruction
Condition, the acquisition virtual machine configuration instruction are that the purpose calculate node obtains virtual machine configuration text from storage cluster
The foundation of part.
A kind of virtual machine migration method, applied to cloud controller, the virtual machine migration method includes：
The information for the calculation of fault node that collector cluster reports is received, the information of the calculation of fault node is described adopt
Storage cluster determines according to the monitoring data that the calculation of fault node reports；
Never purpose calculate node is determined in the calculate node to break down；
Sent to the purpose calculate node and obtain virtual machine configuration instruction, the acquisition virtual machine configuration refers to
Order is the foundation that the purpose calculate node obtains virtual machine configuration from storage cluster；
It is virtual corresponding to calculation of fault node by virtual machine information corresponding to the purpose calculate node of record, adding
Machine information.
A kind of virtual machine migration method, applied to calculate node, the virtual machine migration method includes：
Acquisition monitoring data；
The monitoring data is reported into collector cluster, so as to the collector cluster according to the monitoring data to institute
State calculate node and carry out fault detect, when the calculate node breaks down, report to cloud controller；
When the calculate node does not break down, if receiving the acquisition virtual machine configuration text that the cloud controller is sent
When part instructs, virtual machine configuration is obtained from storage cluster, and configured.
A kind of virtual machine (vm) migration device, applied to collector cluster, the virtual machine (vm) migration device includes：
Receiving module, the monitoring data reported for receiving each calculate node；
Determining module, fault detect is carried out to each calculate node respectively for the monitoring data according to each calculate node, really
Surely the calculation of fault node to break down；
Sending module, for by the information reporting of the calculation of fault node to the cloud controller；The calculation of fault
The information of node is that the triggering cloud controller determines purpose calculate node, is sent to the purpose calculate node and obtains virtual machine
The condition of configuration file instruction, the acquisition virtual machine configuration instruction is that the purpose calculate node obtains from storage cluster
Take the foundation of virtual machine configuration.
A kind of virtual machine (vm) migration device, applied to cloud controller, the virtual machine (vm) migration device includes：
Receiving module, the information of the calculation of fault node reported for receiving collector cluster, the calculation of fault node
Information be that the collector cluster determines according to the monitoring data of the calculation of fault node；
Determining module, in the calculate node that never breaks down, determining purpose calculate node；
Sending module, virtual machine configuration instruction is obtained for being sent to the purpose calculate node, it is described to obtain void
The instruction of plan machine configuration file is the foundation that the purpose calculate node obtains virtual machine configuration from storage cluster.
A kind of virtual machine (vm) migration device, applied to calculate node, the virtual machine (vm) migration device includes：
Acquisition module, for acquisition monitoring data；
Sending module, for the monitoring data to be reported into collector cluster, so that the collector cluster is according to institute
State monitoring data and fault detect is carried out to the calculate node, when the calculate node breaks down, report to cloud control
Device；
Configuration module, for when the calculate node does not break down, if receive that the cloud controller sends obtains
When taking the virtual machine configuration to instruct, virtual machine configuration is obtained from storage cluster, and configured.
Understood via above-mentioned technical scheme, compared with prior art, a kind of virtual machine that the embodiment of the present application provides moves
Shifting system, report the monitoring data of its own respectively by each calculate node；Monitoring number of the collector cluster according to each calculate node
According to, respectively to each calculate node carry out fault detect, the calculation of fault node to break down is determined, by calculation of fault node
Report to cloud controller；Cloud controller determines purpose calculate node, and the virtual machine in calculation of fault node is migrated to described respectively
Purpose calculate node.So that the virtual machine of calculation of fault node normal operation in purpose calculate node, so as to so that
The key business and crucial application for obtaining enterprise continue to run with.
Brief description of the drawings
In order to illustrate more clearly about the embodiment of the present invention or technical scheme of the prior art, below will be to embodiment or existing
There is the required accompanying drawing used in technology description to be briefly described, it should be apparent that, drawings in the following description are only this
The embodiment of invention, for those of ordinary skill in the art, on the premise of not paying creative work, can also basis
The accompanying drawing of offer obtains other accompanying drawings.
Fig. 1 is the block schematic illustration for the virtual machine (vm) migration system that the embodiment of the present application provides；
Fig. 2 is the signaling process figure for the virtual machine migration method that the embodiment of the present application provides；
Fig. 3 provides each calculate node for the embodiment of the present application and the annexation of each collector in collector cluster is illustrated
Figure；
Fig. 4 provides the detailed framework figure of virtual machine (vm) migration system for the embodiment of the present application；
Fig. 5 provides the virtual machine (vm) migration device applied to collector cluster for the embodiment of the present application, applied to cloud controller
Virtual machine (vm) migration device and, the structural representation applied to the virtual machine (vm) migration device of calculate node.
Embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clear, complete
Site preparation describes, it is clear that described embodiment is only part of the embodiment of the present invention, rather than whole embodiments.It is based on
Embodiment in the present invention, those of ordinary skill in the art are obtained every other under the premise of creative work is not made
Embodiment, belong to the scope of protection of the invention.
The virtual machine (vm) migration system that the embodiment of the present application provides includes collector cluster 11, cloud controller 12, storage cluster
13rd, multiple calculate nodes 14.Specific framework is as shown in Figure 1.
Wherein, multiple calculate nodes 14 can be multiple physical servers 14.Each physical server can include one
Or multiple virtual machines.By software virtual machine, one or more virtual service can be simulated on a physical server
Device is virtual machine, and these virtual machines are operated just as real physical server completely, such as can be with installation operation system
System, installation application program, access Internet resources etc..
Collector cluster 11 can include multiple server groups into cluster.Collector cluster 11 can be to each calculate node
It is monitored.
Cloud controller 12 can be the cluster of multiple servers 12 composition.
Storage cluster 13 can be the cluster of multiple servers 13 composition, can be with storage virtual machine configuration file.
Between collector cluster 11, cloud controller 12, storage cluster 13 and multiple calculate nodes can by wireless or
Wired mode connects.
Cluster refers to associate one group of server, them is looked as a service in terms of many in the external world
Device.Generally connected between physical server in cluster by LAN.
Based on above-mentioned framework, virtual machine migration method is illustrated, as shown in Fig. 2 virtual machine migration method includes：
Step S201：Each calculate node gathers respective monitoring data, and each reports to collector cluster.
Whether the network environment that monitoring data can be used for where reflecting calculate node is good；Virtual machine in calculate node
Whether being capable of normal operation；Whether calculate node can excessively cause poor performance because of load capacity.
Each calculate node all includes physics monitoring agent, and the physics monitoring agent can gather the monitoring number of calculate node
According to.
Above-mentioned each calculate node belongs to a Virtual Cluster, and a Virtual Cluster can include one or more physics collection
Group；Certain above-mentioned each calculate node can also belong to a physical cluster, i.e., the virtual machine (vm) migration that the embodiment of the present application provides
In method, migrated between each calculate node in the physical cluster that virtual machine can be where itself, can also be same at other
Migrated between each calculate node in the different physical clusters of one Virtual Cluster of category.
Step S202：Collector cluster carries out event to each calculate node respectively according to the monitoring data that each calculate node reports
Barrier detection, it is determined that the calculation of fault node to break down.
Network environment is broken down where the calculate node, and the virtual machine in the calculate node is not normally functioning, can be with
The calculate node is defined as calculation of fault node；It is not normally functioning when monitoring data embodies virtual machine in calculate node
When, the calculate node can be defined as to calculation of fault node；When calculate node because when load capacity is excessive, this can be calculated
Node is defined as calculation of fault node.
Step S203：Collector cluster is by the information reporting of the calculation of fault node to the cloud controller.
The information of calculation of fault node can include calculation of fault node mark and calculation of fault node in include
The mark of virtual machine.
Calculation of fault nodal information can also include the mark of itself affiliated physical cluster.
Belong to same physical cluster when the embodiment of the present application provides each calculate node in virtual machine migration method, then failure
The mark of itself affiliated physical cluster can not be included by calculating information.When each calculate node belongs to same Virtual Cluster, due to
Virtual Cluster may include multiple physical clusters, so calculation of fault nodal information needs the mark for including itself affiliated physical cluster
Know, so as to follow-up, send after by the virtual machine (vm) migration in calculation of fault node or to purpose calculate node and obtain configuration
After file instruction, the corresponding relation between virtual machine, calculate node and cluster is updated by cloud controller.
Step S204：Cloud controller receives the information of the calculation of fault node, is not sent out from the multiple calculate node
In the calculate node of raw failure, purpose calculate node is determined.
Cloud controller can determine which calculate node is not in multiple calculate nodes by the mark of calculation of fault node
Break down, purpose calculate node is determined in the calculate node not broken down from these.
Step S205：Cloud controller is sent to the purpose calculate node obtains virtual machine configuration instruction.
Step S206：Purpose calculate node receives the acquisition virtual machine configuration instruction that the cloud controller is sent, from
The virtual machine configuration is obtained in the storage cluster, and is configured.
Virtual machine configuration is used for the hardware information for configuring virtual machine, for example, virtual machine configuration is CPU dinuclears,
Internal memory 4G, disk 80G, then when purpose calculate node is configured according to the virtual machine configuration, can configure corresponding CPU,
Internal memory, disk etc..
Above-mentioned acquisition virtual machine configuration instruction, the path letter that purpose calculate node accesses storage cluster can be included
Breath.
Virtual machine configuration in same Virtual Cluster in each calculate node can be unified.
Step S207：Cloud controller adds failure by virtual machine information corresponding to the purpose calculate node of record
Virtual machine information corresponding to calculate node.
Cloud controller can record cluster identity, calculate node mark with virtual machine mark corresponding relation, when will therefore
, it is necessary to update above-mentioned corresponding relation when hindering virtual machine (vm) migration to the purpose calculate node in calculate node.
Calculation of fault node, the calculate node and purpose of normal operation to break down can be included in calculate node
Calculate node.Wherein purpose calculate node is determined from the calculate node of normal operation.In order to clearly draw in Fig. 2
Signaling diagram, calculate node is divided into calculation of fault node, purpose calculate node and normal operation calculate node.
In a kind of virtual machine migration method that the embodiment of the present application provides, the prison of its own is reported respectively by each calculate node
Control data；Monitoring data of the collector cluster according to each calculate node, fault detect is carried out to each calculate node respectively, determined
The calculation of fault node to break down, calculation of fault node is reported into cloud controller；Cloud controller determines purpose calculate node,
Virtual machine in calculation of fault node is migrated to the purpose calculate node respectively.So that the void of calculation of fault node
Plan machine normal operation in purpose calculate node, so that the key business of enterprise and crucial application continue to run with.
Shown in Fig. 1, there is the calculate node of the function of calculate node described in Fig. 2, it is specific to use in acquisition monitoring data
In：
Gather the heartbeat monitor data of the network environment where the calculate node；Gather virtual machine in the calculate node
Control process monitoring data；Gather loadings in the calculate node.
Wherein, gathering the heartbeat monitor data of the network environment where the calculate node includes：By managing network interface card
The heartbeat monitor data of management network port acquisition management network；Supervised by the heartbeat of the data network interface gathered data network of data network card
Control data；By the heartbeat monitor data for the storage network interface collection storage network for storing network interface card.
The heartbeat monitor data of above-mentioned management network can be the Internet packets survey meter ping of continuous management network；Number
Heartbeat monitor data according to network can be the Internet packets survey meter ping of continuous data network；Store the heartbeat of network
Monitoring data can be the Internet packets survey meter ping of continuous storage network.
Ping is an order under Windows, Unix and linux system.Ping falls within a communication protocol, is
A part for ICP/IP protocol.Utilize " ping " order to check whether network connects, can well analyze and judge net
Network failure.
The virtual machine (vm) migration system or method that the embodiment of the present application provides can apply to cloud platform, the networking bag of cloud platform
Include three network levels：Monitoring management net can be passed through by managing network 31, data network 32 and storage network 33, each calculate node
Network, data network and storage network determine whether its network environment is good.
It is as described in Figure 3 each calculate node and the connection relationship diagram of collector cluster.
As can be seen from Figure 3 each calculate node 14 includes three network interfaces, i.e. management network port 141, data network interface 142
With storage network interface 143, the network interface that it can be network interface card in calculate node that these three network interfaces, which are,.Calculate node can be by managing network interface card
Management network port obtain management network heartbeat monitor data；Data network can be obtained by the data network interface of data network card
Heartbeat monitor data；The heartbeat monitor data of storage network can be obtained by storing the storage network interface of network interface card.
Collector cluster 11, which includes each collector 111 of multiple collectors 111, includes management network port 1111, data network interface
1112 and storage network interface 1113, these three network interfaces can also be the network interface of network interface card.
The number of collector 111 in collector cluster 11 can be identical with the number of calculate node, can also be different, adopts
The number of storage 111 can be less than the number of calculate node, i.e. a collector 111 can collect the monitoring of multiple calculate nodes
Data.
Calculate node and collector are to be connected by the management network port of itself with management network, pass through the data network of itself
Mouth is connected with data network, is connected by the storage network interface of itself with storage network.
Management network port in calculate node, for receive calculate control node or cluster controller by manage network to this
The order that calculate node is sent, such as log on command.
Data network interface in calculate node is referred to as service network port, passes through data network for virtual machine in calculate node
Network communication with the outside world, and calculate node communication with the outside world.
Storage network interface in calculate node, communicated for virtual machine by storing network with storage cluster, Yi Jike
So that the configuration file of virtual machine and data in magnetic disk are stored to storage cluster.
Calculate node can be sent the monitoring data of collection to the number of collector by data network by data network interface
According to network interface.
Virtual machine control process monitoring data can include virtual management process, such as in openstack
Libvrit, and, calculate the nova in finger daemon, such as openstack.
Virtual management process is packaged with virtualization technology, if virtual management process goes wrong, calculate node
Without normal direction, the virtual machine sends control instruction to controller, i.e., can not change the current running status of virtual machine, such as need to suspend
The operation of virtual machine, because the process goes wrong, then the virtual machine can not suspend.
Calculate finger daemon, for by the state of calculate node, the state synchronized of virtual machine into calculate node controller,
If the calculating finger daemon goes wrong, show that the calculate node is unavailable in calculate node controller.The calculating is kept
Shield process, which receives, calculates the control instruction that control node is sent, and the control instruction is forwarded into virtual management process.
Loadings, CPU usage, memory usage and/or disk utilization rate etc. can be included.
Calculate node can gather polytype monitoring data in above-mentioned Fig. 3, it is to be understood that in different applications
In scene, shown in Fig. 1, the collector cluster of the function with the collector cluster described in Fig. 2, event is carried out to calculate node
The monitoring data that barrier detection is used is different, for example, in application scenarios A, calculate node includes multiple virtual machines, and calculate node
CPU usage always 90% or so, memory usage 92% or so, but the calculate node by operator be considered as it is non-therefore
Hinder calculate node.
In order to which the virtual machine migration method for providing the embodiment of the present application can more easily be applied to multiple different answer
With scene, it can be that operator shows a monitoring data selection interface, in the monitoring data selection interface, calculating can be shown
Various types of data of node collection, operator can select to need the prison of which type in the monitoring data selection interface
Control data and fault detect is carried out to calculate node, still exemplified by above-mentioned, then the operator can not select loadings, can be with
Only the heartbeat monitor data of selection network environment and virtual machine control process monitoring data, selectable, monitoring data selection
Interface can also the heartbeat monitor data of display management network, the heartbeat monitor data of data network, the heartbeat prison for storing network
Control data, virtual management process, calculate finger daemon, CPU usage, memory usage etc. detailed monitoring data class
Type.
Collector cluster can be in the monitoring data according to each calculate node, data corresponding with the data type, point
It is other that fault detect is carried out to each calculate node.For example, if operator does not select loadings, collector cluster is to meter
When operator node carries out fault detect, fault detect would not be carried out to it according to loadings.
In the embodiment of the present application, collector cluster carries out fault detect, the monitoring data type used to calculate node
Operator can not be allowed to be selected.No matter operator can be allowed to select the type of monitoring data, operator still can not be allowed
The type of monitoring data is selected, shown in Fig. 1, there is the collector cluster of collector clustering functionality described in Fig. 2, according to each meter
The monitoring data of operator node, respectively to each calculate node carry out fault detect, it is determined that break down calculation of fault node when, it is right
It can be specifically used in each calculate node：
The heartbeat monitor data of the management network are not detected by the first preset time, and in the second preset time
When being not detected by the heartbeat monitor data of the storage network, the calculate node is defined as calculation of fault node；Or, described
The heartbeat monitor data of the management network are not detected by first preset time, and number is not detected by the 3rd preset time
According to network heartbeat monitor data when, the calculate node is defined as calculation of fault node；Or, in first preset time
It is not detected by the heartbeat monitor data of the management network, and virtual machine control process is in and stopped in the 4th preset time
Only during running status, the calculate node is defined as calculation of fault node；Or, detect that the calculate node loadings are big
When equal to predetermined threshold value, the calculate node is defined as calculation of fault node.
It is above-mentioned when detecting that the calculate node loadings are more than or equal to predetermined threshold value, the calculate node is defined as
Calculation of fault node, it can include：When detect the calculate node cpu busy percentage be more than or equal to the first predetermined threshold value or
When memory usage is more than or equal to the second predetermined threshold value, the calculate node is defined as calculation of fault node.
Above-mentioned first preset time, the second preset time, the 3rd preset time and the 4th preset time can also may be used with identical
With difference, specifically can be according to actual conditions depending on.
The explanation of each network interface to Fig. 3, it can be seen that when management network goes wrong, calculate node can not connect
Receive the order for calculating control node or cluster controller transmission.In this case it is calculation of fault that the calculate node, which can be determined,
Node.When the calculate node allows to receive the order for calculating control node or cluster controller transmission always,
The calculate node is considered non-faulting calculate node.
When data network goes wrong, calculate node can not interact with the virtual machine of itself with the external world, now
It needs to be determined that the calculate node breaks down.When the calculate node allows not interact with the external world, the calculate node can be with
It is considered as non-faulting node.
When storing network and going wrong, virtual machine can not be communicated with storage cluster, according to different application scenarios,
The calculate node may be considered calculation of fault node by operator, it is also possible to be considered non-faulting node by operator.
I.e. when any of the above-described network goes wrong, in application scenes, operator will be considered that the calculate node is
Calculation of fault node, in other application scenarios, operator will be considered that the calculate node is non-faulting calculate node.If only
One network goes wrong, it is confirmed that the calculate node is calculation of fault node, certain operations person will be considered into non-faulting
Calculate node is defined as calculation of fault node.
Applicant has found by continuous research, is gone wrong simultaneously when managing network and storing two networks of network, or pipe
Reason network and data network go wrong simultaneously, or management network goes wrong and virtual machine control process goes wrong, or negative
Carrying capacity data are excessive, no matter in which kind of application scenarios, this calculating is exactly calculation of fault node, therefore work out above-mentioned to meter
The fault detection method of operator node.
In other application or above-mentioned any combination goes wrong, then assert that calculate node is malfunctioning node, this
When collector clustering functionality collector cluster, in the monitoring data according to each calculate node, each calculate node is carried out respectively
Fault detect, it is determined that break down calculation of fault node when, can be specifically used for for each calculate node：
The heartbeat monitor data of the management network are not detected by the first preset time, in the second preset time not
The heartbeat monitor data of the storage network are detected, the heartbeat monitor number of data network is not detected by the 3rd preset time
According to the virtual machine control process is in run-stopping status and detects that the calculate node is born in the 4th preset time
When carrying capacity data occur more than or equal to one or more situations in predetermined threshold value, the calculate node is defined as calculation of fault section
Point.
It is understood that during calculate node negligible amounts, shown in Fig. 1, with the collector cluster described in Fig. 2
The collector cluster of function, can be in real time by the database in the supervising data storage that calculate node reports to collector cluster
In, when calculate node quantity is larger, if in real time by supervising data storage into database, the friendship with database can be caused
Mutually excessively frequently, therefore, optionally, collector cluster, can be also used for：By each monitoring data of each calculate node received
Cached；When the monitoring data of caching reaches predetermined number, the supervising data storage of the predetermined number is adopted to described
In database in storage cluster.
Optionally, shown in Fig. 1, there is the cloud controller of cloud controller function described in Fig. 2, the collection is received described
After the information for the calculation of fault node that device cluster reports, it is additionally operable to：
Sent to the calculation of fault node and be confirmed whether the information that breaks down；Receive the calculation of fault node feeding back
Confirmation when, triggering perform from the multiple calculate node determine purpose calculate node, to the purpose calculate node
Send and obtain virtual machine configuration instruction.I.e. cloud controller carries out failed synchronization affirmation mechanism with calculation of fault node.
Optionally, the cloud controller that the embodiment of the present application provides can have forcedown mechanism, when calculation of fault node
Quantity when being less than preset failure quantity, in order to avoid non-faulting calculate node is confirmed as calculation of fault section by collector cluster by mistake
Point, cloud controller can carry out failed synchronization affirmation mechanism to calculation of fault node.When calculation of fault number of nodes is more than or equal to
During preset failure quantity, if cloud controller still carries out failed synchronization affirmation mechanism with each calculation of fault node, it may lead
Cause whole transition process slow, therefore, the embodiment of the present application can provide failed synchronization affirmation mechanism selection interface, work as operator
Failed synchronization affirmation mechanism is selected, then performs above-mentioned failed synchronization affirmation mechanism, if operator does not select, does not perform above-mentioned
Failed synchronization affirmation mechanism.I.e. when cloud controller receives calculation of fault node, purpose calculate node is directly determined, and move
Move, and without failed synchronization affirmation mechanism, so as to improve migration velocity.
Shown in Fig. 1, have cloud controller function described in Fig. 2 cloud controller, from the multiple calculate node not
In the calculate node to break down, when determining purpose calculate node, in one implementation, it is specifically used for：
The scheduling parameter of each calculate node is monitored in real time, and the scheduling parameter includes：Resource residual amount or energy input add
Enter the time sequencing of cloud platform；Scheduling parameter is met to the calculate node of scheduling strategy, it is determined that being purpose calculate node.
Scheduling parameter is different, and scheduling strategy is different, and when scheduling parameter is resource residual amount, scheduling strategy can be will money
Surplus maximum calculate node in source is as purpose calculate node, or using the calculate node of resource residual amount minimum as purpose meter
Operator node；When scheduling parameter for that can take, scheduling strategy can be using the minimum calculate node of energy consumption as purpose calculate node；
When scheduling parameter is adds the time sequencing of cloud platform, scheduling strategy can be that will add cloud platform time most long calculating section
Point be used as purpose calculate node (preferentially using old calculate node), or using addition cloud platform time most short calculate node as
Purpose calculate node (preferentially uses new calculate node).
Above-mentioned resource residual amount can be integrated ratio calculating acquisition by CPU, internal memory and disk or refer to
" 1-CPU occupancy " or " utilization rate of 1- internal memories " or " 1- disks occupancy ".
Scheduling parameter have it is a variety of, can be determined in different application scenarios according to different scheduling parameters purpose calculate section
Point.Each calculate node described in Fig. 1 belongs to same Virtual Cluster, but when belonging to different physical clusters, in different things
When managing selection purpose calculate node in cluster, scheduling parameter can be different.Based on this, the virtual machine migration method that the application provides
Or in system, cloud controller is specifically used for when monitoring the scheduling parameter of each calculate node in real time：Determine that purpose calculate node is adjusted
Spend selected purpose scheduling strategy in policy selection interface；Monitor in real time in each calculate node with the purpose scheduling strategy pair
The scheduling parameter answered.
Operator can select current application scene, or different physics in purpose calculate node scheduling strategy selection interface
Scheduling strategy required for cluster, the scheduling strategy that operator selects is referred to as purpose scheduling strategy.Cloud controller can basis
Different purpose scheduling strategies, scheduling parameter corresponding with purpose scheduling strategy in each calculate node is monitored in real time.
Shown in Fig. 1, there is the cloud controller of cloud controller function described in Fig. 2, in the calculation of fault node, often
One virtual machine, cloud controller are meeting scheduling parameter the calculate node of scheduling strategy, it is determined that when being purpose calculate node, tool
Body is used for：Scheduling parameter in the calculate node not broken down currently is met to the calculate node of scheduling strategy, it is virtual as this
The purpose calculate node of machine.
The purpose calculate node that scheduling parameter meets scheduling strategy is chosen in all non-faulting calculate nodes, no matter the mesh
Calculate node whether with calculation of fault node belong to same physical cluster.
Preferably, the purpose that scheduling parameter meets scheduling strategy is first searched in the physical cluster belonging to calculation of fault node
Calculate node, if it is not, searching purpose calculate node in other physical clusters again.For the calculation of fault node
In, each virtual machine, cloud controller is meeting scheduling parameter the calculate node of scheduling strategy, it is determined that being purpose calculate node
When, it is specifically used for：
When the calculate node that scheduling parameter meets scheduling strategy be present in the affiliated cluster of calculation of fault node, will adjust
Degree parameter meets purpose calculate node of the calculate node as the virtual machine of scheduling strategy.
, will when the scheduling parameter of calculate node in the affiliated cluster of calculation of fault node is unsatisfactory for the scheduling strategy
Scheduling parameter meets the calculate node of the scheduling strategy in other clusters, the purpose calculate node as the virtual machine.
To sum up, determine that the method for purpose calculate node can apply in the calculate node that cloud controller never breaks down
In single cluster, i.e. each calculate node mentioned in Fig. 1 belongs to same physical cluster.It can also be applied to across cluster, i.e. Fig. 1
Mentioned in each calculate node belong to same Virtual Cluster, virtual machine cluster can include multiple physical clusters, here across
Cluster refers to across physical cluster.
The embodiment of the present application additionally provides a kind of virtual machine (vm) migration system, as shown in figure 1, including collector cluster 11, cloud
Controller 12, storage cluster 13, multiple calculate nodes 14, wherein：
Collector cluster 11 has the function of collector cluster as shown in Figure 2；Cloud controller 12 has as described in Figure 2
The function of cloud controller；Storage cluster 13 is stored with virtual machine configuration；Each calculate node tool in multiple calculate nodes 14
There is the function of the calculate node described in Fig. 2.
The detailed framework of virtual machine (vm) migration system in collector cluster 11 as shown in figure 4, can include：(the example of database 112
Such as Mysql databases), memory 113, caching system 114 (such as redis), Analysis server for caching monitoring data
115th, multiple collectors 111.
Collector 111 can be server.
When caching system 114 caches monitoring data to predetermined number, by the supervising data storage of predetermined number to number
According to storehouse 112.Caching system 114 can be database.
Queue is included in memory 113.It is understood that each calculate node reports to the monitoring data of collector cluster
Quantity may be very big, therefore queue is set in collector cluster, can be according to each calculate node by each monitoring data
On call time and sorted in queue.
Analysis server 115, for obtaining the monitoring data of each calculate node from database 112, and determine to be out of order
Calculate node.
Calculate node 14 include monitoring agent, calculate finger daemon, virtual management process, monitoring agent include report into
Journey, collection process, monitoring agent is by gathering process acquisition monitoring data, by reporting process that monitoring data is reported into collection
Device 114.
Cloud controller 12 includes election process and scheduling process, and cloud controller determines that purpose calculates section by election process
Point, virtual machine configuration instruction will be obtained by the process of dispatching and sent to purpose calculate node.
Virtual machine (vm) migration system in the embodiment of the present application can cause cloud platform to possess failure automatic switching capabilities, work as meter
When operator node breaks down, the virtual machine in calculation of fault node can be migrated, so as to ensure that the reliable of virtual machine
Property.
The virtual machine (vm) migration device provided below the embodiment of the present application is described, virtual machine (vm) migration dress described below
Putting can be mutually to should refer to above-described virtual machine migration method.
The embodiment of the present application additionally provides the virtual machine (vm) migration device applied to collector cluster, applied to cloud controller
Virtual machine (vm) migration device and, the virtual machine (vm) migration device applied to calculate node.As shown in figure 5, in above three device
The connection relationship diagram of modules.
Virtual machine (vm) migration device 51 applied to collector cluster includes：Receiving module 511, determining module 512 and hair
Send module 513；Virtual machine (vm) migration device 52 applied to cloud controller includes：Receiving module 521, determining module 522 and transmission
Module 523；Virtual machine (vm) migration device 53 applied to calculate node includes：Acquisition module 531, sending module 532 and configuration
Module 533, wherein：
Acquisition module 531, for acquisition monitoring data.
Sending module 532, for the monitoring data to be reported into receiving module 511.
Receiving module 511, the monitoring data reported for receiving sending module 532.
Determining module 512, fault detect is carried out to each calculate node respectively for the monitoring data according to each calculate node,
It is determined that the calculation of fault node to break down.
Sending module 513, for by the information reporting of the calculation of fault node to receiving module 521.
Receiving module 521, the information of the calculation of fault node reported for receiving sending module 513.
Determining module 522, in the calculate node that never breaks down, determining purpose calculate node.
Sending module 523, virtual machine configuration is obtained for being sent to the configuration module 533 of the purpose calculate node
Instruction.
Configuration module 533, for when the calculate node does not break down, if receiving what the cloud controller was sent
When obtaining virtual machine configuration instruction, virtual machine configuration is obtained from storage cluster, and configured.
The virtual machine (vm) migration device applied to calculate node of the embodiment of the present application offer, the void applied to collector cluster
Intend moving apparatus and applied in the virtual moving apparatus of cloud controller, report acquisition module 531 to gather by each sending module 532
Monitoring data；Monitoring data of the determining module 512 according to each calculate node, fault detect is carried out to each calculate node respectively,
The calculation of fault node to break down is determined, calculation of fault node is reported to receiving module 521 by sending module 513；It is determined that
Module 522 determines purpose calculate node, and sending module 523 sends to the configuration module 533 of the purpose calculate node and obtains void
Plan machine configuration file instructs, and configuration module 533 obtains virtual machine configuration from storage cluster, and is configured.So as to realize
Virtual machine in calculation of fault node migrated to the purpose calculate node respectively.So that the void of calculation of fault node
Plan machine normal operation in purpose calculate node, so that the key business of enterprise and crucial application continue to run with.
The embodiment of the present application provides the alternative construction of acquisition module in the virtual machine (vm) migration device applied to calculate node,
It is specific as follows：Acquisition module can include：
First collecting unit, for gathering the heartbeat monitor data of the network environment where the calculate node.
Second collecting unit, process monitoring data are controlled for gathering virtual machine in the calculate node.
Second collecting unit, for gathering loadings in the calculate node.
The embodiment of the present application additionally provides the alternative construction applied to the first collecting unit, specific as follows：First collection is single
Member includes：
First collection subelement, the heartbeat monitor data for the management network port acquisition management network by managing network interface card.
Second collection subelement, the heartbeat monitor data for the data network interface gathered data network by data network card.
3rd collection subelement, the heartbeat monitor data for the storage network interface collection storage network by storing network interface card.
The embodiment of the present application additionally provides in the virtual machine (vm) migration device applied to collector cluster that determining module one kind can
Structure is selected, it is specific as follows：Determining module includes：
Receiving unit, for receiving the data type being selected in display monitoring data selection interface；
Detection unit, for data corresponding with the data type in the monitoring data according to each calculate node, difference
Fault detect is carried out to each calculate node.
The embodiment of the present application additionally provides the another of determining module in the virtual machine (vm) migration device applied to collector cluster
Kind alternative construction, it is specific as follows：
First determining unit, for being not detected by the heartbeat monitor data of the management network in the first preset time,
And when the heartbeat monitor data of the storage network are not detected by the second preset time, the calculate node is defined as failure
Calculate node；
Or, second determining unit, for being not detected by the heartbeat prison of the management network in first preset time
When controlling data, and being not detected by the 3rd preset time the heartbeat monitor data of data network, the calculate node is defined as
Calculation of fault node；
Or, the 3rd determining unit, for being not detected by the heartbeat prison of the management network in first preset time
Data are controlled, and when virtual machine control process is in run-stopping status in the 4th preset time, the calculate node is true
It is set to calculation of fault node；
Or, the 4th determining unit, during for detecting that the calculate node operation load capacity is more than or equal to predetermined threshold value, general
The calculate node is defined as calculation of fault node.
The embodiment of the present application additionally provides can also include following knot applied to the virtual machine (vm) migration device of collector cluster
Structure, it is specific as follows：
Cache module, for each monitoring data of each calculate node received to be cached；
Data module is sent, for when the monitoring data of caching reaches predetermined number, by the monitoring of the predetermined number
In database in data storage to the collector cluster.
The embodiment of the present application additionally provides can also include following structure applied to the virtual machine (vm) migration device of cloud controller,
It is specific as follows：
Confirmation module is sent, is confirmed whether the information that breaks down for being sent to the calculation of fault node；
First trigger module, during confirmation for receiving the calculation of fault node feeding back, trigger cloud controller
In determining module.
The embodiment of the present application additionally provides can also include following structure applied to the virtual machine (vm) migration device of cloud controller,
It is specific as follows：
Second trigger module, for receiving in failed synchronization affirmation mechanism selection interface, failed synchronization affirmation mechanism quilt
During selection, triggering sends confirmation module.
One kind that the embodiment of the present application additionally provides determining module in the virtual machine (vm) migration device applied to cloud controller can
Structure is selected, it is specific as follows：Determining module includes：
Monitoring unit, for monitoring the scheduling parameter of each calculate node in real time, the scheduling parameter includes：Resource residual amount
Or energy input or the time sequencing for adding cloud platform；
Determining unit, for scheduling parameter to be met to the calculate node of scheduling strategy, it is determined that being purpose calculate node.
The embodiment of the present application additionally provides the monitoring in determining module in the virtual machine (vm) migration device applied to cloud controller
A kind of alternative construction of unit, it is specific as follows：Monitoring unit includes：
First determination subelement, for determining selected purpose scheduling in purpose calculate node scheduling strategy selection interface
Strategy；
Subelement is monitored, for monitoring scheduling parameter corresponding with the purpose scheduling strategy in each calculate node in real time.
The embodiment of the present application additionally provides the determination in determining module in the virtual machine (vm) migration device applied to cloud controller
A kind of alternative construction of unit, it is specific as follows：Determining unit includes：
Second determination subelement, for scheduling parameter in the currently calculate node of non-failure to be met to the calculating of scheduling strategy
Node, the purpose calculate node as the virtual machine.
The embodiment of the present application additionally provides the determination in determining module in the virtual machine (vm) migration device applied to cloud controller
A kind of alternative construction of unit, it is specific as follows：Determining unit includes：
3rd determination subelement, for when exist in the affiliated cluster of calculation of fault node scheduling parameter meet scheduling plan
During the calculate node omited, scheduling parameter is met to purpose calculate node of the calculate node as the virtual machine of scheduling strategy；
4th determination subelement, for being discontented with when the scheduling parameter of calculate node in the affiliated cluster of calculation of fault node
During the foot scheduling strategy, scheduling parameter in other clusters is met to the calculate node of the scheduling strategy, as described virtual
The purpose calculate node of machine.
Finally, it is to be noted that, herein, such as first and second or the like relational terms be used merely to by
One entity or operation make a distinction with another entity or operation, and not necessarily require or imply these entities or operation
Between any this actual relation or order be present.Moreover, term " comprising ", "comprising" or its any other variant meaning
Covering including for nonexcludability, so that process, method, article or equipment including a series of elements not only include that
A little key elements, but also the other element including being not expressly set out, or also include for this process, method, article or
The intrinsic key element of equipment.In the absence of more restrictions, the key element limited by sentence "including a ...", is not arranged
Except other identical element in the process including the key element, method, article or equipment being also present.
Each embodiment is described by the way of progressive in this specification, what each embodiment stressed be and other
The difference of embodiment, between each embodiment identical similar portion mutually referring to.
The foregoing description of the disclosed embodiments, professional and technical personnel in the field are enable to realize or using the application.
A variety of modifications to these embodiments will be apparent for those skilled in the art, as defined herein
General Principle can be realized in other embodiments in the case where not departing from spirit herein or scope.Therefore, the application
The embodiments shown herein is not intended to be limited to, and is to fit to and principles disclosed herein and features of novelty phase one
The most wide scope caused.
Claims (27)
- A kind of 1. virtual machine (vm) migration system, it is characterised in that including：Collector cluster, cloud controller, storage cluster and multiple Calculate node, wherein：The collector cluster, the monitoring data reported for receiving each calculate node, the monitoring data according to each calculate node Fault detect is carried out to each calculate node respectively, it is determined that the calculation of fault node to break down, by the calculation of fault node Information reporting is to the cloud controller；The storage cluster, for storage virtual machine configuration file；The cloud controller, for receiving the information of the calculation of fault node, event does not occur from the multiple calculate node In the calculate node of barrier, purpose calculate node is determined, is sent to the purpose calculate node and obtains virtual machine configuration instruction, And virtual machine letter corresponding to calculation of fault node in virtual machine information corresponding to the purpose calculate node of record, will be added Breath；The purpose calculate node, the acquisition virtual machine configuration instruction sent for receiving the cloud controller, from described The virtual machine configuration is obtained in storage cluster, and is configured.
- 2. virtual machine (vm) migration system according to claim 1, it is characterised in that the cloud controller is receiving the failure meter After the information of operator node, it is additionally operable to：Sent to the calculation of fault node and be confirmed whether the information that breaks down；When receiving the confirmation of the calculation of fault node feeding back, triggering performs not to be occurred from the multiple calculate node In the calculate node of failure, purpose calculate node is determined, sending acquisition virtual machine configuration to the purpose calculate node refers to Order；The calculation of fault node is additionally operable to, receive that the cloud controller sends be confirmed whether to break down information when, to The cloud controller feedback acknowledgment information.
- 3. virtual machine (vm) migration system according to claim 2, it is characterised in that the cloud controller is additionally operable to：Receive in failed synchronization affirmation mechanism selection interface, during the selected message of failed synchronization affirmation mechanism, triggering performs Sent to the calculation of fault node and be confirmed whether the information that breaks down.
- 4. virtual machine (vm) migration system according to claim 1, it is characterised in that the cloud controller is from the multiple calculating In the calculate node not broken down in node, when determining purpose calculate node, it is specifically used for：The scheduling parameter of each calculate node is monitored in real time, and the scheduling parameter includes：Resource residual amount or energy input add cloud The time sequencing of platform；Scheduling parameter is met to the calculate node of scheduling strategy, it is determined that being purpose calculate node.
- 5. virtual machine (vm) migration system according to claim 4, it is characterised in that the cloud controller is monitoring each calculating in real time During the scheduling parameter of node, it is specifically used for：Determine selected purpose scheduling strategy in purpose calculate node scheduling strategy selection interface；Scheduling parameter corresponding with the purpose scheduling strategy in each calculate node is monitored in real time.
- 6. according to the virtual machine (vm) migration system of claim 4 or 5, it is characterised in that the cloud controller is by scheduling parameter Meet the calculate node of scheduling strategy, it is determined that when being purpose calculate node, be specifically used for：When the calculate node that the scheduling parameter meets the scheduling strategy be present in the affiliated cluster of calculation of fault node, The scheduling parameter is met to purpose calculate node of the calculate node as the virtual machine of the scheduling strategy；, will when the scheduling parameter of calculate node in the affiliated cluster of calculation of fault node is unsatisfactory for the scheduling strategy Scheduling parameter meets the calculate node of the scheduling strategy described in other clusters, and the purpose as the virtual machine calculates section Point.
- 7. according to any virtual machine (vm) migration system of claim 1, it is characterised in that the calculate node is by monitoring data When reporting to collector cluster, it is specifically used for：Gather the heartbeat monitor data of the network environment where the calculate node；Gather virtual machine in the calculate node and control process monitoring data；Gather loadings in the calculate node.
- 8. virtual machine (vm) migration system according to claim 7, it is characterised in that the network environment includes management network, number According to network and storage network, the heartbeat monitor data of network environment of the calculate node where the calculate node is gathered When, it is specifically used for：Management network port by managing network interface card gathers the heartbeat monitor data of the management network；The heartbeat monitor data of the data network are gathered by the data network interface of data network card；Storage network interface by storing network interface card gathers the heartbeat monitor data of the storage network.
- 9. virtual machine (vm) migration system according to claim 8, it is characterised in that the collector cluster is calculating section according to each Point monitoring data respectively to each calculate node carry out fault detect, it is determined that break down calculation of fault node when, for every One calculate node is specifically used for：The heartbeat monitor data of the management network are not detected by the first preset time, and are not examined in the second preset time When measuring the heartbeat monitor data of the storage network, the calculate node is defined as calculation of fault node；Or, the heartbeat monitor data of the management network are not detected by first preset time, and when the 3rd is default In when being not detected by the heartbeat monitor data of data network, the calculate node is defined as calculation of fault node；Or, the heartbeat monitor data of the management network are not detected by first preset time, and when the 4th is default When the interior virtual machine control process is in run-stopping status, the calculate node is defined as calculation of fault node；Or, when detecting that the calculate node operation load capacity is more than or equal to predetermined threshold value, the calculate node is defined as event Hinder calculate node.
- 10. virtual machine (vm) migration system according to claim 1, it is characterised in that the collector cluster is according to each calculating When the monitoring data of node carries out fault detect to each calculate node respectively, it is specifically used for：Receive the data type being selected in display monitoring data selection interface；According to data corresponding with the data type in the monitoring data of each calculate node, event is carried out to each calculate node respectively Barrier detection.
- 11. a kind of virtual machine migration method, it is characterised in that applied to collector cluster, the virtual machine migration method includes：Receive the monitoring data that each calculate node reports respectively；Monitoring data according to each calculate node carries out fault detect to each calculate node respectively, it is determined that the failure meter to break down Operator node；By the information reporting of the calculation of fault node to the cloud controller；The information of the calculation of fault node is triggering institute State cloud controller and determine purpose calculate node, the bar for obtaining virtual machine configuration instruction is sent to the purpose calculate node Part, the acquisition virtual machine configuration instruction is that the purpose calculate node obtains virtual machine configuration from storage cluster Foundation.
- 12. the virtual machine migration method according to claim 11, it is characterised in that the monitoring number according to each calculate node Include according to fault detect is carried out to each calculate node respectively：Receive the data type being selected in display monitoring data selection interface；According to data corresponding with the data type in the monitoring data of each calculate node, event is carried out to each calculate node respectively Barrier detection.
- 13. according to the virtual machine migration method of claim 11 or 12, it is characterised in that the monitoring data includes management net It is virtual in the heartbeat monitor data of network, the heartbeat monitor data of data network, the heartbeat monitor data for storing network, calculate node Machine controls process monitoring data and calculate node loadings, and the monitoring data according to each calculate node is respectively to each meter Operator node carries out fault detect, it is determined that the calculation of fault node to break down, includes for each calculate node：The heartbeat monitor data of the management network are not detected by the first preset time, and are not examined in the second preset time When measuring the heartbeat monitor data of the storage network, the calculate node is defined as calculation of fault node；Or, the heartbeat monitor data of the management network are not detected by first preset time, and when the 3rd is default In when being not detected by the heartbeat monitor data of data network, the calculate node is defined as calculation of fault node；Or, the heartbeat monitor data of the management network are not detected by first preset time, and when the 4th is default When the interior virtual machine control process is in run-stopping status, the calculate node is defined as calculation of fault node；Or, when detecting that the calculate node operation load capacity is more than or equal to predetermined threshold value, the calculate node is defined as event Hinder calculate node.
- 14. according to any virtual machine migration method of claim 11 to 13, it is characterised in that also include：Each monitoring data of each calculate node received is cached；When the monitoring data of caching reaches predetermined number, by the supervising data storage of the predetermined number to the collector collection In database in group.
- 15. a kind of virtual machine migration method, it is characterised in that applied to cloud controller, the virtual machine migration method includes：The information for the calculation of fault node that collector cluster reports is received, the information of the calculation of fault node is the collector Cluster determines according to the monitoring data that the calculation of fault node reports；Never purpose calculate node is determined in the calculate node to break down；Sent to the purpose calculate node and obtain virtual machine configuration instruction, the acquisition virtual machine configuration instruction is The purpose calculate node obtains the foundation of virtual machine configuration from storage cluster；In virtual machine information corresponding to the purpose calculate node of record, virtual machine letter corresponding to calculation of fault node will be added Breath.
- 16. the virtual machine migration method according to claim 15, it is characterised in that received described on the collector cluster After the information of the calculation of fault node of report, in addition to：Sent to the calculation of fault node and be confirmed whether the information that breaks down；When receiving the confirmation of the calculation of fault node feeding back, triggering performs and determines mesh from the multiple calculate node Calculate node, to the purpose calculate node send obtain virtual machine configuration instruction.
- 17. the virtual machine migration method according to claim 16, it is characterised in that also include：Receive in failed synchronization affirmation mechanism selection interface, when failed synchronization affirmation mechanism is chosen, triggering is performed to described Calculate node, which is sent, is confirmed whether the information that breaks down.
- 18. the virtual machine migration method according to claim 15, it is characterised in that the calculate node never to break down Middle determination purpose calculate node includes：The scheduling parameter of each calculate node is monitored in real time, and the scheduling parameter includes：Resource residual amount or energy input add cloud The time sequencing of platform；The scheduling parameter is met to the calculate node of the scheduling strategy, it is determined that being purpose calculate node.
- 19. the virtual machine migration method according to claim 18, it is characterised in that the tune for monitoring each calculate node in real time Degree parameter includes：Determine selected purpose scheduling strategy in purpose calculate node scheduling strategy selection interface；Scheduling parameter corresponding with the purpose scheduling strategy in each calculate node is monitored in real time.
- 20. according to the virtual machine migration method of claim 18 or 19, it is characterised in that for the calculation of fault node In, each virtual machine, the calculate node that the scheduling parameter is met to the scheduling strategy, it is determined that being purpose calculate node Including：By currently scheduling parameter meets the calculate node of the scheduling strategy described in the calculate node of non-failure, as the void The purpose calculate node of plan machine.
- 21. according to the virtual machine migration method of claim 18 or 19, it is characterised in that for the calculation of fault node In, each virtual machine, the calculate node that the scheduling parameter is met to the scheduling strategy, it is determined that being purpose calculate node Including：When the calculate node that the scheduling parameter meets the scheduling strategy be present in the affiliated cluster of calculation of fault node, The scheduling parameter is met to purpose calculate node of the calculate node as the virtual machine of the scheduling strategy；, will when the scheduling parameter of calculate node in the affiliated cluster of calculation of fault node is unsatisfactory for the scheduling strategy Scheduling parameter meets the calculate node of the scheduling strategy described in other clusters, and the purpose as the virtual machine calculates section Point.
- 22. a kind of virtual machine migration method, it is characterised in that applied to calculate node, the virtual machine migration method includes：Acquisition monitoring data；The monitoring data is reported into collector cluster, so as to the collector cluster according to the monitoring data to the meter Operator node carries out fault detect, when the calculate node breaks down, reports to cloud controller；When the calculate node does not break down, refer to if receiving the acquisition virtual machine configuration that the cloud controller is sent When making, virtual machine configuration is obtained from storage cluster, and configured.
- 23. the virtual machine migration method according to claim 22, it is characterised in that the acquisition monitoring data include：Gather the heartbeat monitor data of the network environment where the calculate node；Gather virtual machine in the calculate node and control process monitoring data；Gather loadings in the calculate node.
- 24. the virtual machine migration method according to claim 23, it is characterised in that the network environment include management network, Data network and storage network, the heartbeat monitor data of the network environment where the collection calculate node include：Management network port by managing network interface card gathers the heartbeat monitor data of the management network；The heartbeat monitor data of the data network are gathered by the data network interface of data network card；Storage network interface by storing network interface card gathers the heartbeat monitor data of the storage network.
- 25. a kind of virtual machine (vm) migration device, it is characterised in that applied to collector cluster, the virtual machine (vm) migration device includes：Receiving module, the monitoring data reported for receiving each calculate node；Determining module, fault detect is carried out to each calculate node respectively for the monitoring data according to each calculate node, it is determined that hair The calculation of fault node of raw failure；Sending module, for by the information reporting of the calculation of fault node to the cloud controller；The calculation of fault node Information be that the triggering cloud controller determines purpose calculate node, sent to the purpose calculate node and obtain virtual machine configuration The condition of file instruction, the acquisition virtual machine configuration instruction is that the purpose calculate node obtains void from storage cluster The foundation of plan machine configuration file.
- 26. a kind of virtual machine (vm) migration device, it is characterised in that applied to cloud controller, the virtual machine (vm) migration device includes：Receiving module, the information of the calculation of fault node reported for receiving collector cluster, the letter of the calculation of fault node Breath is that the collector cluster determines according to the monitoring data of the calculation of fault node；Determining module, in the calculate node that never breaks down, determining purpose calculate node；Sending module, virtual machine configuration instruction, the acquisition virtual machine are obtained for being sent to the purpose calculate node Configuration file instruction is the foundation that the purpose calculate node obtains virtual machine configuration from storage cluster.
- 27. a kind of virtual machine (vm) migration device, it is characterised in that applied to calculate node, the virtual machine (vm) migration device includes：Acquisition module, for acquisition monitoring data；Sending module, for the monitoring data to be reported into collector cluster, so that the collector cluster is according to the prison Control data and fault detect is carried out to the calculate node, when the calculate node breaks down, report to cloud controller；Configuration module, for when the calculate node does not break down, if it is empty to receive the acquisition that the cloud controller is sent When plan machine configuration file instructs, virtual machine configuration is obtained from storage cluster, and configured.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610481831.4A CN107544839B (en) | 2016-06-27 | 2016-06-27 | Virtual machine migration system, method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610481831.4A CN107544839B (en) | 2016-06-27 | 2016-06-27 | Virtual machine migration system, method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107544839A true CN107544839A (en) | 2018-01-05 |
| CN107544839B CN107544839B (en) | 2021-05-25 |
Family
ID=60962526
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610481831.4A Active CN107544839B (en) | 2016-06-27 | 2016-06-27 | Virtual machine migration system, method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107544839B (en) |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108132829A (en) * | 2018-01-11 | 2018-06-08 | 郑州云海信息技术有限公司 | A kind of high available virtual machine realization method and system based on OpenStack |
| CN108334402A (en) * | 2018-03-07 | 2018-07-27 | 山东超越数控电子股份有限公司 | A kind of the virtual management system and its resource regulating method of non-stop layer framework |
| CN108762993A (en) * | 2018-06-06 | 2018-11-06 | 山东超越数控电子股份有限公司 | A kind of virtual-machine fail moving method and device based on artificial intelligence |
| CN109151045A (en) * | 2018-09-07 | 2019-01-04 | 北京邮电大学 | A kind of distribution cloud system and monitoring method |
| CN109818785A (en) * | 2019-01-15 | 2019-05-28 | 无锡华云数据技术服务有限公司 | A kind of data processing method, server cluster and storage medium |
| CN110445662A (en) * | 2019-08-29 | 2019-11-12 | 上海仪电（集团）有限公司中央研究院 | OpenStack control node is adaptively switched to the method and device of calculate node |
| CN110659109A (en) * | 2019-09-26 | 2020-01-07 | 上海仪电（集团）有限公司中央研究院 | Openstack cluster virtual machine monitoring system and method |
| CN110837451A (en) * | 2018-08-16 | 2020-02-25 | 中国移动通信集团重庆有限公司 | Processing method, device, equipment and medium for high availability of virtual machine |
| CN112073518A (en) * | 2020-09-09 | 2020-12-11 | 杭州海康威视系统技术有限公司 | Cloud storage system, cloud storage system management method and central management node |
| CN112994977A (en) * | 2021-02-24 | 2021-06-18 | 紫光云技术有限公司 | Method for high availability of server host |
| CN113407301A (en) * | 2021-05-22 | 2021-09-17 | 济南浪潮数据技术有限公司 | Virtual machine monitoring method, system, storage medium and equipment |
| CN114064217A (en) * | 2021-11-29 | 2022-02-18 | 建信金融科技有限责任公司 | Node virtual machine migration method and device based on OpenStack |
| CN114090184A (en) * | 2021-11-26 | 2022-02-25 | 中国电信集团系统集成有限责任公司 | Method and equipment for realizing high availability of virtualization cluster |
| CN114217905A (en) * | 2021-12-17 | 2022-03-22 | 北京志凌海纳科技有限公司 | High-availability recovery processing method and system for virtual machine |
| CN114760313A (en) * | 2020-12-29 | 2022-07-15 | 中国联合网络通信集团有限公司 | Service scheduling method and service scheduling device |
| CN115766405A (en) * | 2023-01-09 | 2023-03-07 | 苏州浪潮智能科技有限公司 | Fault processing method, device, equipment and storage medium |
Citations (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102170474A (en) * | 2011-04-22 | 2011-08-31 | 广州杰赛科技股份有限公司 | Method and system for dynamic scheduling of virtual resources in cloud computing network |
| CN102708000A (en) * | 2012-04-19 | 2012-10-03 | 北京华胜天成科技股份有限公司 | System and method for realizing energy consumption control through virtual machine migration |
| CN102819465A (en) * | 2012-06-29 | 2012-12-12 | 华中科技大学 | Failure recovery method in virtualization environment |
| CN103064733A (en) * | 2011-10-20 | 2013-04-24 | 电子科技大学 | Cloud computing virtual machine live migration technology |
| CN103677993A (en) * | 2012-08-31 | 2014-03-26 | 鸿富锦精密工业（深圳）有限公司 | Virtual machine resource load balancing system and method |
| CN103729280A (en) * | 2013-12-23 | 2014-04-16 | 国云科技股份有限公司 | High availability mechanism for virtual machine |
| CN104113596A (en) * | 2014-07-15 | 2014-10-22 | 华侨大学 | Cloud monitoring system and method for private cloud |
| CN104253860A (en) * | 2014-09-11 | 2014-12-31 | 武汉噢易云计算有限公司 | Shared storage message queue-based implementation method for high availability of virtual machines |
| CN104301389A (en) * | 2014-09-19 | 2015-01-21 | 华侨大学 | Energy efficiency monitoring and managing method and system of cloud computing system |
| US20150113531A1 (en) * | 2013-10-18 | 2015-04-23 | Power-All Networks Limited | System for migrating virtual machine and method thereof |
| CN104660690A (en) * | 2015-02-06 | 2015-05-27 | 中国农业大学 | Cloud video service monitoring system |
| US20150278042A1 (en) * | 2014-03-28 | 2015-10-01 | Vmware, Inc. | Vm availability during migration and vm network failures in host computing systems |
| CN105095001A (en) * | 2014-05-08 | 2015-11-25 | 中国银联股份有限公司 | Virtual machine exception recovery method under distributed environment |
| US20160026493A1 (en) * | 2011-07-06 | 2016-01-28 | Microsoft Technology Licensing, Llc | Planned virtual machines |
| US9286104B1 (en) * | 2015-01-05 | 2016-03-15 | International Business Machines Corporation | Selecting virtual machines to be relocated based on memory volatility |
| US20160179635A1 (en) * | 2014-12-17 | 2016-06-23 | American Megatrends, Inc. | System and method for performing efficient failover and virtual machine (vm) migration in virtual desktop infrastructure (vdi) |
-
2016
- 2016-06-27 CN CN201610481831.4A patent/CN107544839B/en active Active
Patent Citations (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102170474A (en) * | 2011-04-22 | 2011-08-31 | 广州杰赛科技股份有限公司 | Method and system for dynamic scheduling of virtual resources in cloud computing network |
| US20160026493A1 (en) * | 2011-07-06 | 2016-01-28 | Microsoft Technology Licensing, Llc | Planned virtual machines |
| CN103064733A (en) * | 2011-10-20 | 2013-04-24 | 电子科技大学 | Cloud computing virtual machine live migration technology |
| CN102708000A (en) * | 2012-04-19 | 2012-10-03 | 北京华胜天成科技股份有限公司 | System and method for realizing energy consumption control through virtual machine migration |
| CN102819465A (en) * | 2012-06-29 | 2012-12-12 | 华中科技大学 | Failure recovery method in virtualization environment |
| CN103677993A (en) * | 2012-08-31 | 2014-03-26 | 鸿富锦精密工业（深圳）有限公司 | Virtual machine resource load balancing system and method |
| US20150113531A1 (en) * | 2013-10-18 | 2015-04-23 | Power-All Networks Limited | System for migrating virtual machine and method thereof |
| CN103729280A (en) * | 2013-12-23 | 2014-04-16 | 国云科技股份有限公司 | High availability mechanism for virtual machine |
| US20150278042A1 (en) * | 2014-03-28 | 2015-10-01 | Vmware, Inc. | Vm availability during migration and vm network failures in host computing systems |
| CN105095001A (en) * | 2014-05-08 | 2015-11-25 | 中国银联股份有限公司 | Virtual machine exception recovery method under distributed environment |
| CN104113596A (en) * | 2014-07-15 | 2014-10-22 | 华侨大学 | Cloud monitoring system and method for private cloud |
| CN104253860A (en) * | 2014-09-11 | 2014-12-31 | 武汉噢易云计算有限公司 | Shared storage message queue-based implementation method for high availability of virtual machines |
| CN104301389A (en) * | 2014-09-19 | 2015-01-21 | 华侨大学 | Energy efficiency monitoring and managing method and system of cloud computing system |
| US20160179635A1 (en) * | 2014-12-17 | 2016-06-23 | American Megatrends, Inc. | System and method for performing efficient failover and virtual machine (vm) migration in virtual desktop infrastructure (vdi) |
| US9286104B1 (en) * | 2015-01-05 | 2016-03-15 | International Business Machines Corporation | Selecting virtual machines to be relocated based on memory volatility |
| CN104660690A (en) * | 2015-02-06 | 2015-05-27 | 中国农业大学 | Cloud video service monitoring system |
Non-Patent Citations (4)
| Title |
|---|
| KEJIANG YE ET AL.: "《VC-Migration: Live Migration of Virtual Clusters in the Cloud》", < 2012 ACM/IEEE 13TH INTERNATIONAL CONFERENCE ON GRID COMPUTING> * |
| 卢军: "《云计算关键技术研究》", 30 November 2015, 电子科技大学出版社 * |
| 张磊: "《存储系统的一种故障检测与服务迁移的研究》", 《中国优秀硕士学位论文全文数据库（电子期刊）信息科技辑》 * |
| 陈晶晶: "《云数据中心的能耗资源调度策略研究》", 《中国优秀硕士论文全文数据库（电子期刊）信息科技辑》 * |
Cited By (23)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108132829A (en) * | 2018-01-11 | 2018-06-08 | 郑州云海信息技术有限公司 | A kind of high available virtual machine realization method and system based on OpenStack |
| CN108334402A (en) * | 2018-03-07 | 2018-07-27 | 山东超越数控电子股份有限公司 | A kind of the virtual management system and its resource regulating method of non-stop layer framework |
| CN108762993A (en) * | 2018-06-06 | 2018-11-06 | 山东超越数控电子股份有限公司 | A kind of virtual-machine fail moving method and device based on artificial intelligence |
| CN110837451B (en) * | 2018-08-16 | 2023-08-15 | 中国移动通信集团重庆有限公司 | Processing method, device, equipment and medium for high availability of virtual machine |
| CN110837451A (en) * | 2018-08-16 | 2020-02-25 | 中国移动通信集团重庆有限公司 | Processing method, device, equipment and medium for high availability of virtual machine |
| CN109151045A (en) * | 2018-09-07 | 2019-01-04 | 北京邮电大学 | A kind of distribution cloud system and monitoring method |
| CN109151045B (en) * | 2018-09-07 | 2020-05-19 | 北京邮电大学 | Distributed cloud system and monitoring method |
| CN109818785A (en) * | 2019-01-15 | 2019-05-28 | 无锡华云数据技术服务有限公司 | A kind of data processing method, server cluster and storage medium |
| CN110445662A (en) * | 2019-08-29 | 2019-11-12 | 上海仪电（集团）有限公司中央研究院 | OpenStack control node is adaptively switched to the method and device of calculate node |
| CN110659109A (en) * | 2019-09-26 | 2020-01-07 | 上海仪电（集团）有限公司中央研究院 | Openstack cluster virtual machine monitoring system and method |
| CN110659109B (en) * | 2019-09-26 | 2023-07-04 | 上海仪电（集团）有限公司中央研究院 | System and method for monitoring openstack virtual machine |
| CN112073518A (en) * | 2020-09-09 | 2020-12-11 | 杭州海康威视系统技术有限公司 | Cloud storage system, cloud storage system management method and central management node |
| CN112073518B (en) * | 2020-09-09 | 2023-06-02 | 杭州海康威视系统技术有限公司 | Cloud storage system, cloud storage system management method and central management node |
| CN114760313A (en) * | 2020-12-29 | 2022-07-15 | 中国联合网络通信集团有限公司 | Service scheduling method and service scheduling device |
| CN114760313B (en) * | 2020-12-29 | 2023-11-24 | 中国联合网络通信集团有限公司 | Service scheduling method and service scheduling device |
| CN112994977A (en) * | 2021-02-24 | 2021-06-18 | 紫光云技术有限公司 | Method for high availability of server host |
| CN113407301A (en) * | 2021-05-22 | 2021-09-17 | 济南浪潮数据技术有限公司 | Virtual machine monitoring method, system, storage medium and equipment |
| CN114090184A (en) * | 2021-11-26 | 2022-02-25 | 中国电信集团系统集成有限责任公司 | Method and equipment for realizing high availability of virtualization cluster |
| WO2023092772A1 (en) * | 2021-11-26 | 2023-06-01 | 中电信数智科技有限公司 | Method and device for implementing high availability of virtualized cluster |
| CN114064217A (en) * | 2021-11-29 | 2022-02-18 | 建信金融科技有限责任公司 | Node virtual machine migration method and device based on OpenStack |
| CN114064217B (en) * | 2021-11-29 | 2024-04-19 | 建信金融科技有限责任公司 | OpenStack-based node virtual machine migration method and device |
| CN114217905A (en) * | 2021-12-17 | 2022-03-22 | 北京志凌海纳科技有限公司 | High-availability recovery processing method and system for virtual machine |
| CN115766405A (en) * | 2023-01-09 | 2023-03-07 | 苏州浪潮智能科技有限公司 | Fault processing method, device, equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107544839B (en) | 2021-05-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107544839A (en) | Virtual machine (vm) migration system, method and device | |
| CN105187249B (en) | A kind of fault recovery method and device | |
| CN107786616A (en) | Main frame intelligent monitor system based on high in the clouds | |
| US7797572B2 (en) | Computer system management method, management server, computer system, and program | |
| CN110389838A (en) | A kind of Real-Time Scheduling suitable for virtual resource and online migration management-control method | |
| US20140165054A1 (en) | Method and system for analyzing root causes of relating performance issues among virtual machines to physical machines | |
| US20140215077A1 (en) | Methods and systems for detecting, locating and remediating a congested resource or flow in a virtual infrastructure | |
| CN110493080A (en) | A kind of block chain node monitoring method, device and electronic equipment and storage medium | |
| CN108039964A (en) | Fault handling method and device, system based on network function virtualization | |
| CN110177020A (en) | A kind of High-Performance Computing Cluster management method based on Slurm | |
| CN111200526B (en) | Monitoring system and method of network equipment | |
| WO2016058318A1 (en) | Elastic virtual machine (vm) resource scaling method, apparatus and system | |
| CN106027328A (en) | Cluster monitoring method and system based on application container deployment | |
| CN107870832A (en) | Multipath storage device based on various dimensions Gernral Check-up method | |
| CN110515702A (en) | A kind of automatic evacuation method and device of calculate node fault virtual machine | |
| CN103973815A (en) | Method for unified monitoring of storage environment across data centers | |
| CN105893113A (en) | Management system and management method of virtual machine | |
| CN107947998A (en) | A kind of real-time monitoring system based on application system | |
| CN105516293A (en) | Cloud resource monitoring system of intelligent substation | |
| CN109271256A (en) | A kind of cloud resource management and monitoring system and method based on distributed deployment | |
| CN114154035A (en) | Data processing system for dynamic loop monitoring | |
| CN103414739B (en) | Use Cloud Server automatic monitored control system and the method for automatic drift | |
| CN107332707A (en) | A kind of acquisition method and device of SDN measurement data | |
| CN106982244A (en) | The method and apparatus that the message mirror of dynamic flow is realized under cloud network environment | |
| KR20220166760A (en) | Apparatus and method for managing trouble using big data of 5G distributed cloud system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20230918 Address after: 518000 Tencent Building, No. 1 High-tech Zone, Nanshan District, Shenzhen City, Guangdong Province, 35 Floors Patentee after: TENCENT TECHNOLOGY (SHENZHEN) Co.,Ltd. Patentee after: TENCENT CLOUD COMPUTING (BEIJING) Co.,Ltd. Address before: 2, 518000, East 403 room, SEG science and Technology Park, Zhenxing Road, Shenzhen, Guangdong, Futian District Patentee before: TENCENT TECHNOLOGY (SHENZHEN) Co.,Ltd. |