CN106227582B - Elastic telescopic method and system - Google Patents

Elastic telescopic method and system Download PDF

Info

Publication number
CN106227582B
CN106227582B CN201610657659.3A CN201610657659A CN106227582B CN 106227582 B CN106227582 B CN 106227582B CN 201610657659 A CN201610657659 A CN 201610657659A CN 106227582 B CN106227582 B CN 106227582B
Authority
CN
China
Prior art keywords
virtual machine
layers
management system
flexible
resource management
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.)
Active
Application number
CN201610657659.3A
Other languages
Chinese (zh)
Other versions
CN106227582A (en
Inventor
吴学启
徐荣华
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Huawei Technologies Co Ltd
Original Assignee
Huawei Technologies Co Ltd
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 Huawei Technologies Co Ltd filed Critical Huawei Technologies Co Ltd
Priority to CN201610657659.3A priority Critical patent/CN106227582B/en
Publication of CN106227582A publication Critical patent/CN106227582A/en
Application granted granted Critical
Publication of CN106227582B publication Critical patent/CN106227582B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/44—Arrangements for executing specific programs
    • G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
    • G06F9/45533—Hypervisors; Virtual machine monitors
    • G06F9/45558—Hypervisor-specific management and integration aspects
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/01—Protocols
    • H04L67/10—Protocols in which an application is distributed across nodes in the network
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/44—Arrangements for executing specific programs
    • G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
    • G06F9/45533—Hypervisors; Virtual machine monitors
    • G06F9/45558—Hypervisor-specific management and integration aspects
    • G06F2009/45562—Creating, deleting, cloning virtual machine instances
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/44—Arrangements for executing specific programs
    • G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
    • G06F9/45533—Hypervisors; Virtual machine monitors
    • G06F9/45558—Hypervisor-specific management and integration aspects
    • G06F2009/45575—Starting, stopping, suspending or resuming virtual machine instances

Landscapes

  • Engineering & Computer Science (AREA)
  • Software Systems (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Debugging And Monitoring (AREA)

Abstract

The invention discloses a kind of elastic telescopic method and system, belong to the communications field.The described method includes: the monitoring data for the application that flexible pond is supported in acquisition；It wherein, include at least one flexible group in flexible pond, each flexible group of virtual machine being pre-created including at least one, and a kind of application is deployed on each virtual machine in advance；Flexible decision is carried out to each flexible group according to monitoring data；When decision dilatation target stretches the virtual machine in group, the first virtual machine is selected in the virtual machine in a dormant state included from the flexible group of target, and wake up the first virtual machine.The present invention solves the problems, such as that the elastic telescopic scheme that the prior art provides is unable to satisfy the demand of Quick extension；By pre-creating virtual machine and pre- application deployment, waking up virtual machine in a dormant state can be realized the Quick extension of application node, has reached the stretching speed for improving application node, has met the technical effect of the needs of Quick extension.

Description

Elastic telescopic method and system
Technical field
The present invention relates to the communications field, in particular to a kind of elastic telescopic method and system.
Background technique
As the cloudization of every profession and trade application is gradually goed deep into, elastic telescopic (Auto Scaling) characteristic is at application cloud One of key property.The elastic telescopic characteristic of application refers to that on-demand application resource is using the size according to service traffics to increase Or reduce the number of nodes of operation application.
The elastic telescopic scheme that the prior art provides is as follows: when initial deployment is completed, not having on the host in flexible pond There is any virtual machine (Virtual Machine, VM) of operation.At runtime, when application service traffics rise when, cloud platform from Host is selected in flexible pond, and creates virtual machine on the host, then disposes and matches on the virtual machine that creation is completed Set application.In addition, it is necessary to application is transformed, so that applying can load and the configuration data that comes into force in stretching, extension automatically, So that it is guaranteed that using can normally provide business.For the application complicated for configuration data, using the larger transformation ability of needs Realization loads automatically in stretching, extension and the configuration data that comes into force.
For the application of certain industries, such as the application of telecommunications, finance, enterprise, have to the speed of elastic telescopic Higher requirement.Application node needs quick installation and deployment in stretching, extension, if Extensional periods are too long, can miss Business Stream Time to peak window is measured, the purpose used as needed is not achieved.For example, the telephone traffic of sms center can increase within 5 minutes 3 to 5 times, if the time of one application node of stretching, extension is more than 3 minutes, service traffics time to peak window is missed substantially.
However, the elastic telescopic scheme that the prior art provides, is unable to satisfy the demand of Quick extension.
Summary of the invention
In order to solve the problems, such as that the elastic telescopic scheme of prior art offer is unable to satisfy the demand of Quick extension, the present invention Embodiment provides a kind of elastic telescopic method and system.
On the one hand, the embodiment of the invention provides a kind of elastic telescopic methods, this method comprises: acquisition is to flexible Chi Suozhi The monitoring data for the application held；It wherein, include at least one flexible group in flexible pond, each flexible group is pre- including at least one The virtual machine first created, and a kind of application is deployed on each virtual machine in advance；According to monitoring data to each flexible group into The flexible decision of row；It is included in a dormant state from the flexible group of target when decision dilatation target stretches the virtual machine in group Virtual machine in select the first virtual machine, and wake up the first virtual machine.
Scheme provided in an embodiment of the present invention, by the way that virtual machine is pre-created in flexible pond, and in each virtual machine Upper preparatory application deployment stretch from target when decision dilatation target stretches the virtual machine in group and are in suspend mode included by group The first virtual machine is selected in the virtual machine of state, and wakes up first virtual machine；The elastic telescopic of prior art offer is provided Scheme, the problem of being unable to satisfy the demand of Quick extension；By pre-creating virtual machine and pre- application deployment, wakes up and be in suspend mode shape The Quick extension of application node can be realized in the virtual machine of state, has reached the stretching speed for improving application node, and satisfaction is quickly stretched The technical effect of the demand of exhibition.
Optionally, the first virtual machine is in a dormant state and to belong to the virtual machine of the flexible group of target on the first host. First host is that current idle resource is most in flexible pond, and exists in a dormant state and belong to the virtual of the flexible group of target The host of machine.
In a possible design, when decision dilatation target stretches the virtual machine in group, wrapped from the flexible group of target The first virtual machine is selected in the virtual machine in a dormant state included, and wakes up the first virtual machine, comprising: when decision dilatation target When virtual machine in flexible group, P layer resource management system are to I layer resource management systems transmission wake up instruction, wake up instruction use It is waken up in a dormant state in I layers of resource management system of instruction and belongs to the virtual machine of the flexible group of target；I layers of resource management system After receiving wake up instruction, the first virtual machine is selected from flexible pond, and wake up the first virtual machine.
In one example, I layers of resource management system select after receiving wake up instruction from the flexible pond One virtual machine, and wake up the first virtual machine, comprising: I layers of resource management system are after receiving wake up instruction, from flexible pond The virtual machine of the flexible group of target is selected in a dormant state and belonged to, and wakes up the virtual machine selected；I layers of resource management system After receiving the migration instruction that P layers of resource management system are sent, detect whether the virtual machine being waken up is located at the first host On machine；If the virtual machine being waken up is not located on the first host, the void that I layers of resource management system will be waken up Quasi- machine is migrated from original host to the first host.
By the above-mentioned means, being selected in a dormant state in the host most from current idle resource in stretching, extension Virtual machine is stretched, so that the load relative equilibrium of each host in flexible pond.
In a possible design, after carrying out flexible decision to each flexible group according to monitoring data, further includes: when Decision volume reduction target stretch the virtual machine in group when, select in the virtual machine in wake-up states included from the flexible group of target Second virtual machine, and the second virtual machine of suspend mode.Wherein, the second virtual machine is to be in wake-up states on the second host and belong to mesh The virtual machine of the flexible group of mark.Second host is that current idle resource is minimum in flexible pond, and exists and be in wake-up states and category In the host of the virtual machine of the flexible group of target.
By the above-mentioned means, selection is in wake-up states from current idle resource least host when shrinking Virtual machine is shunk, so that the load relative equilibrium of each host in flexible pond.
In one example, if I layers of resource management system are provided automatically from the least host of current idle resource The function of virtual machine suspend mode of the selection in wake-up states, then when decision volume reduction target stretches the virtual machine in group, from target The second virtual machine, and the second virtual machine of suspend mode are selected in the included virtual machine in wake-up states of flexible group, comprising: when certainly Plan volume reduction target stretch the virtual machine in group when, P layer resource management system refer to the first suspend mode of transmission of I layers of resource management system It enables, the first dormancy instruction is used to indicate I layers of resource management system suspend mode and is in wake-up states and belongs to the virtual of the flexible group of target Machine；I layers of resource management system select the second host after receiving the first dormancy instruction from flexible pond；I layers of resource pipe Reason system selects the second virtual machine, and the second virtual machine of suspend mode from the second host.
In another example, if I layers of resource management system provide connecing for the current idle resource for inquiring each host Mouthful, then when decision volume reduction target stretches the virtual machine in group, stretch from target virtual in wake-up states included by group The second virtual machine, and the second virtual machine of suspend mode are selected in machine, comprising: when decision volume reduction target stretches the virtual machine in group, P layers Resource management system obtains the current idle resource of each host in flexible pond from I layers of resource management system inquiry；P layers of resource Management system selects the second host according to the current idle resource of each host；P layers of resource management system are to I layers of resource Management system sends the second dormancy instruction, and the mark of the second host is carried in the second dormancy instruction, and the second dormancy instruction is used for Indicate the virtual machine for being in wake-up states on I layers of the second host of resource management system suspend mode and belonging to the flexible group of target；I layers of money Management system selects the second virtual machine after receiving the second dormancy instruction from the second host, and suspend mode second is empty Quasi- machine.
In another example, it if I layers of resource management system neither support the function in above-mentioned first example, does not also prop up The function in above-mentioned second example is held, then when decision volume reduction target stretches the virtual machine in group, is wrapped from the flexible group of target The second virtual machine, and the second virtual machine of suspend mode are selected in the virtual machine in wake-up states included, comprising: when decision volume reduction target When virtual machine in flexible group, P layer resource management system are from the virtual machine in wake-up states included by the flexible group of target Select the virtual machine started earliest as the second virtual machine；P layers of resource management system send third to I layers of resource management system and stop It sleeps and instructs, the mark of the second virtual machine is carried in third dormancy instruction, third dormancy instruction is used to indicate I layers of resource management system The second virtual machine of suspend mode；I layers of resource management system are after receiving third dormancy instruction, the second virtual machine of suspend mode.
The embodiment of the invention provides above-mentioned three kinds of modes, realize selection place from current idle resource least host In the virtual machine suspend mode of wake-up states.In practical applications, the function that can be supported according to I layers of resource management system, selection are appointed A kind of mode.
In a possible design, the monitoring data for the application supported flexible pond is acquired, comprising: P layers of resource pipe The operation data for the application that the flexible pond of reason system acquisition is supported, operation data are used to indicate the operation conditions of application；And/or P Layer resource management system receives the performance data for the application that the flexible pond that I layers of resource management system is sent is supported, performance data It is used to indicate the resource occupation situation of application.
By the above-mentioned means, operation data and performance data two of the P layers of resource management system in combination with the application got Aspect information carries out flexible decision, so that the result of decision is more accurate.
In a possible design, before the monitoring data for acquiring the application supported flexible pond, further includes: obtaining After getting virtual machine creating instruction, the virtual machine of destination number is pre-created in shared storage；According to the application section got Administration's request, disposes a kind of application, and be pre-configured application in advance on each virtual machine respectively；The temporary unused virtual machine of suspend mode.
By pre-creating virtual machine, pre- application deployment and pre-configuration application, eliminates and application is realized in stretching, extension certainly The requirement of dynamic load and the configuration data that comes into force, reduces the retrofit work amount of application.In addition, since each virtual machine creating is altogether It enjoys in storage, virtual machine occupied resource in suspend mode can discharge automatically, so that other application and virtual machine can be seized Its resource discharged has achieved the purpose that on demand, reasonably using resource.
On the other hand, the embodiment of the invention provides a kind of elastic telescopic system, which, which has, realizes that the above method shows The function of example.The function can also execute corresponding software realization by hardware realization by hardware.The hardware or Software includes one or more modules corresponding with above-mentioned function.
Another aspect includes processor, hair in the structure of control node the embodiment of the invention provides a kind of control node Emitter and receiver, the control node are configured as supporting to execute corresponding function in the above method.It the transmitter and connects Device is received to be used to support the communication between control node and other equipment.Further, control node can also include memory, institute Memory is stated for coupling with processor, saves the necessary program instruction of control node and data.
Another aspect, the embodiment of the invention provides a kind of computer storage mediums, above-mentioned for controlling for being stored as Computer software instructions used in node, it includes for executing program designed by above-mentioned aspect.
Compared to the prior art, in scheme provided in an embodiment of the present invention, by the way that virtual machine is pre-created in flexible pond, And the preparatory application deployment on each virtual machine are stretched group when decision dilatation target stretches the virtual machine in group from target The first virtual machine is selected in included virtual machine in a dormant state, and wakes up first virtual machine；Solves existing skill The elastic telescopic scheme that art provides, the problem of being unable to satisfy the demand of Quick extension；It is answered by pre-creating virtual machine and pre- deployment With waking up virtual machine in a dormant state can be realized the Quick extension of application node, has reached and has improved stretching for application node Contracting speed meets the technical effect of the needs of Quick extension.
Detailed description of the invention
To describe the technical solutions in the embodiments of the present invention more clearly, make required in being described below to embodiment Attached drawing is briefly described, it should be apparent that, drawings in the following description are only some embodiments of the invention, for For those of ordinary skill in the art, without creative efforts, it can also be obtained according to these attached drawings other Attached drawing.
Fig. 1 is a kind of schematic diagram of system architecture shown according to an exemplary embodiment；
Fig. 2 is the implementation diagram of technical solution provided by the invention；
Fig. 3 is a kind of flow chart of elastic telescopic method shown according to an exemplary embodiment；
Fig. 4 is a kind of flow chart of the elastic telescopic method shown according to another exemplary embodiment；
Fig. 5 is a kind of flow chart of the elastic telescopic method shown according to a further exemplary embodiment；
Fig. 6 is a kind of block diagram of elastic telescopic system shown according to an exemplary embodiment；
Fig. 7 is a kind of block diagram of the elastic telescopic system shown according to another exemplary embodiment；
Fig. 8 is a kind of block diagram of the elastic telescopic system shown according to another exemplary embodiment；
Fig. 9 is a kind of structural schematic diagram of control node shown according to an exemplary embodiment.
Specific embodiment
To make the object, technical solutions and advantages of the present invention clearer, below in conjunction with attached drawing to embodiment party of the present invention Formula is described in further detail.
There is cloud environment software to service (Software as a Service, SaaS), platform i.e. service (Platform As a Service, PaaS) and infrastructure device i.e. service (Infrastructure as a Service, IaaS) three service greatly Mode.PaaS and IaaS can directly pass through Services Oriented Achitecture (Service-Oriented Architecture, SOA) or network server is to platform user offer service, can also be used as the support platform of SaaS mode indirectly to end user Service.
Fig. 1 is a kind of schematic diagram of system architecture shown according to an exemplary embodiment.As shown in Figure 1, the system tray Structure includes: hardware layer 11, I layers of platform 12 and P layers of platform 13.
Physical equipment, including calculating, storage and the network equipment are disposed in hardware layer 11.
I layers of platform (IaaS Platform) 12, refer in cloud environment, are responsible for device resource virtualization, and carry out resource tune Spend the platform of management.Typical I layers of platform have OpenStack+kvm, vCloud+vCenter etc..The virtualization of I layers of platform 12 Management program (vitualization hypervisor) is for realizing the virtualization of equipment.
P layers of platform (PaaS Platform) 13, refer in cloud environment, are responsible for life cycle management (exploitation, portion of application Administration, monitoring, flexible etc.) platform.Typical P layers of platform have Google GAE, CloudFoundry etc..Wherein, apply and release Virtual machine is put, the application programming interface (Application for calling I layers of platform 12 to provide by P layers of platform 13 Programming Interface, API) request is initiated, and executed by I layers of platform 12.
It is operated on virtual machine created using (Application, App).
Fig. 2 shows the implementation diagrams of technical solution provided by the invention.
In the initial deployment stage, the virtual machine of destination number is pre-created on host by operation maintenance personnel, and completes to apply Pre- deployment and pre-configuration.For example, 4 virtual machines are pre-created in host Host1 and Host2 respectively, and respectively each A kind of application is disposed on a virtual machine in advance, as shown in Fig. 2, App1 is disposed in advance on virtual machine VM1, VM3, VM5 and VM6, in void Pre- deployment APP2 on quasi- machine VM2, VM4, VM7 and VM8.
After the completion of commissioning, the temporarily different virtual machine of suspend mode discharges the resource of host.
In the operation phase, by P layers of platform to application carry out KPI Key Performance Indicator (Key Performance Indicator, KPI) and condition monitoring, and whether decision needs to stretch or shrink, and initiates wake-ups or the request of suspend mode virtual machine by P layer platform Extension and contraction control is carried out to I layers of platform, it is preferred that I layers of platform execute resource.In embodiments of the present invention, it wakes up virtual machine and starts fortune Row virtual machine, suspend mode virtual machine, that is, virtual machine out of service.
Following is embodiment of the present invention method, and the executing subject of each step can be control node, which is used for The application node controlled in flexible pond carries out elastic telescopic.Wherein, application node refers in cloud environment, has run application program Virtual machine instance.Optionally, control node can be realized by an equipment, can also be realized by multiple devices reciprocal.Control Node processed can be deployed in the P layer platform in system shown in Figure 1 framework, can also be deployed in I layers of platform, or use distribution side Formula is deployed in P layers of platform and I layers of platform.Control node can be the equipment for having calculating and processing capacity, such as server.
Fig. 3 is a kind of flow chart of elastic telescopic method shown according to an exemplary embodiment.This method may include Following steps:
Step 301, the monitoring data for the application that flexible pond is supported in acquisition.It wherein, include at least one in flexible pond Flexible group, each flexible group includes the virtual machine that at least one is pre-created, and is deployed with one in advance on each virtual machine Kind application.
Step 302, flexible decision is carried out to each flexible group according to monitoring data.
Step 303, it when decision dilatation target stretches the virtual machine in group, stretches from target and is in suspend mode included by group The first virtual machine is selected in the virtual machine of state, and wakes up the first virtual machine.
In conclusion method provided in this embodiment, by the way that virtual machine is pre-created in flexible pond, and in each void Preparatory application deployment on quasi- machine are in when decision dilatation target stretches the virtual machine in group from the flexible group of target is included The first virtual machine is selected in the virtual machine of dormant state, and wakes up first virtual machine；The elasticity of prior art offer is provided Flexible scheme, the problem of being unable to satisfy the demand of Quick extension；By pre-creating virtual machine and pre- application deployment, wake up in not The Quick extension of application node can be realized in the virtual machine of dormancy state, has reached the stretching speed for improving application node, meets fast The technical effect of the demand of speed stretching, extension.
Fig. 4 is a kind of flow chart of the elastic telescopic method shown according to another exemplary embodiment.This method can wrap Include following steps:
Step 401, after getting virtual machine creating instruction, the virtual of destination number is pre-created in shared storage Machine.
Operation maintenance personnel initiates virtual machine creating instruction, such as virtual machine creating instruction is used to indicate control node creation mesh Mark the virtual machine of quantity.Control node creates target in I layers of platform after getting virtual machine creating instruction, according to the instruction The virtual machine of quantity.
In addition, virtual machine creating so as to subsequent in wake-up or suspend mode virtual machine, dynamically seized and is released in shared storage Put resource.
Step 402, according to the application deployment request got, a kind of application is disposed in advance on each virtual machine respectively, And it is pre-configured above-mentioned application.
Operation maintenance personnel initiates application deployment request, such as application deployment request for requesting control node specified virtual Specified application is disposed on machine.Control node completes the pre- deployment of application according to the application deployment request got.Operation maintenance personnel can Manually or using deployment tool application deployment on each virtual machine.
The pre-configuration of application includes IP (Internet Protocol, the internet protocol of the configuration and virtual machine using itself View) configuration etc..Using completing after being pre-configured, where virtual machine corresponding to can preserve in virtual machine file it is relevant Configuration data directly therefrom obtains configuration data when waking up virtual machine in order to subsequent.Operation maintenance personnel can be manually or using portion Administration's tool configures application on the virtual machine for be equipped with application.
In addition, the maximum that the quantity for the application node disposed in advance may need within a period of time in future according to corresponding service Quantity determines.For example, it is assumed that it is corresponding to estimate the App1 that may be needed in a period of time in future according to history service statistical result The maximum quantity of application node is that the maximum quantity of the corresponding application node of 4, App2 is 6, then 10 virtual machines is pre-created, at it In dispose App1 in advance on 4 virtual machines, and dispose App2 in advance on other 6 virtual machines.
Complete it is above-mentioned pre-create, dispose and be pre-configured in advance after, form flexible pond.It is stretched in flexible pond including at least one Contracting group, each flexible group includes the virtual machine that at least one is pre-created, and is deployed with one kind in advance on each virtual machine Using.Flexible group is the identical virtual machine set of one group of function.That is, being deployed with the Imaginary Mechanism of same application in flexible pond At one flexible group.Fig. 2 is referred to for example, combining, includes two flexible group in the initial deployment stage, pond of stretching, wherein first Flexible group is the virtual machine for being deployed with App1, including VM1, VM3, VM5 and VM6, and second stretches group to be deployed with the virtual of App2 Machine, including VM2, VM4, VM7 and VM8.
Step 403, the temporary unused virtual machine of suspend mode.
For control node according to current business demand, determining needs virtual machine to be used, and suspend mode is temporarily unused virtual Machine.After virtual machine suspend mode, the resource originally occupied can discharge automatically.
Step 404, the monitoring data for the application that flexible pond is supported in acquisition.
The monitoring data for the application that flexible pond is supported in control node acquisition.Optionally, monitoring data includes application The performance data of operation data and application.The operation data of application is used to indicate the operation conditions of application, such as the current shape of application State and KPI.The performance data of application is used to indicate the resource occupation situation of application, such as the occupancy of CPU, memory, network resource Situation.
Step 405, flexible decision is carried out to each flexible group according to monitoring data.
Control node carries out flexible decision to each flexible group respectively according to the monitoring data of collected various applications. For example, when the operation data instruction current business of target application is busy and/or the performance data of target application indicates resource occupation When rate is higher than the first threshold value, decision stretches the quantity of the application node corresponding to the target application in flexible group.In another example When the operation data instruction current business of target application is idle and/or the performance data instruction resources occupation rate of target application is lower than When the second threshold value, decision shrinks the quantity of the application node corresponding to the target application in flexible group.
In embodiments of the present invention, the specific strategy for carrying out flexible decision to each flexible group according to monitoring data is not made It limits, can be pre-configured with according to actual needs.In addition, in embodiments of the present invention, passing through the operation data of connected applications Flexible decision is carried out with two aspect information of performance data, so as to the real busy of aware application, so that the result of decision is more Accurately.
Step 406, it when decision dilatation target stretches the virtual machine in group, stretches from target and is in suspend mode included by group The first virtual machine is selected in the virtual machine of state, and wakes up the first virtual machine.
Fig. 2 is referred to for example, combining, is still the virtual machine for being deployed with App1 with the first flexible group, the second flexible group is portion There is the virtual machine instance of App2 in administration.Assuming that the virtual machine in current first flexible group in wake-up states is VM1 and VM5, such as Fig. 2 Shown, VM1 is located at Host1, and VM5 is located at Host2.When the virtual machine in group is stretched in decision dilatation first, in first flexible group The virtual machine for being currently at dormant state includes VM3 and VM6, and control node selects a virtual machine and called out from VM3 and VM6 It wakes up.
Optionally, the first virtual machine is in a dormant state and to belong to the virtual machine of the flexible group of target on the first host. First host is that current idle resource is most in flexible pond, and exists in a dormant state and belong to the virtual of the flexible group of target The host of machine.That is, selecting virtual machine in a dormant state in the host most from current idle resource in stretching, extension It wakes up.For example, as shown in Fig. 2, when the virtual machine in group is stretched in decision dilatation first, it is assumed that the current idle resource of Host1 More than Host2, then control node selection wakes up VM3.
Step 407, it when decision volume reduction target stretches the virtual machine in group, is waken up from included being in of the flexible group of target The second virtual machine, and the second virtual machine of suspend mode are selected in the virtual machine of state.
Fig. 2 is referred to for example, combining, is still the virtual machine for being deployed with App1 with the first flexible group, the second flexible group is portion There is the virtual machine instance of App2 in administration.Assuming that the virtual machine in current second flexible group in wake-up states is VM2 and VM7, such as Fig. 2 Shown, VM2 is located at Host1, and VM7 is located at Host2.When the virtual machine in group is stretched in decision volume reduction second, control node is from VM2 With one virtual machine of selection and suspend mode in VM7.
Optionally, the second virtual machine is the virtual machine for being in wake-up states on the second host and belonging to the flexible group of target. Second host is that current idle resource is minimum in flexible pond, and exists in wake-up states and belong to the virtual of the flexible group of target The host of machine.That is, selection is in the virtual machine of wake-up states from current idle resource least host when shrinking Suspend mode.For example, as shown in Fig. 2, when the virtual machine in group is stretched in decision volume reduction second, it is assumed that the current idle resource of Host1 More than Host2, then control node selects suspend mode VM7.
In addition, the current idle resource of host, can be combined with from any one or more in the resources such as memory, CPU Angle measured.
In conclusion method provided in this embodiment, by the way that virtual machine is pre-created in flexible pond, and in each void Preparatory application deployment on quasi- machine are in when decision dilatation target stretches the virtual machine in group from the flexible group of target is included The first virtual machine is selected in the virtual machine of dormant state, and wakes up first virtual machine；The elasticity of prior art offer is provided Flexible scheme, the problem of being unable to satisfy the demand of Quick extension；By pre-creating virtual machine and pre- application deployment, wake up in not The Quick extension of application node can be realized in the virtual machine of dormancy state, has reached the stretching speed for improving application node, meets fast The technical effect of the demand of speed stretching, extension.
In addition, selecting virtual machine in a dormant state to call out in the host most from current idle resource in stretching, extension It wakes up；When shrinking, selection is in the virtual machine suspend mode of wake-up states from current idle resource least host, so that flexible The load relative equilibrium of each host in pond.
In addition, being applied by pre-creating virtual machine, pre- application deployment and being pre-configured, eliminates and stretched to using realization The requirement of Shi Zidong load and the configuration data that comes into force, reduces the retrofit work amount of application.
In addition, occupied resource can be automatic in suspend mode for virtual machine since each virtual machine creating is in shared storage Release has reached the mesh on demand, reasonably utilizing resource so that other application and virtual machine can seize the resource of its release 's.
Fig. 5 is a kind of flow chart of the elastic telescopic method shown according to a further exemplary embodiment.In the present embodiment, I layers of resource management system are equivalent to the I layer platform in system shown in Figure 2 framework, and P layers of resource management system are equivalent to shown in Fig. 2 P layer platform in system architecture.This method may include steps of:
Step 501, after getting virtual machine creating instruction, I layers of resource management system pre-create in shared storage The virtual machine of destination number.
Step 502, I layers of resource management system are according to the application deployment request got, respectively on each virtual machine A kind of application is disposed in advance, and is pre-configured above-mentioned application.
Step 503, the temporary unused virtual machine of I layers of resource management system suspend mode.
In the initial deployment stage, operation maintenance personnel creates virtual machine in I layers of resource management system in advance, and completes the pre- of application Deployment and pre-configuration.Later, after all virtual machines of operation maintenance personnel commissioning and application are available, I layers of resource management system suspend mode Temporary unused virtual machine.
Step 504, the monitoring data for the application that flexible pond is supported in P layers of resource management system acquisition.
In the present embodiment, it is located at P layers of resource management system for the control node for decision of stretching.P layers of resource management system System can be used the following two kinds mode and acquire monitoring data:
First, P layers of resource management system acquire the operation data for the application that flexible pond is supported；
The operation data of application is used to indicate the operation conditions of application, the current state and KPI of such as application.The operation of application Data are actively acquired by P layers of resource management system.
Second, P layers of resource management system receive the performance for the application that the flexible pond that I layers of resource management system is sent is supported Data.
The performance data of application is used to indicate the resource occupation situation of application, such as the occupancy of CPU, memory, network resource Situation.The performance data of application is acquired by I layers of resource management system, and is reported to P layers of resource management system.
Step 505, P layers of resource management system carry out flexible decision to each flexible group according to monitoring data.
P layers of resource management system are stretched in combination with the operation data and two aspect information of performance data of the application got Contracting decision, so that the result of decision is more accurate.
Step 506, when decision dilatation target stretches the virtual machine in group, P layers of resource management system are to I layers of resource management System sends wake up instruction.
Wake up instruction is used to indicate I layers of resource management system and wakes up in a dormant state and belong to the virtual of the flexible group of target Machine.For example, the mark of application corresponding to the mark or the flexible group of target of the flexible group of target can be carried in wake up instruction, accused with this Know that I layers of resource management system need which flexible group to carry out dilatation to.
Correspondingly, I layers of resource management system receive the wake up instruction that P layers of resource management system is sent.I layers of resource management system System is after receiving wake up instruction, and selection first is virtual in the virtual machine in a dormant state included from the flexible group of target Machine, and wake up the first virtual machine.
In one example, it in such a way that following steps 507 to step 509 provide, realizes from current idle resource most Virtual machine in a dormant state is selected to wake up in more hosts.
Step 507, I layers of resource management system are after receiving wake up instruction, and selection is in suspend mode shape from flexible pond State and the virtual machine for belonging to the flexible group of target, and wake up the virtual machine selected.
Fig. 2 is referred to for example, combining, is still the virtual machine for being deployed with App1 with the first flexible group, the second flexible group is portion There is the virtual machine instance of App2 in administration.Assuming that the virtual machine in current first flexible group in wake-up states is VM1 and VM5, such as Fig. 2 Shown, VM1 is located at Host1, and VM5 is located at Host2.When the virtual machine that P layers of resource management system decision dilatation first are stretched in group When, P layers of resource management system send wake up instruction to I layers of resource management system, and I layers of resource management system receive wake up instruction Later, obtain and be currently at the virtual machine (including VM3 and VM6) of dormant state in first flexible group, I layers of resource management system from Middle any one virtual machine of selection wakes up.For example, I layers of resource management system selection wake up VM3.
Step 508, I layers of resource management system are after receiving the migration instruction that P layers of resource management system are sent, detection Whether the virtual machine being waken up is located on the first host.
P layers of resource management system are after issuing wake up instruction to I layers of resource management system, also to I layers of resource management system System sends migration instruction.Migration instruction is used to indicate I layers of resource management system for the virtual machine (vm) migration being waken up to the first host Machine；Wherein, the first host is that current idle resource is most in flexible pond, and exists in a dormant state and belong to target and stretch The host of the virtual machine of group.Whether I layers of resource management system detect the virtual machine that is waken up after receiving migration instruction On the first host.For example, as illustrated in fig. 2, it is assumed that Host2 is that current idle resource is most, and exists and be in suspend mode State and the host for belonging to the first flexible virtual machine organized.Then I layers of resource management system are after receiving migration instruction, inspection Survey whether the virtual machine being waken up in step 507 is located at Host2.
Step 509, if the virtual machine being waken up is not located on the first host, I layers of resource management system will be called out Awake virtual machine is migrated from original host to the first host.
For example, I layers of resource management system wake up VM3 in step 507, VM3 is located at Host1 rather than is located at Host2, then I Layer resource management system migrates VM3 to Host2 from Host1, so that the load of each host in flexible pond is relatively equal Weighing apparatus.
In addition, I layers of resource management system do not execute virtual machine if the virtual machine being waken up is located on the first host Migration operation.
When P layers of resource management system decision volume reduction target stretch the virtual machine in group, any one following side can be used Formula realizes virtual machine suspend mode of the selection in wake-up states from current idle resource least host.
In the first possible embodiment, if I layers of resource management system provide automatically from current idle resource most The function of virtual machine suspend mode of the selection in wake-up states in few host, then include the following steps 510 to step 512:
Step 510, when decision volume reduction target stretches the virtual machine in group, P layers of resource management system are to I layers of resource management System sends the first dormancy instruction.
First dormancy instruction is used to indicate I layers of resource management system suspend mode and is in wake-up states and belongs to the flexible group of target Virtual machine.For example, the mark of application corresponding to the mark or the flexible group of target of the flexible group of target can be carried in the first dormancy instruction Know, informs that I layers of resource management system need which flexible group to carry out volume reduction to this.
Correspondingly, I layers of resource management system receive the first dormancy instruction that P layers of resource management system is sent.
Step 511, I layers of resource management system select the second place after receiving the first dormancy instruction from flexible pond Host.
Second host is that current idle resource is minimum in flexible pond, and exists to be in wake-up states and belong to target and stretch The host of the virtual machine of group.
Fig. 2 is referred to for example, combining, is still the virtual machine for being deployed with App1 with the first flexible group, the second flexible group is portion There is the virtual machine instance of App2 in administration.Assuming that the virtual machine in current second flexible group in wake-up states is VM2 and VM7, such as Fig. 2 Shown, VM2 is located at Host1, and VM7 is located at Host2.When the virtual machine in group is stretched in decision volume reduction second, P layers of resource management system Unite and send the first dormancy instructions to I layers of resource management system, I layers of resource management system after receiving the first dormancy instruction, Obtain the current idle resource of Host1 and Host2.Assuming that the current idle resource of Host1 is more than Host2, then I layers of resource management Systematic selection Host2.
Step 512, I layers of resource management system select the second virtual machine, and the second virtual machine of suspend mode from the second host.
Second virtual machine is the virtual machine for being in wake-up states on the second host and belonging to the flexible group of target.
In second of possible embodiment, if I layers of resource management system, which provide, is inquired the current of each host The interface of idling-resource then includes the following steps 513 to step 516:
Step 513, when decision volume reduction target stretches the virtual machine in group, P layers of resource management system are from I layers of resource management System queries obtain the current idle resource of each host in flexible pond.
For example, when P layers of resource management system decision volume reduction target stretch the virtual machine in group, P layers of resource management system The above-mentioned interface for calling I layers of resource management system to provide sends inquiry request to I layers of resource management system, which uses It stretches the current idle resource of each host in pond in request.Correspondingly, I layers of resource management system receive inquiry and ask After asking, the current idle resource of each host in flexible pond is sent to P layers of resource management system.
Step 514, P layers of resource management system select the second host according to the current idle resource of each host.
Second host is that current idle resource is minimum in flexible pond, and exists to be in wake-up states and belong to target and stretch The host of the virtual machine of group.
Step 515, P layers of resource management system send the second dormancy instruction to I layers of resource management system.
The mark of the second host is carried in second dormancy instruction, the second dormancy instruction is used to indicate I layers of resource management system Wake-up states are on system the second host of suspend mode and belong to the virtual machine of the flexible group of target.Optionally, in the second dormancy instruction The mark that application corresponding to the mark or the flexible group of target of the flexible group of target can also be carried, informs I layers of resource management system with this System needs which flexible group to carry out volume reduction to.Alternatively, can also be by the mark of the second virtual machine of selection suspend mode in the second dormancy instruction Know, informs I layers of resource management system need which virtual machine of suspend mode with this.Wherein, the second virtual machine is on the second host In wake-up states and belong to the flexible virtual machine organized of target.
Step 516, I layers of resource management system select after receiving the second dormancy instruction from the second host Two virtual machines, and the second virtual machine of suspend mode.
In the third possible embodiment, if I layers of resource management system neither support that above-mentioned the first is possible Function in embodiment does not support the function in above-mentioned second possible embodiment yet, then include the following steps 517 to Step 519:
Step 517, when decision volume reduction target stretches the virtual machine in group, P layers of resource management system stretch group from target Select the virtual machine started earliest as the second virtual machine in the included virtual machine in wake-up states.
Step 518, P layers of resource management system send third dormancy instruction to I layers of resource management system.
The mark of the second virtual machine is carried in third dormancy instruction, third dormancy instruction is used to indicate I layers of resource management system System the second virtual machine of suspend mode.
Step 519, I layers of resource management system are after receiving third dormancy instruction, the second virtual machine of suspend mode.
It is the least host of current idle resource in view of the host where the virtual machine that starts earliest has very maximum probability Machine, therefore select the virtual machine started earliest as the second virtual machine suspend mode, it can largely meet preferred demand.
In conclusion method provided in this embodiment, by the way that virtual machine is pre-created in flexible pond, and in each void Preparatory application deployment on quasi- machine are in when decision dilatation target stretches the virtual machine in group from the flexible group of target is included The first virtual machine is selected in the virtual machine of dormant state, and wakes up first virtual machine；The elasticity of prior art offer is provided Flexible scheme, the problem of being unable to satisfy the demand of Quick extension；By pre-creating virtual machine and pre- application deployment, wake up in not The Quick extension of application node can be realized in the virtual machine of dormancy state, has reached the stretching speed for improving application node, meets fast The technical effect of the demand of speed stretching, extension.
In addition, stretching, extension when, using thermophoresis mode by virtual machine (vm) migration to the most host of current idle resource it On, so that the load relative equilibrium of each host in flexible pond.
In addition, the present embodiment additionally provides three kinds of modes, selection place from current idle resource least host is realized In the virtual machine suspend mode of wake-up states.In practical applications, the function that can be supported according to I layers of resource management system, selection are appointed A kind of mode.
In a typical application scenarios, for traditional electrical business Enabler SMS service center (Short Message Service Center, SMSC) for, since short message service flow may increase by 2 to 10 times in several minutes, because This external demand is: Carrier Requirements SMSC can stretch out new application node in 3 minutes.And it is scalable due to SMSC Several configuration parameters of module are all privately owned configurations, i.e., related to module instance, NE ID, primary charging section such as this node Point etc., internal demands first is that: SMSC require without transformation automatically configure load can be achieved with business stretching, extension when automatic wire charging； It is larger additionally, due to the virtual machine image template of SMSC, be 100G or so, therefore when creating virtual machine time-consuming at 20 to 30 points Clock can not stretch out application node, therefore internal demands if just creating virtual machine in run mode stretching, extension in 3 minutes Second is that: SMSC requires not create virtual machine in stretching, extension, but starts the virtual machine for realizing that deployment commissioning finishes.
Technical solution provided by the invention can meet the demand well.Operation maintenance personnel is as requested in realization business Virtual machine before line pre-creates, business is disposed and is pre-configured in advance；P layers of platform realize business monitoring and flexible decision, and call I Stop virtual machine when starting virtual machine, business are shunk when layer platform realizes business stretching, extension；I layers of platform realize that resources are preferred, when connecing After the request for receiving P layers of platform starting virtual machine, idling-resource (CPU and/or memory) most host is found in flexible pond Start virtual machine on machine, (the mark of the virtual machine wherein with stopping in need after receiving P layers of platform and stopping the request of virtual machine Know), stop corresponding virtual machine.
It is following that this hair is please referred to for undisclosed details in present system embodiment for present system embodiment Bright embodiment of the method.
Fig. 6 is a kind of block diagram of elastic telescopic system shown according to an exemplary embodiment, which may include: to adopt Collect module 610, decision-making module 620 and wake-up module 630.
Acquisition module 610, for acquiring the monitoring data for the application supported flexible pond.Wherein, in the flexible pond Including at least one group of stretching, each flexible group includes the virtual machine that at least one is pre-created, and on each virtual machine It is deployed with a kind of application in advance.
Decision-making module 620, for carrying out flexible decision to each described flexible group according to the monitoring data.
Wake-up module 630, for being wrapped from the flexible group of the target when decision dilatation target stretches the virtual machine in group The first virtual machine is selected in the virtual machine in a dormant state included, and wakes up first virtual machine.
In conclusion system provided in this embodiment, by the way that virtual machine is pre-created in flexible pond, and in each void Preparatory application deployment on quasi- machine are in when decision dilatation target stretches the virtual machine in group from the flexible group of target is included The first virtual machine is selected in the virtual machine of dormant state, and wakes up first virtual machine；The elasticity of prior art offer is provided Flexible scheme, the problem of being unable to satisfy the demand of Quick extension；By pre-creating virtual machine and pre- application deployment, wake up in not The Quick extension of application node can be realized in the virtual machine of dormancy state, has reached the stretching speed for improving application node, meets fast The technical effect of the demand of speed stretching, extension.
In one example, first virtual machine is in a dormant state and to belong to the target on the first host and stretch The virtual machine of contracting group.First host is that current idle resource is most in the flexible pond, and exists in a dormant state And belong to the host of the virtual machine of the flexible group of the target.
In the alternative embodiment provided based on embodiment illustrated in fig. 6, as shown in fig. 7, the system also includes: stop Dormancy module 640.
The sleep block 640, for the target described in the decision volume reduction stretch the virtual machine in group when, stretched from the target The second virtual machine, and the second virtual machine described in suspend mode are selected in virtual machine included by contracting group in wake-up states.
Wherein, second virtual machine is the void for being in wake-up states on the second host and belonging to the flexible group of the target Quasi- machine；Second host is that current idle resource is minimum in the flexible pond, and exists in wake-up states and belong to institute State the host of the virtual machine of the flexible group of target.
In another alternative embodiment provided based on embodiment illustrated in fig. 6, as shown in fig. 7, the system also includes: it is pre- Processing module 650.
The preprocessing module 650, is specifically used for: upper pre- in shared storage after getting virtual machine creating instruction Create the virtual machine of destination number；According to the application deployment request got, one kind is disposed in advance on each virtual machine respectively Using, and it is pre-configured the application；The temporary unused virtual machine of suspend mode.
Fig. 8 is a kind of block diagram of the elastic telescopic system shown according to another exemplary embodiment, which includes: P layers Resource management system 81 and I layers of resource management system 82.
P layers of resource management system 81 include: acquisition module 811, decision-making module 812 and wake up instruction sending module 813.
Acquisition module 811, for acquiring the monitoring data for the application supported flexible pond.Wherein, in the flexible pond Including at least one group of stretching, each flexible group includes the virtual machine that at least one is pre-created, and on each virtual machine It is deployed with a kind of application in advance.
Decision-making module 812, for carrying out flexible decision to each described flexible group according to the monitoring data.
Wake up instruction sending module 813, for when decision dilatation target stretches the virtual machine in group, I layers of Xiang Suoshu to be provided Management system 82 sends wake up instruction, and the wake up instruction is used to indicate the I layers of resource management system 82 and wakes up in not Dormancy state and the virtual machine for belonging to the flexible group of the target.
I layers of resource management system 82 include: wake up instruction receiving module 821 and virtual machine wake-up module 822.
Wake up instruction receiving module 821, for receiving the wake up instruction.
Virtual machine wake-up module 822, for selecting the first virtual machine from the flexible pond, and it is virtual to wake up described first Machine.
Optionally, first virtual machine is in a dormant state and to belong to the flexible group of the target on the first host Virtual machine.First host is that current idle resource is most in the flexible pond, and exists in a dormant state and belong to The host of the virtual machine of the flexible group of the target.
In one example, the virtual machine wake-up module 822, is specifically used for: selection is in not from the flexible pond Dormancy state and the virtual machine for belonging to the flexible group of the target, and wake up the virtual machine selected；It is connect in the wake up instruction After receiving the migration instruction that module 821 receives the transmission of P layers of resource management system 81, the virtual machine being waken up is detected Whether it is located on first host；It, will if the virtual machine being waken up is not located on first host The virtual machine being waken up is migrated from original host to first host.The P layers of resource management system 81 is called out Instruction sending module 813 of waking up is also used to send the migration instruction to the I layers of resource management system 82.
In a kind of possible embodiment, as shown in figure 8, P layers of resource management system 81 further include: dormancy instruction is sent Module 814.I layers of resource management system 82 further include: dormancy instruction receiving module 823 and virtual machine sleep block 824.
Dormancy instruction sending module 814, for the target described in the decision volume reduction stretch the virtual machine in group when, Xiang Suoshu I Layer resource management system 82 sends the first dormancy instruction, and first dormancy instruction is used to indicate the I layers of resource management system Suspend mode is in wake-up states and belongs to the virtual machine of the flexible group of the target.
Dormancy instruction receiving module 823, for receiving first dormancy instruction.
Virtual machine sleep block 824, for selecting second host from the flexible pond, from second host The second virtual machine, and the second virtual machine described in suspend mode are selected on machine.
Optionally, second virtual machine is to be in wake-up states on the second host and belong to the flexible group of the target Virtual machine.Second host is that current idle resource is minimum in the flexible pond, and exists in wake-up states and belong to The host of the virtual machine of the flexible group of the target.
In another possible embodiment, P layers of resource management system 81 further include: enquiry module (not shown), Selecting module (not shown) and dormancy instruction sending module.I layers of resource management system 82 further include: dormancy instruction receives mould Block and virtual machine sleep block.
Enquiry module, for the target described in the decision volume reduction stretch the virtual machine in group when, from the I layers of resource management system 82 inquiry of system obtains the current idle resource of each host in the flexible pond.
Selecting module selects second host for the current idle resource according to each host.It is described Second host is that current idle resource is minimum in the flexible pond, and exists to be in wake-up states and belong to the target and stretch The host of the virtual machine of group.
Dormancy instruction sending module, for the I layer resource management system 82 transmission the second dormancy instruction, described second The mark of second host is carried in dormancy instruction, second dormancy instruction is used to indicate the I layers of resource management system Wake-up states are on second host described in system suspend mode and belong to the virtual machine of the flexible group of the target.
Dormancy instruction receiving module, for receiving second dormancy instruction.
Virtual machine sleep block, for selecting the second virtual machine from second host, and the second void described in suspend mode Quasi- machine.Second virtual machine is the virtual machine for being in wake-up states on the second host and belonging to the flexible group of the target.
In another possible embodiment, P layers of resource management system 81 further include: selecting module (not shown) and Dormancy instruction sending module.I layers of resource management system 82 further include: dormancy instruction receiving module and virtual machine sleep block.
Selecting module, for the target described in the decision volume reduction stretch the virtual machine in group when, from the flexible group institute of the target Including the virtual machine in wake-up states in select the virtual machine started earliest as the second virtual machine.
Dormancy instruction sending module, for sending third dormancy instruction, the third to the I layers of resource management system 82 The mark of second virtual machine is carried in dormancy instruction, the third dormancy instruction is used to indicate the I layers of resource management system Second virtual machine described in system suspend mode.
Dormancy instruction receiving module, for receiving the third dormancy instruction.
Virtual machine sleep block, for the second virtual machine described in suspend mode.
In one example, the acquisition module 811, is specifically used for: the fortune for the application that the acquisition flexible pond is supported Row data, the operation data are used to indicate the operation conditions of application；And/or receive the institute that I layers of resource management system 82 is sent The performance data for the application that flexible pond is supported is stated, the performance data is used to indicate the resource occupation situation of application.
Optionally, as shown in figure 8, the I layers of resource management system 82 further include: preprocessing module 825.The pretreatment Module 825, is specifically used for: after getting virtual machine creating instruction, pre-creating the virtual of destination number in shared storage Machine；According to the application deployment request got, a kind of application is disposed in advance on each virtual machine respectively, and answer described in pre-configuration With；The temporary unused virtual machine of suspend mode.
In conclusion system provided in this embodiment, by the way that virtual machine is pre-created in flexible pond, and in each void Preparatory application deployment on quasi- machine are in when decision dilatation target stretches the virtual machine in group from the flexible group of target is included The first virtual machine is selected in the virtual machine of dormant state, and wakes up first virtual machine；The elasticity of prior art offer is provided Flexible scheme, the problem of being unable to satisfy the demand of Quick extension；By pre-creating virtual machine and pre- application deployment, wake up in not The Quick extension of application node can be realized in the virtual machine of dormancy state, has reached the stretching speed for improving application node, meets fast The technical effect of the demand of speed stretching, extension.
It should be noted that system provided by the above embodiment is when realizing its function, only with above-mentioned each functional module It divides and carries out for example, can according to need in practical application and be completed by different functional modules above-mentioned function distribution, The internal structure of equipment is divided into different functional modules, to complete all or part of the functions described above.In addition, System and method for embodiment provided by the above embodiment belongs to same design, and specific implementation process is detailed in embodiment of the method, this In repeat no more.
It is above-mentioned that mainly scheme provided in an embodiment of the present invention is described from the angle of control node.It is understood that It is that in order to realize the above functions, it comprises execute the corresponding hardware configuration of each function and/or software module for control node. In conjunction with disclosed in this invention embodiment describe each exemplary module and algorithm steps, the embodiment of the present invention can be with hardware Or the combining form of hardware and computer software is realized.Some function drives hardware actually with hardware or computer software Mode executes, specific application and design constraint depending on technical solution.Those skilled in the art can be to each spy Described function is realized in fixed application using different methods, but this is realized it is not considered that implementing beyond the present invention The range of the technical solution of example.
Fig. 9 is a kind of structural schematic diagram of control node shown according to an exemplary embodiment.For example, the control node It can be server, for realizing the exemplary function of the above method.Control node 900 may include: emitter/receiver 901 With processor 902.
Emitter/receiver 901 is for supporting to receive and send messages between control node 900 and external equipment.Processor 902 is used In the various functions for realizing control node 900, for example flexible decision, resource are preferably etc..The processor 902 is also used in execution State other steps of each step or described in the invention technical solution of the Fig. 3 into embodiment illustrated in fig. 5.
Further, control node 900 can also include memory 903, and memory 903 is for storing control node 900 Program code and data.
In addition, control node 900 can also include bus 904.The memory 903 and the emitter/receiver 901 It is connected by bus 904 with the processor 902.
It is designed it is understood that Fig. 9 illustrate only simplifying for control node 900.In practical applications, control node 900 may include any number of transmitter, receiver, processor, memory etc., and all the embodiment of the present invention may be implemented Equipment all within the protection scope of the embodiment of the present invention.
It should be understood that referenced herein " multiple " refer to two or more."and/or", description association The incidence relation of object indicates may exist three kinds of relationships, for example, A and/or B, can indicate: individualism A exists simultaneously A And B, individualism B these three situations.Character "/" typicallys represent the relationship that forward-backward correlation object is a kind of "or".
The serial number of the above embodiments of the invention is only for description, does not represent the advantages or disadvantages of the embodiments.
Those of ordinary skill in the art will appreciate that realizing that all or part of the steps of above-described embodiment can pass through hardware It completes, relevant hardware can also be instructed to complete by program, the program can store in a kind of computer-readable In storage medium, storage medium mentioned above can be read-only memory, disk or CD etc..
The foregoing is merely presently preferred embodiments of the present invention, is not intended to limit the invention, it is all in spirit of the invention and Within principle, any modification, equivalent replacement, improvement and so on be should all be included in the protection scope of the present invention.

Claims (18)

1. a kind of elastic telescopic method, which is characterized in that the described method includes:
Acquire the monitoring data for the application supported flexible pond；It wherein, include at least one flexible group in the flexible pond, often One flexible group includes the virtual machine that at least one is pre-created, and is deployed with a kind of application on each virtual machine in advance；
Flexible decision is carried out to each described flexible group according to the monitoring data；
When decision dilatation target stretches the virtual machine in group, the void in a dormant state included from the flexible group of the target The first virtual machine is selected in quasi- machine, and wakes up first virtual machine；
It is described when decision dilatation target stretches the virtual machine in group, it is included in a dormant state from the flexible group of the target Virtual machine in select the first virtual machine, and wake up first virtual machine, comprising:
The target described in the decision dilatation stretch the virtual machine in group when, P layers of resource management system are sent to I layers of resource management system Wake up instruction, the wake up instruction are used to indicate the I layers of resource management system and wake up in a dormant state and belong to the mesh The virtual machine of the flexible group of mark；
The I layers of resource management system selects first void after receiving the wake up instruction from the flexible pond Quasi- machine, and wake up first virtual machine.
2. the method according to claim 1, wherein first virtual machine is on the first host in suspend mode State and the virtual machine for belonging to the flexible group of the target；First host is that current idle resource is most in the flexible pond It is more, and there is in a dormant state and belong to the host of the virtual machine of the flexible group of the target.
3. according to the method described in claim 2, it is characterized in that, the I layers of resource management system is receiving the wake-up After instruction, first virtual machine is selected from the flexible pond, and wakes up first virtual machine, comprising:
The I layers of resource management system is after receiving the wake up instruction, and selection is in suspend mode shape from the flexible pond State and the virtual machine for belonging to the flexible group of the target, and wake up the virtual machine selected；
The I layers of resource management system after receiving the migration instruction that the P layers of resource management system is sent, called out by detection Whether the awake virtual machine is located on first host；
If the virtual machine being waken up is not located on first host, the I layers of resource management system will be called out The awake virtual machine is migrated from original host to first host.
4. the method according to claim 1, wherein it is described according to the monitoring data to each described flexible group After carrying out flexible decision, further includes:
The target described in the decision volume reduction stretch the virtual machine in group when, from the flexible group of the target it is included be in wake-up states Virtual machine in select the second virtual machine, and the second virtual machine described in suspend mode；
Wherein, second virtual machine is to be in wake-up states on the second host and belong to the virtual of the flexible group of the target Machine；Second host is that current idle resource is minimum in the flexible pond, and exists in wake-up states and belong to described The host of the virtual machine of the flexible group of target.
5. according to the method described in claim 4, it is characterized in that, the target described in the decision volume reduction stretch it is virtual in group When machine, the second virtual machine is selected in the virtual machine in wake-up states included from the flexible group of the target, and described in suspend mode Second virtual machine, comprising:
The target described in the decision volume reduction stretch the virtual machine in group when, the P layers of resource management system is to the I layers of resource management System sends the first dormancy instruction, and first dormancy instruction is used to indicate the I layers of resource management system suspend mode and is in wake-up State and the virtual machine for belonging to the flexible group of the target；
The I layers of resource management system selects described after receiving first dormancy instruction from the flexible pond Two hosts；
The I layers of resource management system selects second virtual machine from second host, and the second void described in suspend mode Quasi- machine.
6. according to the method described in claim 4, it is characterized in that, the target described in the decision volume reduction stretch it is virtual in group When machine, the second virtual machine is selected in the virtual machine in wake-up states included from the flexible group of the target, and described in suspend mode Second virtual machine, comprising:
The target described in the decision volume reduction stretch the virtual machine in group when, the P layers of resource management system is from the I layers of resource management System queries obtain the current idle resource of each host in the flexible pond；
The P layers of resource management system selects second host according to the current idle resource of each host；
The P layers of resource management system sends the second dormancy instruction, second dormancy instruction to the I layers of resource management system The middle mark for carrying second host, second dormancy instruction are used to indicate the I layers of resource management system suspend mode institute State the virtual machine for being in wake-up states on the second host and belonging to the flexible group of the target；
The I layers of resource management system selects institute from second host after receiving second dormancy instruction State the second virtual machine, and the second virtual machine described in suspend mode.
7. according to the method described in claim 4, it is characterized in that, the target described in the decision volume reduction stretch it is virtual in group When machine, the second virtual machine is selected in the virtual machine in wake-up states included from the flexible group of the target, and described in suspend mode Second virtual machine, comprising:
The target described in the decision volume reduction stretch the virtual machine in group when, the P layer resource management system is from the flexible group of the target Select the virtual machine started earliest as second virtual machine in the included virtual machine in wake-up states；
The P layers of resource management system sends third dormancy instruction, the third dormancy instruction to the I layers of resource management system The middle mark for carrying second virtual machine, the third dormancy instruction are used to indicate the I layers of resource management system suspend mode institute State the second virtual machine；
The I layers of resource management system is after receiving the third dormancy instruction, the second virtual machine described in suspend mode.
8. method according to any one of claims 1 to 7, which is characterized in that the acquisition is answered what flexible pond was supported Monitoring data, comprising:
The P layers of resource management system acquires the operation data for the application that the flexible pond is supported, the operation data is used for Indicate the operation conditions of application；And/or
The P layers of resource management system receives the application that the flexible pond that the I layers of resource management system is sent is supported Performance data, the performance data are used to indicate the resource occupation situation of application.
9. method according to any one of claims 1 to 7, which is characterized in that the acquisition is answered what flexible pond was supported Before monitoring data, further includes:
After getting virtual machine creating instruction, the virtual machine of destination number is pre-created in shared storage；
According to the application deployment request got, a kind of application is disposed in advance on each virtual machine respectively, and described in pre-configuration Using；
The temporary unused virtual machine of suspend mode.
10. a kind of elastic telescopic system, which is characterized in that the system comprises:
Acquisition module, for acquiring the monitoring data for the application supported flexible pond；It wherein, include at least in the flexible pond One flexible group, each flexible group includes the virtual machine that at least one is pre-created, and is disposed in advance on each virtual machine There is a kind of application；
Decision-making module, for carrying out flexible decision to each described flexible group according to the monitoring data；
Wake-up module, for from when decision dilatation target stretches the virtual machine in group, stretching included by group from the target The first virtual machine is selected in the virtual machine of dormant state, and wakes up first virtual machine；
The system comprises: P layers of resource management system and I layers of resource management system；The wake-up module includes: to belong to the P The wake up instruction sending module of layer resource management system, and belong to the wake up instruction reception mould of the I layers of resource management system Block and virtual machine wake-up module；
The wake up instruction sending module, for the target described in the decision dilatation stretch the virtual machine in group when, I layer of Xiang Suoshu provide Management system sends wake up instruction, and the wake up instruction is used to indicate the I layers of resource management system and wakes up in suspend mode shape State and the virtual machine for belonging to the flexible group of the target；
The wake up instruction receiving module, for receiving the wake up instruction；
The virtual machine wake-up module, for selecting first virtual machine from the flexible pond, and it is empty to wake up described first Quasi- machine.
11. system according to claim 10, which is characterized in that first virtual machine is on the first host in not Dormancy state and the virtual machine for belonging to the flexible group of the target；First host is that current idle resource is most in the flexible pond It is more, and there is in a dormant state and belong to the host of the virtual machine of the flexible group of the target.
12. system according to claim 11, which is characterized in that the virtual machine wake-up module is specifically used for:
It selects in a dormant state and belongs to the virtual machine of the flexible group of the target from the flexible pond, and wake up and selected The virtual machine；
After the migration instruction that the wake up instruction receiving module receives that the P layers of resource management system is sent, quilt is detected Whether the virtual machine waken up is located on first host；
If the virtual machine being waken up is not located on first host, by the virtual machine being waken up from original Host is migrated to first host.
13. system according to claim 10, which is characterized in that the system also includes: sleep block；
The sleep block, for the target described in the decision volume reduction stretch the virtual machine in group when, from the flexible group institute of the target Including the virtual machine in wake-up states in select the second virtual machine, and the second virtual machine described in suspend mode；
Wherein, second virtual machine is to be in wake-up states on the second host and belong to the virtual of the flexible group of the target Machine；Second host is that current idle resource is minimum in the flexible pond, and exists in wake-up states and belong to described The host of the virtual machine of the flexible group of target.
14. system according to claim 13, which is characterized in that the sleep block includes: to belong to the P layers of resource pipe The dormancy instruction sending module of reason system, and belong to the dormancy instruction receiving module of the I layers of resource management system and virtual Machine sleep block；
The dormancy instruction sending module, for the target described in the decision volume reduction stretch the virtual machine in group when, I layer of Xiang Suoshu provide Management system sends the first dormancy instruction, and first dormancy instruction is used to indicate at the I layers of resource management system suspend mode In wake-up states and belong to the flexible virtual machine organized of the target；
The dormancy instruction receiving module, for receiving first dormancy instruction；
The virtual machine sleep block, for selecting second host from the flexible pond, from second host Upper selection second virtual machine, and the second virtual machine described in suspend mode.
15. system according to claim 13, which is characterized in that the sleep block includes: to belong to the P layers of resource pipe Enquiry module, selecting module and the dormancy instruction sending module of reason system, and belong to the suspend mode of the I layers of resource management system Command reception module and virtual machine sleep block；
The enquiry module, for the target described in the decision volume reduction stretch the virtual machine in group when, from the I layers of resource management system System inquiry obtains the current idle resource of each host in the flexible pond；
The selecting module selects second host for the current idle resource according to each host；
The dormancy instruction sending module, for sending the second dormancy instruction to the I layers of resource management system, described second stops It sleeps and carries the mark of second host in instructing, second dormancy instruction is used to indicate the I layers of resource management system It is in wake-up states on second host described in suspend mode and belongs to the virtual machine of the flexible group of the target；
The dormancy instruction receiving module, for receiving second dormancy instruction；
The virtual machine sleep block, for selecting second virtual machine from second host, and described in suspend mode Two virtual machines.
16. system according to claim 13, which is characterized in that the sleep block includes: to belong to the P layers of resource pipe The selecting module and dormancy instruction sending module of reason system, and belong to the dormancy instruction reception of the I layers of resource management system Module and virtual machine sleep block；
The selecting module, for the target described in the decision volume reduction stretch the virtual machine in group when, from the flexible group institute of the target Including the virtual machine in wake-up states in select the virtual machine started earliest as second virtual machine；
The dormancy instruction sending module, for sending third dormancy instruction to the I layers of resource management system, the third is stopped It sleeps and carries the mark of second virtual machine in instructing, the third dormancy instruction is used to indicate the I layers of resource management system Second virtual machine described in suspend mode；
The dormancy instruction receiving module, for receiving the third dormancy instruction；
The virtual machine sleep block, for the second virtual machine described in suspend mode.
17. system according to any one of claims 10 to 16, which is characterized in that the acquisition module is specifically used for:
The P layers of resource management system acquires the operation data for the application that the flexible pond is supported, the operation data is used for Indicate the operation conditions of application；And/or
The P layers of resource management system receives the application that the flexible pond that the I layers of resource management system is sent is supported Performance data, the performance data are used to indicate the resource occupation situation of application.
18. system according to any one of claims 10 to 16, which is characterized in that the system also includes: pretreatment mould Block；The preprocessing module, is specifically used for:
After getting virtual machine creating instruction, the virtual machine of destination number is pre-created in shared storage；
According to the application deployment request got, a kind of application is disposed in advance on each virtual machine respectively, and described in pre-configuration Using；
The temporary unused virtual machine of suspend mode.
CN201610657659.3A 2016-08-10 2016-08-10 Elastic telescopic method and system Active CN106227582B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201610657659.3A CN106227582B (en) 2016-08-10 2016-08-10 Elastic telescopic method and system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201610657659.3A CN106227582B (en) 2016-08-10 2016-08-10 Elastic telescopic method and system

Publications (2)

Publication Number Publication Date
CN106227582A CN106227582A (en) 2016-12-14
CN106227582B true CN106227582B (en) 2019-06-11

Family

ID=57548085

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201610657659.3A Active CN106227582B (en) 2016-08-10 2016-08-10 Elastic telescopic method and system

Country Status (1)

Country Link
CN (1) CN106227582B (en)

Families Citing this family (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108628660B (en) * 2017-03-24 2021-05-18 华为技术有限公司 Virtual machine capacity expansion and reduction method and virtual management equipment
CN108574718B (en) * 2017-04-19 2021-03-26 北京金山云网络技术有限公司 Cloud host creation method and device
CN108574719B (en) * 2017-04-19 2021-04-20 北京金山云网络技术有限公司 Cloud host creation method and device
CN109002354B (en) * 2017-06-07 2022-05-03 中国科学院信息工程研究所 OpenStack-based computing resource capacity elastic expansion method and system
CN109302302B (en) * 2017-07-25 2022-08-12 中兴通讯股份有限公司 Method, system and computer readable storage medium for scaling service network element
CN107810488A (en) * 2017-08-11 2018-03-16 深圳前海达闼云端智能科技有限公司 A kind of method of state management of virtual machine, device and intelligent terminal
CN109995564B (en) * 2017-12-31 2022-04-15 中国移动通信集团辽宁有限公司 Method, device, equipment and medium for automatically opening maintenance port
CN109873718A (en) * 2019-01-23 2019-06-11 平安科技（深圳）有限公司 A kind of container self-adapting stretching method, server and storage medium
CN111580925B (en) * 2019-02-18 2023-11-17 华为技术有限公司 Method and device for stretching application
CN113296931A (en) * 2020-07-14 2021-08-24 阿里巴巴集团控股有限公司 Resource control method, system, computing device and storage medium
CN112000606A (en) * 2020-07-22 2020-11-27 中国建设银行股份有限公司 Computer cluster and infrastructure cluster suitable for deploying application cluster
CN115904717B (en) * 2022-11-30 2023-11-17 湖南长银五八消费金融股份有限公司 Application system deployment method and device in cloud platform and computer equipment

Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101719081A (en) * 2009-12-01 2010-06-02 北京大学 Method for scheduling virtual machines
CN102662750A (en) * 2012-03-23 2012-09-12 上海交通大学 Virtual machine resource optimal control method and control system based on elastic virtual machine pool
JP2012208781A (en) * 2011-03-30 2012-10-25 Internatl Business Mach Corp <Ibm> Information processing system, information processing apparatus, scaling method, program, and recording medium
CN103034453A (en) * 2011-09-30 2013-04-10 国际商业机器公司 Method and device for managing persistent data of preinstalled application in virtual machine instance
KR20130046040A (en) * 2011-10-27 2013-05-07 삼성에스디에스 주식회사 Fuzzy control based virtual machine auto scaling system and method
CN103220337A (en) * 2013-03-22 2013-07-24 合肥工业大学 Cloud computing resource optimizing collocation method based on self-adaptation elastic control
CN103810020A (en) * 2014-02-14 2014-05-21 华为技术有限公司 Virtual machine elastic scaling method and device
CN103870341A (en) * 2014-03-12 2014-06-18 汉柏科技有限公司 Method and system of adjusting resources of virtual machine
CN104219290A (en) * 2014-08-19 2014-12-17 南京邮电大学 Flexible configuring method of multi-module cloud application
CN104317638A (en) * 2014-10-17 2015-01-28 华为技术有限公司 Application stretching management method and device

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9372735B2 (en) * 2012-01-09 2016-06-21 Microsoft Technology Licensing, Llc Auto-scaling of pool of virtual machines based on auto-scaling rules of user associated with the pool

Patent Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101719081A (en) * 2009-12-01 2010-06-02 北京大学 Method for scheduling virtual machines
JP2012208781A (en) * 2011-03-30 2012-10-25 Internatl Business Mach Corp <Ibm> Information processing system, information processing apparatus, scaling method, program, and recording medium
CN103034453A (en) * 2011-09-30 2013-04-10 国际商业机器公司 Method and device for managing persistent data of preinstalled application in virtual machine instance
KR20130046040A (en) * 2011-10-27 2013-05-07 삼성에스디에스 주식회사 Fuzzy control based virtual machine auto scaling system and method
CN102662750A (en) * 2012-03-23 2012-09-12 上海交通大学 Virtual machine resource optimal control method and control system based on elastic virtual machine pool
CN103220337A (en) * 2013-03-22 2013-07-24 合肥工业大学 Cloud computing resource optimizing collocation method based on self-adaptation elastic control
CN103810020A (en) * 2014-02-14 2014-05-21 华为技术有限公司 Virtual machine elastic scaling method and device
CN103870341A (en) * 2014-03-12 2014-06-18 汉柏科技有限公司 Method and system of adjusting resources of virtual machine
CN104219290A (en) * 2014-08-19 2014-12-17 南京邮电大学 Flexible configuring method of multi-module cloud application
CN104317638A (en) * 2014-10-17 2015-01-28 华为技术有限公司 Application stretching management method and device

Also Published As

Publication number Publication date
CN106227582A (en) 2016-12-14

Similar Documents

Publication Publication Date Title
CN106227582B (en) Elastic telescopic method and system
CN108737270B (en) Resource management method and device for server cluster
CN111796908B (en) System and method for automatic elastic expansion and contraction of resources and cloud platform
US8881149B2 (en) Control of java resource runtime usage
CN111385114B (en) VNF service instantiation method and device
US20180367460A1 (en) Data flow processing method and apparatus, and system
US20210117231A1 (en) Task processing method and apparatus
EP3584998B1 (en) Method for virtual machine capacity expansion and reduction and virtual management device
CN100517221C (en) Efficient multiprocessor system and methods thereof
WO2018121334A1 (en) Web application service providing method, apparatus, electronic device and system
CN109766172B (en) Asynchronous task scheduling method and device
CN110166507B (en) Multi-resource scheduling method and device
WO2016155291A1 (en) Method and apparatus for managing virtualized network function scaling
WO2018196462A1 (en) Resource scheduling apparatus, resource scheduling system and resource scheduling method
CN109960579B (en) Method and device for adjusting service container
CN112600761A (en) Resource allocation method, device and storage medium
CN106911741B (en) Method for balancing virtual network management file downloading load and network management server
WO2019109948A1 (en) Paas management method and device, and storage medium
CN106664259B (en) Method and device for expanding virtual network function
CN112424749A (en) On-demand code execution with limited memory footprint
JP6488910B2 (en) Control method, control program, and information processing apparatus
CN113726881B (en) Communication connection establishment method, related device and computer readable storage medium
CN111045778B (en) Virtual machine creation method, device, server and storage medium
WO2021057405A1 (en) Resource sharing method and device
US20180329758A1 (en) Method and device for transmitting managing instruction and managing automatic scaling function of vnf

Legal Events

Date Code Title Description
C06 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