CN107818013A - A kind of application scheduling method thereof and device - Google Patents

Abstract

Embodiments of the invention provide a kind of application scheduling method thereof and device, are related to communication technical field, it is possible to increase the success rate of application deployment.This method includes：Dispatching device receives set of applications from deployment managing device, the set of applications include in the set of applications it is each treat application deployment and this treat corresponding relation between the resource requirement of application deployment；The dispatching device, which obtains the idling-resource amount of each resource node and default deployment conditions, the deployment conditions in multiple resource nodes, includes the default SLA conditions of each resource node；Each application deployment are treated for this, the dispatching device is according to the deployment conditions, this is treated that application deployment are deployed in idling-resource amount and meet that this is treated on the resource node of the resource requirement of application deployment, wherein, the service-level agreement SLA parameters for disposing the resource node after application deployment meet the default SLA conditions of the resource node.This method is applied in application schedules scene.

Description

A kind of application scheduling method thereof and device

Technical field

The present invention relates to communication technical field, more particularly to a kind of application scheduling method thereof and device.

Background technology

With the rapid development of virtualization technology, virtualization technology develops into operating system from the virtualization of physical machine The containerization stage.Containerization technique is a kind of resource virtualizing technique of the lightweight of operating system layer, can be realized single The isolation of process and resource is supported on main frame so that main frame is in the case where performance loss is minimum, there is provided thousands of individual high density every From the container application (application run in a reservoir) of operation, resource utilization ratio is at utmost improved.

Container cluster management, it is the key technology of layout and the scheduling of completing container.At present, the most widely used container Cluster management system is Kubernetes (k8s) system.K8s systems support that one kind is " hard " selector, and it ensures selected Main frame must is fulfilled for needing the affine sexual intercourse between the container application disposed, for example, container application A and container application B tools There is compatibility (can be understood as binding relationship), then container application A and container application B must be deployed on same main frame, held Device application A and container application C has anti-compatibility (can be understood as resource contention relation), then container application A and container application C It must be deployed on different main frames.

However, when the host resource in system is insufficient for some affine sexual intercourse treated between deployment container application When, for example, no idling-resource amount in system is capable of the main frame of deployment container application A and container application B simultaneously, or system In there is no enough host numbers so that container application A and container application B are separately disposed, " hard " selector will be unable to complete this The deployment of a little container applications, so as to cause these container applications not run.

The content of the invention

Embodiments of the invention provide a kind of application scheduling method thereof and device, it is possible to increase the success rate of application deployment.

To reach above-mentioned purpose, embodiments of the invention adopt the following technical scheme that：

In a first aspect, the embodiment of the present invention provides a kind of application scheduling method thereof, this method includes：

Dispatching device receives set of applications from deployment managing device, and the set of applications includes each treating in the set of applications Application deployment and this treat corresponding relation between the resource requirement of application deployment；The dispatching device obtains every in multiple resource nodes The idling-resource amount and default deployment conditions of individual resource node, the deployment conditions include the default SLA of each resource node Condition；Application deployment each are treated for this, this is treated that application deployment are deployed in idle money by the dispatching device according to the deployment conditions Source amount meets that this is treated on the resource node of the resource requirement of application deployment, wherein, dispose the resource section after application deployment The service-level agreement SLA parameters of point meet the default SLA conditions of the resource node.

The embodiment of the present invention provides a kind of application scheduling method thereof, dispatching device can the SLA conditions based on resource node, will Treat that application deployment are deployed in and meet that this is treated on the resource node of the resource requirement of application deployment so that the deployment of application is no longer limited In fixed affine sexual intercourse, so as to which dispatching device can be in the case where meeting SLA conditions, more efficient deployment is more Using, improve using deployment success rate.

In a kind of possible design, dispatching device obtains the idling-resource of each resource node in multiple resource nodes Amount, including：The dispatching device receives the idling-resource amount of each resource node from detection means.

In a kind of possible design, set of applications also includes treating that the compatibility between application deployment is closed in the set of applications System, sexual intercourse that this is affine include the binding relationship treated between application deployment or resource contention relation in the set of applications, should answer Treat that application deployment treat application deployment for one group with binding relationship with set；Application deployment, dispatching device are treated for each According to deployment conditions, it will treat that application deployment are deployed in idling-resource amount and meet that this treats the resource node of the resource requirement of application deployment On, wherein, the SLA parameters for disposing the resource node after application deployment meet the default SLA conditions of the resource node, bag Include：Application deployment are treated for each group in the set of applications, the dispatching device treats that deployment should according to the deployment conditions by one group Met with idling-resource amount is deployed on the resource node of one group of total resources demand for treating application deployment, wherein, one group of deployment is treated The SLA parameters of the resource node after application deployment meet the default SLA conditions of the resource node.

In a kind of possible design, if the dispatching device treats that application deployment are disposed according to the deployment conditions by this group Before idling-resource amount meets on the resource node of this group total resources demand for treating application deployment, on the plurality of resource node With the application disposed；The dispatching device treats that application deployment are deployed in idling-resource amount according to the deployment conditions by this group On the resource node for meeting this group total resources demand for treating application deployment, including：The dispatching device, will according to the deployment conditions This group is treated that application deployment are deployed in idling-resource amount and meet this group total resources demand for treating application deployment, and that has disposed answers Treat that application deployment do not possess on the resource node of the resource contention relation with this group.

In a kind of possible design, set of applications also includes the affine sexual intercourse that application deployment are treated in the set of applications, Sexual intercourse that this is affine includes treating the resource contention relation between application deployment in the set of applications, if the dispatching device is according to the portion Administration's condition, this is treated that application deployment are deployed in idling-resource amount and meet that this treats on the resource node of the resource requirement of application deployment it Before, there is the application disposed, this is treated application deployment portion by the dispatching device according to the deployment conditions on the plurality of resource node Administration meets that this is treated on the resource node of the resource requirement of application deployment in idling-resource amount, including：The dispatching device is according to the portion Administration's condition, this is treated that application deployment are deployed in idling-resource amount and meet that this treats the resource requirement of application deployment, and that has disposed answers Treat that application deployment do not possess the resource node of the resource contention relation for this.

In above-mentioned three kinds of possible designs, dispatching device can be pressed in the case where meeting the SLA conditions of resource node According to the resource requirement for treating application deployment and the affine sexual intercourse between application deployment is treated, application deployment, which are disposed, to be treated to this, from And further ensure after application deployment after completing to dispose, the quality of service in running.

In a kind of possible design, dispatching device will treat that application deployment are deployed in idling-resource amount according to deployment conditions Before meeting that this is treated on the resource node of the resource requirement of application deployment, this method also includes：The dispatching device determines the plurality of It is not present in resource node and disposes this after application deployment, the resource node of SLA conditions is preset corresponding to SLA parameters satisfaction；Should Dispatching device redeploys to the certain applications disposed on the plurality of resource node, to cause in the plurality of resource node Part resource node idling-resource amount increase；This is treated that application deployment are deployed in by the dispatching device according to the deployment conditions Idling-resource amount meets that this is treated on the resource node of the resource requirement of application deployment, including：The dispatching device is according to the deployment bar Part, this is treated that application deployment are deployed in idle stock number in the part resource node and meet that this treats the resource requirement of application deployment On resource node.

In this possible design, when the idling-resource amount of multiple resource nodes is unsatisfactory for treating that the deployment of application deployment will When asking, by way of the application that can have been disposed redeploying the part on the plurality of resource node, increase multiple resource sections The idling-resource amount of part resource node in point, to enable the part resource node in the case where meeting SLA conditions Affix one's name to this and treat application deployment, so as to further increase treat application deployment be deployed to power.

In a kind of possible design, this method also includes：The scheduling that the dispatching device receives detection means transmission is determined Plan message, the scheduling decision message include adjust instruction and the mark of at least one application to be adjusted, and the adjust instruction is used for Indicate that the dispatching device redeploys at least one application to be adjusted of the mark instruction of at least one application to be adjusted；It is right Each application to be adjusted at least one application to be adjusted, the dispatching device is according to the adjust instruction and the deployment bar Part, the application to be adjusted is redeployed to another resource section for meeting the resource requirement of the application to be adjusted in idling-resource amount On point, wherein, the SLA parameters for another resource node disposed after the application to be adjusted meet another resource node Default SLA conditions.

In this possible design, dispatching device can be discontented with to performance data in time according to the instruction of detection means The application to be adjusted of SLA conditions corresponding to foot is redeployed, and is improved the reliability of application in the process of running, is optimized system Performance.

Second aspect, the embodiment of the present invention provide a kind of application scheduling method thereof, and this method includes：Detection means obtain with it is more Monitoring Data corresponding to each resource node in individual resource node, what the Monitoring Data included having disposed on the resource node answers Performance data；The detection means calculates the idling-resource amount of the resource node according to the Monitoring Data；The detection means to Dispatching device sends the idling-resource amount of each resource node calculated, and the idling-resource amount of each resource node is used for should Application deployment are treated in dispatching device deployment.

By application scheduling method thereof provided in an embodiment of the present invention, detection means can be by monitoring each resource section in real time The Monitoring Data of point, calculates the current idling-resource amount of each resource node, and by the current idling-resource of each resource node Amount is sent to dispatching device, and to enable effective idling-resource amount of the dispatching device according to each resource node, portion is treated in deployment Administration's application, so as to improve the resource utilization of resource node.

In a kind of possible design, this method also includes：The detection means is according to corresponding with each resource node Monitoring Data and default SLA conditions corresponding with each application disposed, determine what is disposed on the plurality of resource node At least one application to be adjusted in all applications be present, the performance data of the application to be adjusted is unsatisfactory for the application pair to be adjusted The SLA conditions answered；The detection means sends scheduling decision message to the dispatching device, and the scheduling decision message includes this at least The mark and adjust instruction of one application to be adjusted, the adjust instruction are at least one for indicating that the dispatching device redeploys this Application to be adjusted.

In this possible design, detection means can monitor the running status of each application disposed in real time, and When it is determined that the application to be adjusted that performance data is unsatisfactory for corresponding SLA conditions in the application disposed be present, instruction in time is adjusted Degree device redeploys to the application to be adjusted, improves the reliability of application in the process of running, optimizes the performance of system.

The third aspect, the embodiment of the present invention provide a kind of dispatching device, including：Receiving unit, for being filled from deployment management Reception set of applications is put, the set of applications includes each treating that the resource of application deployment is treated in application deployment with this in the set of applications Corresponding relation between demand；Acquiring unit, for obtain in multiple resource nodes the idling-resource amount of each resource node and Default deployment conditions, the deployment conditions include the default SLA conditions of each resource node；Deployment unit, for for this Application deployment each are treated, according to the deployment conditions, this is treated that application deployment are deployed in idling-resource amount and meet that this treats application deployment Resource requirement resource node on, wherein, dispose the service-level agreement SLA ginsengs of the resource node after application deployment Number meets the default SLA conditions of the resource node.

In a kind of possible design, acquiring unit obtains the idling-resource amount of each resource node in multiple resource nodes When, specifically include：The acquiring unit receives the idling-resource amount of each resource node from detection means.

In a kind of possible design, the set of applications that receiving unit receives also includes treating that deployment should in the set of applications Affine sexual intercourse between, sexual intercourse that this is affine include the binding relationship or money treated between application deployment in the set of applications Source competitive relation, have binding relationship in the set of applications treats that application deployment treat application deployment for one group, the deployment unit, tool Body is used to treat application deployment for each group in the set of applications, treats that application deployment are deployed in idling-resource amount and met by one group This group treated on the resource node of total resources demand of application deployment, wherein, dispose resource section of this group after application deployment The SLA parameters of point meet the default SLA conditions of the resource node.

In a kind of possible design, if deployment unit treats that application deployment are deployed in the free time according to deployment conditions by one group Before stock number meets that this group is treated on the resource node of the total resources demand of application deployment, have on the plurality of resource node The application of deployment, the deployment unit treat that application deployment are deployed in idling-resource amount and meet this according to the deployment conditions by one group Group is treated on the resource node of total resources demand of application deployment, is specifically included：The deployment unit is according to the deployment conditions, by one group Treat that application deployment are deployed in idling-resource amount and meet that this group treats the total resources demand of application deployment, and the application disposed and this One group is treated that application deployment do not possess on the resource node of the resource contention relation.

In a kind of possible design, the set of applications that receiving unit receives also includes treating that deployment should in the set of applications Affine sexual intercourse, sexual intercourse that this is affine include treating the resource contention relation between application deployment in the set of applications, if should This is treated that application deployment are deployed in idling-resource amount and meet that this treats the resource need of application deployment by deployment unit according to the deployment conditions Before on the resource node asked, there is the application disposed on the plurality of resource node, the deployment unit according to the deployment conditions, This is treated that application deployment are deployed in idling-resource amount and meet that this is treated on the resource node of the resource requirement of application deployment, specific bag Include：This is treated that application deployment are deployed in idling-resource amount and meet that this treats application deployment by the deployment unit according to the deployment conditions Resource requirement, and that has disposed treats that application deployment do not possess on the resource node of the resource contention relation applied to this.

In a kind of possible design, deployment unit, it is additionally operable to, according to deployment conditions, to treat that application deployment are deployed in sky Before not busy stock number meets that this is treated on the resource node of the resource requirement of application deployment, determine to be not present in the plurality of resource node This is disposed after application deployment, the resource node of SLA conditions is preset corresponding to SLA parameters satisfaction；And to the plurality of resource node On the certain applications disposed redeployed, to cause the idling-resource of the part resource node in the plurality of resource node Amount increase；This is treated that application deployment are deployed in idling-resource amount and meet that this treats that deployment should by the deployment unit according to the deployment conditions On the resource node of resource requirement, specifically include：This is treated that application deployment are disposed by the deployment unit according to the deployment conditions Idle stock number meets that this is treated on the resource node of the resource requirement of application deployment in the part resource node.

In a kind of possible design, receiving unit, the scheduling decision message that receiving detection device is sent, the tune are additionally operable to Degree decision information includes adjust instruction and the mark of at least one application to be adjusted, and the adjust instruction is used to indicate the deployment list Member redeploys at least one application to be adjusted of the mark instruction of at least one application to be adjusted；The deployment unit, also For for each application to be adjusted at least one application to be adjusted, according to the adjust instruction and the deployment conditions, inciting somebody to action The application to be adjusted redeploys to be met on another resource node of the resource requirement of the application to be adjusted in idling-resource amount, Wherein, the SLA parameters for another resource node disposed after the application to be adjusted meet the default of another resource node SLA conditions.

The technique effect of dispatching device provided in an embodiment of the present invention may refer to above-mentioned first aspect or first aspect The technique effect of each optional mode, here is omitted.

Fourth aspect, the embodiment of the present invention provide a kind of detection means, including：Acquiring unit, for obtaining and multiple moneys Monitoring Data corresponding to each resource node in source node, the Monitoring Data include the application disposed on the resource node Performance data；Computing unit, the Monitoring Data for being obtained according to the acquiring unit calculate the idling-resource amount of the resource node； Transmitting element, the idling-resource amount of each resource node for sending computing unit calculating to dispatching device, each resource The idling-resource amount of node is used for dispatching device deployment and treats application deployment.

In a kind of possible design, detection means also includes determining unit, the determining unit, for according to acquiring unit The Monitoring Data corresponding with each resource node and the default grade of service corresponding with each application disposed obtained is assisted SLA conditions are discussed, at least one application to be adjusted in all applications for determining to have disposed on the plurality of resource node be present, this waits to adjust The performance data of whole application is unsatisfactory for SLA conditions corresponding to the application to be adjusted；The transmitting element, it is additionally operable to fill to the scheduling Put and send scheduling decision message, the scheduling decision message includes the mark of at least one application to be adjusted and adjustment refers to Order, the adjust instruction are used to indicate that the dispatching device redeploys at least one application to be adjusted.

The technique effect of detection means provided in an embodiment of the present invention may refer to above-mentioned second aspect or second aspect The technique effect of each optional mode, here is omitted.

It should be noted that above-mentioned first aspect to treat application deployment and the application to be adjusted described in fourth aspect can be with It is in container application or the application operated in non-container environment.

5th aspect, the embodiments of the invention provide a kind of dispatching device, the dispatching device can realize above-mentioned aspect institute It is related in method the function performed by dispatching device, the function can be realized by hardware, can also perform phase by hardware The software answered is realized.The hardware or software include one or more above-mentioned corresponding modules of function.

In a kind of possible design, the structure of the dispatching device includes processor and communication interface, the processor quilt It is configured to support the dispatching device to perform corresponding function in the above method.The communication interface is used to support the dispatching device and its Communication between his network element.The dispatching device can also include memory, and the memory is used to couple with processor, and it is preserved should The necessary programmed instruction of dispatching device and data.

6th aspect, the embodiment of the present invention provides a kind of computer-readable recording medium, including computer executed instructions, when dispatching device Described in computing device during computer executed instructions, the dispatching device performs any of such as above-mentioned first aspect or first aspect Dispositions method is applied described in a kind of implementation.

7th aspect, the embodiments of the invention provide a kind of detection means, the detection means can realize above-mentioned aspect institute It is related in embodiment of the method the function performed by detection means, the function can be realized by hardware, can also pass through hardware Corresponding software is performed to realize.The hardware or software include one or more above-mentioned corresponding modules of function.

In a kind of possible design, the structure of the detection means includes processor and communication interface, the processor quilt It is configured to support the detection means to perform corresponding function in the above method.The transceiver is used to support the detection means and other Communication between network element.The detection means can also include memory, and the memory is used to couple with processor, and it preserves the inspection Survey the necessary programmed instruction of device and data.

Eighth aspect, the embodiment of the present invention provides a kind of computer-readable recording medium, including computer executed instructions, when detection means Described in computing device during computer executed instructions, the detection means performs any of such as above-mentioned second aspect or second aspect Dispositions method is applied described in a kind of implementation.

9th aspect, the embodiments of the invention provide a kind of communication system, the system includes the scheduling described in above-mentioned aspect Device and detection means.

Compared to prior art, in communication system provided in an embodiment of the present invention, dispatching device can be based on resource section The SLA conditions of point, will treat that application deployment are deployed in meets that this is treated on the resource node of the resource requirement of application deployment so that application Deployment be no longer limited by the affine sexual intercourse of fixation, can more increase in the case where meeting SLA conditions so as to dispatching device The deployment of effect is more applied, and improves the success rate using deployment.

Brief description of the drawings

Technical scheme in order to illustrate the embodiments of the present invention more clearly, below will be required in the description to embodiment The accompanying drawing used is briefly described, it should be apparent that, drawings in the following description are only some embodiments of the present invention, and The embodiment being not all of.

Fig. 1 is a kind of communication system block diagram provided in an embodiment of the present invention；

Fig. 2 is a kind of structural representation of physical machine provided in an embodiment of the present invention；

Fig. 3 is a kind of interaction figure of application scheduling method thereof provided in an embodiment of the present invention；

Fig. 4 is a kind of schematic diagram of a scenario one of application scheduling method thereof provided in an embodiment of the present invention；

Fig. 5 is a kind of schematic diagram of a scenario two of application scheduling method thereof provided in an embodiment of the present invention；

Fig. 6 is a kind of schematic diagram of a scenario three of application scheduling method thereof provided in an embodiment of the present invention；

Fig. 7 is a kind of schematic diagram of a scenario four of application scheduling method thereof provided in an embodiment of the present invention；

Fig. 8 is a kind of schematic diagram of a scenario five of application scheduling method thereof provided in an embodiment of the present invention；

Fig. 9 is a kind of interaction figure of monitoring method provided in an embodiment of the present invention；

Figure 10 is a kind of interaction figure of dynamic adjusting method provided in an embodiment of the present invention；

Figure 11 is that the network architecture of a kind of Apache container applications and MySQL container applications that the present invention implements to provide is illustrated Figure；

Figure 12 A are a kind of structural representation one of dispatching device provided in an embodiment of the present invention；

Figure 12 B are a kind of structural representation two of dispatching device provided in an embodiment of the present invention；

Figure 12 C are a kind of structural representation three of dispatching device provided in an embodiment of the present invention；

Figure 13 A are a kind of structural representation one of detection means provided in an embodiment of the present invention；

Figure 13 B are a kind of structural representation two of detection means provided in an embodiment of the present invention；

Figure 13 C are a kind of structural representation three of detection means provided in an embodiment of the present invention；

Figure 14 A are a kind of structural representation one for disposing managing device provided in an embodiment of the present invention；

Figure 14 B are a kind of structural representation two for disposing managing device provided in an embodiment of the present invention；

Figure 14 C are a kind of structural representation three for disposing managing device provided in an embodiment of the present invention.

Embodiment

Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is clearly retouched State, it is clear that described embodiment is only part of the embodiment of the present invention, rather than whole embodiments.

As shown in figure 1, be a kind of communication system provided in an embodiment of the present invention, including deployment managing device, dispatching device, Detection means, multiple agent apparatus and resource node corresponding to each agent apparatus.Managing device is disposed with dispatching device to connect Connect, dispatching device is connected with detection means and each agent apparatus, and detection means is also connected with each agent apparatus.

Wherein, resource node can be virtual machine or physical machine, and multiple answer can be disposed on each resource node With, for example, using A, using B and application C etc..

Deployment managing device provides the user configuration interface, supports user's input, modification to treat application deployment set, deployment bar The information such as the affine sexual intercourse between part and application, and the information of user configuration is sent to dispatching device, to trigger scheduling Device calls related configuration condition and affine sexual intercourse, treats application deployment and is disposed.

Wherein, the affine sexual intercourse between can include compatibility and anti-compatibility, and compatibility is used to indicate at least Two application requirements are deployed on same resource node, for example, at least two applications of shared storage, share at least the two of network Individual application, or needing to carry out at least two applications of information exchange etc. possesses the application of height delay sensitive, during in order to meet Prolong requirement, these applications are generally bundled on same resource node, i.e., in embodiments of the present invention, will can need to be deployed in Relation is referred to as binding relationship between application on same resource node.

For example, ordering front end applications (Order Web App) and order back-end services (Order Service) is two and ties up Fixed application is, it is necessary to carry out information exchange, then is imitated to improve to order front end applications and order the communication between back-end services Rate, can will order front end applications and order back-end services are deployed on same resource node.

Anti- compatibility is used to indicate that at least two application requirements are deployed in different resource nodes, for example, port be present At least two applications of conflict, the application that at least two applications of CPU contention etc. have resource contention be present, it is each in order to ensure The process of application has independent high reliability resource, and these applications are usually deployed on different resource nodes.In this hair In bright embodiment, relation is referred to as resource contention relation between the application that will can need to be deployed on different resource nodes.

For example, there is resource contention relation between application in same cluster, then every in same cluster in order to ensure The quality of service of individual application, and in order to avoid a resource node breaks down, and cause the application in whole cluster all to stop Run, the application requirement in same cluster is deployed on different resource nodes.

Dispatching device is responsible for the affine sexual intercourse between the related configuration condition of storage and application, and according to configuration condition And the affine sexual intercourse between application, treat application deployment and disposed, and the application to having disposed but having needed to adjust is entered Row redeploys.

Detection means carries out research and application by the performance data of the application to having disposed, and analyzing in the application disposed is The no application adjusted in the presence of needs, and trigger the application that dispatching device adjusts to needs in time and redeploy.

Agent apparatus is responsible for monitoring the application that runs on corresponding resource node in real time, and by the property of the application of collection Energy data are sent to detection means.

In embodiments of the present invention, each agent apparatus can may be integrated on its corresponding resource node, Can be the virtual machine or physical machine independently of its corresponding resource node.

It can be respectively independent physical machine to dispose managing device, dispatching device and detection means, can also be respectively to transport The independent process or independent thread of row in same physical machine etc..

As shown in Fig. 2 be a kind of physical machine provided in an embodiment of the present invention, including：Processor, memory, bus, input Output interface and communication interface.Bus is connection processor, memory, input/output interface and communication interface, and is being handled Data transfer is realized between device, memory, input/output interface and communication interface.For example, processor is defeated from inputting by bus Outgoing interface receives order, decrypts the order received, and calculating or data processing are performed according to the order of decryption.Memory can be with Including program module, such as kernel (kernel), middleware (middleware), application programming interfaces (AP) and application etc..Journey Sequence module can be made up of software, firmware, hardware or therein at least two.Input/output interface forwards user to be set by input The order of standby (such as inductor, keyboard, touch-screen) input or data.Communication interface by this physical machine and other physical machines, Network is attached.For example, communication interface can be connected to outside other physics by being wired or wirelessly connected to network Machine.Optionally, display device can also be included in the physical machine, for user show by user input configuration information and To user's operation display interface etc..

What deserves to be explained is applying dispositions method provided in an embodiment of the present invention, suitable for container application, it is also applied for The application operated in non-container environment.

For container application, in embodiments of the present invention, at least one container operated in same container can be answered With being considered as, an application is overall, and application entirety is equally applicable to by application dispositions method provided in an embodiment of the present invention, That is the process provided in an embodiment of the present invention that container deployment is equally applicable to using dispositions method.

It is provided in an embodiment of the present invention to apply dispositions method based on communication system as shown in Figure 1, as shown in figure 3, the party Method can include：

S101, deployment managing device send set of applications to dispatching device, and the set of applications is included in the set of applications It is each treat application deployment and this treat corresponding relation between the resource requirement of application deployment.

Wherein, set of applications is that the configuration interface that user is provided by the deployment managing device is input to deployment management dress In putting, then the set of applications is sent to dispatching device by the deployment managing device, for indicating dispatching device to application Treat that application deployment are disposed in set.

The set of applications includes at least one resource requirement treated application deployment and each treat application deployment, each treats portion Administration's application and its resource requirement are mutually corresponding, and dispatching device can treat that application deployment are deployed according to the corresponding relation by each On the resource node that disclosure satisfy that its corresponding resource requirement.

Exemplary, the cpu resource and internal memory needed for application deployment can be treated including this by treating the resource requirement of application deployment Resource, can also including other deployments, this treats physical resource or virtual resource needed for application deployment.

S102, dispatching device obtain the idling-resource amount of each resource node and default deployment bar in multiple resource nodes Part, the deployment conditions include the default SLA conditions of each resource node.

Wherein, the default SLA conditions of each resource node can include the upper limit, interior of the CPU usage of the resource node Deposit the upper limit or other upper limit parameters of utilization rate etc..

Optionally, dispatching device can obtain deployment conditions from deployment managing device, can also be obtained in advance from local The deployment conditions of storage.

Exemplary, the deployment conditions are that user is input to deployment managing device by disposing the interface that managing device provides In, when the dispatching device acquisition request deployment conditions, the deployment conditions can be sent to scheduling by deployment managing device Device.Or deployment managing device directly can also send the deployment conditions after the deployment conditions of user's input are received To dispatching device, the deployment conditions are stored in local by dispatching device, when dispatching device, which treats application deployment, to be disposed, The dispatching device can be directly from the local deployment conditions for reading storage.

Optionally, dispatching device can also obtain the idling-resource amount of each resource node in several ways.For example, adjust The idling-resource amount of each resource node can be received from detection means by spending device, can also be from multiple resource sections of local maintenance Obtained in the resource service condition database of point, wherein, resource service condition database includes the attribute of each resource node Information, the information for the application disposed on each resource node, and idling-resource amount of each resource node etc..Detection means can Periodically to report the idling-resource amount of each resource node, so as to which dispatching device is to the resource service condition number of local maintenance It is updated according to storehouse.

S103, application deployment are treated for each, dispatching device will treat that application deployment are deployed in idle money according to deployment conditions Source amount meets that this is treated on the resource node of the resource requirement of application deployment, wherein, dispose the resource node after application deployment SLA parameters meet the default SLA conditions of the resource node.

Exemplary, by taking application A as an example, it is assumed that treat application deployment using A for one in the set of applications, this applies A Resource requirement include 5% cpu resource, and 300M memory source, the SLA conditions of each resource node include：It is interior The upper limit for depositing utilization rate is 75%, and the upper limit of CPU usage is 80%.Dispatching device is according to the sky of each resource node of acquisition Not busy stock number, determine that idling-resource amount disclosure satisfy that the resource requirement using A, and after CPU usage increase by 5% still less than Resource node of the utilization rate still less than 75% after 80%, the usage amount increase 300M of internal memory.

For example, dispatching device determines that the CPU usage of the resource node 1 in the plurality of resource node is 20%, there is 80% Idle cpu resource, also have 1.5G to be not used in the 2G of resource node 1 internal memory, i.e. the idling-resource amount of resource node 1 can Meet the resource requirement using A, and be 25% after CPU usage increase by 5%, still less than 80%, the internal memory of resource node 1 After usage amount increase 300M, memory usage 40%, still less than 75%, dispatching device can determine that resource node 1 is Meet the resource node of the deployment request using A, so as to which dispatching device will can be deployed on resource node 1 using A, and can Guarantee is deployed using after A, the SLA parameters of resource node 1, including CPU usage and memory usage, meets resource node 1 default SLA conditions.

It should be noted that multiple moneys that disclosure satisfy that the deployment request using A are there may be in the plurality of resource node Source node, dispatching device can randomly choose one of resource node application deployment A, can also it is determined that first can Meet application deployment A on the resource node using A deployment request, on the other hand, the embodiment of the present invention is not limited.

By application scheduling method thereof provided in an embodiment of the present invention, dispatching device can the SLA conditions based on resource node, To treat that application deployment are deployed in and meet that this is treated on the resource node of the resource requirement of application deployment so that the deployment of application no longer by It is limited to the affine sexual intercourse of fixation, so as to which dispatching device can be in the case where meeting SLA conditions, more efficient deployment is more Application, improve using deployment success rate.

In one example, if also including treating that deployment should in the set of applications in the set of applications that deployment managing device is sent Affine sexual intercourse between, then dispatching device can also be according in deployment conditions and the compatibility relation pair set of applications Treat that application deployment are disposed.

In embodiments of the present invention, application deployment are treated with binding relationship if existing in the set of applications, can incited somebody to action Treat that application deployment are defined as one group and treat application deployment with binding relationship, for example, there is binding relationship using A and using B, then It is one group using A and application B and treats application deployment.If there are 3 examples respectively using A and using B in the set of applications, to answer With A1, using A2, using A3 and application B1, using B2, using B3, then using A1 and application B1 can be one group treat deployment should With, using A2 and application B2 can be one group treat application deployment, using A3 and application B3 can be one group treat application deployment.

It should be noted that it is identical application using A1, using A2 and application A3, all for using A, in the present invention In embodiment, for the ease of description, these three examples are referred to as using A1, using A2 and using A3.Similarly, using B1, It is also identical application using B2, using B3, all for using B.

Application deployment are treated for each group in the set of applications, dispatching device can be according to deployment conditions, by one group Treat that application deployment are deployed in idling-resource amount and meet that this group is treated on the resource node of total resources demand of application deployment, wherein, The SLA parameters for disposing resource node of this group after application deployment meet the default SLA conditions of the resource node.

Exemplary, exemplified by by the application A1 in set of applications and applying this group of B1 to treat application deployment, it is assumed that using A1 Resource requirement include 5% cpu resource, and 300M memory source, 8% CPU is included using B1 resource requirement and is provided Source, and 200M memory source, then the memory source and 13% that the total resources demand using A1 and application B1 is 500M Cpu resource, the SLA conditions of each resource node include：The upper limit of memory usage is 75%, and the upper limit of CPU usage is 80%.Dispatching device determines that idling-resource amount disclosure satisfy that using A1 according to the idling-resource amount of each resource node of acquisition With application B1 total resources demand, and still less than 80% after CPU usage increase by 13%, the usage amount of internal memory increase 500M it Resource node of the utilization rate still less than 75% afterwards.

For example, dispatching device determines that the idling-resource amount of the resource node 1 in resource pool disclosure satisfy that using A1 and application B1 total resources demand, and be 33% after CPU usage increase by 13%, still less than 80%, the internal memory of resource node 1 uses After amount increase 500M, memory usage 49%, still less than 75%, dispatching device can determine that resource node 1 meets to answer The resource node of deployment request with A1 and using B1, so as to which dispatching device can will apply A1 and be deployed in resource section using B1 On point 1, and it can ensure to deploy using after A1 and application B1, the SLA parameters of resource node 1, including 33% CPU are used Rate, 49% memory usage, still meet the default SLA conditions of resource node 1.

Similarly, dispatching device can be according to the deployment process to application A1 and application B1, to applying A2 in set of applications Treat that application deployment are disposed with using B2, and using A3 and using this two groups of B3, it is exemplary, as shown in figure 4, scheduling dress Putting application A2 and will be deployed in using B2 and can will apply A3 and using B3 on application deployment A2 and application B2 resource node 2 Be deployed in can application deployment A3 and application B3 resource node 3 on.

It is exemplary at one, if dispatching device deployment each group treat application deployment before, have on the plurality of resource node There is the application disposed, then for dispatching device when disposing each group when application deployment, the dispatching device can be according to deployment bar Part, treat that application deployment are deployed in idling-resource amount and meet that this group treats the total resources demand of application deployment by one group, and disposed Application treat that application deployment do not possess on the resource node of resource contention relation with this group.

For example, using A 3 exemplary application A1, using A2, using possessing resource contention relation between A3, using the 3 of B Individual exemplary application B1, using B2, using also possessing resource contention relation between B3.So, as shown in figure 4, when dispatching device is incited somebody to action After being deployed in using A1 and using B1 on resource node 1, in application deployment A2 and application B2, due on resource node 1 Deploy with using A2 possess resource contention relation apply A1, with using B2 possess resource contention relation apply B1, therefore, Dispatching device need select resource pool in other resource node application deployment A2 and using B2.Assuming that non-portion on resource node 2 Possesses the application of resource contention relation between administration and application A2 and/or application B2, and the idling-resource amount of resource node 2 can expire Foot is using A2 and the total resources demand using B2, and can ensure after deploying using A2 and application B2, resource node 2 SLA parameters still meet the SLA conditions of resource node 2, then dispatching device can will apply A2 and be deployed in resource using B2 On node 2.

Similarly, after dispatching device will apply A2 and will be deployed in using B2 on resource node 2, in application deployment A3 and application , it is necessary to select the resource node in addition to resource node 1 and resource node 2 to carry out application deployment A3 and using B3 during B3.

Further, in set of applications except each group have binding relationship treat application deployment in addition to each Application deployment are treated, or, for being not present in set of applications with binding relationship when application deployment, in the set of applications Each treats application deployment, and the dispatching device can will treat that application deployment are deployed in idling-resource amount satisfaction according to deployment conditions This treats the resource requirement of application deployment, and the application disposed treats that application deployment do not possess the resource section of resource contention relation with this Point on.

It is understood that the application disposed on the resource node treats do not possess compatibility pass between application deployment with this System, or possess binding relationship.

It is exemplary, when 3 exemplary application C1 in dispatching device application deployment C clusters, using C2 and application C3 when, Dispatching device is needed according to deployment conditions, will be applied C1, is deployed in the application disposed respectively with answering using C2 and application C3 Do not possess with C on three different resource nodes of resource contention relation.As shown in figure 5, dispatching device according to deployment conditions with And affine sexual intercourse, C1 will be applied, be deployed in resource node 1, resource node 2 and resource section respectively using C2 and using C3 On point 3.

In one example, dispatching device can first judge multiple resource sections before disposing each and treating application deployment With the presence or absence of the resource node for the deployment request for meeting to treat application deployment in point, i.e., application deployment are treated with the presence or absence of deploying this Afterwards, the resource node of SLA conditions is preset corresponding to SLA parameters satisfaction.

If dispatching device, which determines to be not present in the plurality of resource node, deploys this after application deployment, SLA parameters meet The resource node of corresponding default SLA conditions, the certain applications that the dispatching device can be to having disposed on the plurality of resource node Redeployed, to cause the idling-resource amount of the part resource node in the plurality of resource node to increase, and then the scheduling This can be treated that application deployment are deployed in idle stock number in the part resource node and meet that this treats portion by device according to deployment conditions On the resource node for affixing one's name to the resource requirement of application.

It is exemplary, when dispatching device is in application deployment A3, determine resource node 1 in resource pool, resource node 2 with And resource node 3, it can not meet the deployment request using A3.Dispatching device can to resource node 1, resource node 2 and The certain applications disposed on resource node 3 are redeployed.

Specifically, as shown in fig. 6, dispatching device determines that if the application C3 that will have been disposed on resource node 3 is deleted so that money The idling-resource amount increase of source node 3, resource node 3 then meets the deployment request using A3, and resource node 2 disclosure satisfy that and answer With C3 deployment request.So as to which dispatching device can be redeployed having been disposed on resource node 3 in resource node using C3 On 2.Wherein, disposed on resource node 3 using C3 be using C an example, dispatching device can be by resource node 3 The application C3 disposed is out of service and deletes, then one of the application deployment C new exemplary application C4 on resource node 2, It will be deployed in again using A3 on resource node 2.

In the example present, when the idling-resource amount of multiple resource nodes is unsatisfactory for when the deployment request of application deployment, It can be increased by way of redeploying the certain applications disposed on the plurality of resource node in the middle part of multiple resource nodes The idling-resource amount of point resource node, with enable presence in the part resource node this treat the deployment requirements of application deployment Resource node, so as to complete the deployment that this treats application deployment, further increase treat application deployment be deployed to power.

Based on the above-mentioned application scheduling method thereof as described in Fig. 3 to Fig. 6, the embodiment of the present invention provides a kind of preferably application and adjusted Degree method, it is specially：

After dispatching device receives set of applications, treated first in deployment conditions and set of applications between application deployment Affine sexual intercourse, to treating that application deployment are disposed in set of applications, including：Method shown in reference picture 4, will have and tie up That determines relation treats that application deployment are deployed on same resource node, the method shown in reference picture 5, by with resource contention relation Treat that application deployment are deployed on different resource nodes, and each treat that the SLA parameters of the resource node where application deployment meet The SLA conditions of the resource node.

If dispatching device can not meet to complete to treat application deployment in the case of deployment conditions and affine sexual intercourse at the same time Deployment, dispatching device then can ensure meet deployment conditions in the case of, preferentially determine whether to close according to compatibility System complete deployment, when according to affine sexual intercourse carry out deployment can not complete deployment when, then can not consider affine sexual intercourse.

Exemplary, as shown in fig. 7, working as dispatching device according to deployment conditions and affine sexual intercourse, will have in set of applications There is the application A1 of binding relationship and be deployed in using B1 on resource node 1, by the application A2 with binding relationship and application B2 portions After administration is on resource node 2, for the remaining one group application deployment for the treatment of with binding relationship in set of applications, i.e., using A3 and Using B3, dispatching device determine in this resource node be not present be capable of A3 and application B3 deployment request resource node (i.e. Idling-resource amount meets the total resources demand using A3 and application B3, and after deploying using A3 and application B3, SLA parameters Disclosure satisfy that the resource node of corresponding SLA conditions), then, dispatching device can will apply A3 and using B3 according to deployment conditions Separately deployment, for example, will be deployed in using A3 on resource node 4, deploy the SLA parameters satisfaction using A3 resource node 4 It the SLA conditions of resource node 4, will be deployed in using B3 on resource node 5, and deploy the SLA parameters of the resource node 5 using B3 Meet the SLA conditions of resource node 5.

When dispatching device is according to deployment conditions and affine sexual intercourse, the exemplary application C1 in application C clusters is deployed in money On source node 6, after being deployed in using C2 on resource node 7, for a remaining exemplary application C3 in application C clusters, if Dispatching device determines that the resource node that disclosure satisfy that the deployment request using C3 is not present in the plurality of resource node, then, adjust Resource requirement of the idling-resource amount satisfaction using C3 can will be deployed according to deployment conditions using C3 by spending device, and in portion Affix one's name to using after C3, SLA parameters disclosure satisfy that on the resource node 7 of corresponding SLA conditions.

Further, if dispatching device can not be completed to treat the deployment of application deployment in the case where meeting deployment conditions, Dispatching device then can be according to method as shown in Figure 6, and the application disposed to multiple node upper parts redeploys, with The idling-resource amount of the part resource node in the plurality of resource node is adjusted, further according to deployment conditions, this is treated into application deployment It is deployed on the part resource node and is disposed.

By the preferable application scheduling method thereof, dispatching device can realize the Dynamical Deployment for treating application deployment, be based on The SLA conditions of resource node, deployment scheme is neatly selected, complete to treat the deployment of application deployment, be no longer limited by fixation Affine sexual intercourse, so as to improve the success rate using deployment.

Further, the deployment conditions that dispatching device obtains can also include the tenant space belonging to each resource node.

In embodiments of the present invention, the set of applications that dispatching device receives also includes the tenant's letter for each treating application deployment Breath.Dispatching device to each tenant's information for treating application deployment by authenticating, to determine each to treat that application deployment request is visited The tenant space asked.

Exemplary, the dispatching device can include multi-tenant resource selection submodule, the multi-tenant resource selection submodule The integrated third-party multi-tenant management system of docking of block, it can treat that deployment should to each under the control of the multi-tenant management system Tenant's information is authenticated, it is determined that each tenant space treated application deployment request and accessed, and treat application deployment each The resource node that application deployment can access is treated in inquiry in the tenant space for asking to access.And then dispatching device can be according to upper The scheme shown in Fig. 3 to Fig. 7 is stated, in each tenant space treated application deployment and be able to access that, application deployment, which are carried out, to be treated to this Deployment.

It is exemplary, as shown in figure 8, after dispatching device receives set of applications, by set of applications using A's Tenant's information, tenant's information using B and tenant's information using C are authenticated, it is determined that being accessed using A and application B requests Resource node in tenant space 1, the resource node in tenant space 2 is accessed using C access requests.Wherein, using A and application B has binding relationship, and dispatching device can exist application A and application B multiple example deployments according to method as shown in Figure 4 On resource node in tenant space 1.Be an application cluster using C, dispatching device can according to method as shown in Figure 5, Application C multiple examples are deployed on the different resource nodes in tenant space 2 respectively.

It is appreciated that when, treat that application deployment request accesses different tenant spaces corresponding to different tenant's information representations In resource node, i.e. tenant's information is different, and the resource node that application can be disposed is with regard to same, so as to effectively carry out treating portion Isolation between administration's application, realizes the security control during the deployment of application.

Optionally, deployment conditions provided in an embodiment of the present invention, other bars set according to deployment requirements can also be included Part, such as, it is allowed to the number upper limit, permission in application of the same resource node deployment with resource contention relation are in different resource The ratio upper limit, the particular ip address for filtered host, mark and the title of application of the deployment with binding relationship on node Deng on the other hand, the present invention is not limited.

Based on the method shown in above-mentioned Fig. 3 to Fig. 8, the embodiment of the present invention also provides a kind of monitoring method, can be to each The resource service condition of resource node is monitored, as shown in figure 9, this method can include：

S201, the corresponding resource node of each agent apparatus collection Monitoring Data.

Wherein, Monitoring Data includes the performance data for the application disposed on corresponding resource node, for example, on resource node The application disposed actual CPU usage operationally, actual memory utilization rate etc..

S202, each agent apparatus send the Monitoring Data of collection to detection means.

The sky of the resource node of S203, detection means according to corresponding to the Monitoring Data received calculates each Monitoring Data Not busy stock number.

It is exemplary, by taking resource node 1 as an example, run on resource node 1 using A1, using B1, using C1, using C2 And it is 20% using the actual CPU usages of A1 in the process of running, actual memory usage amount is 300M, application using D1 The actual CPU usages of B1 in the process of running are 30%, and actual memory usage amount is 180M, using C1 and in the process of running Actual CPU usage be 10%, actual memory usage amount is 160M, using the actual CPU usages of C2 in the process of running For 10%, actual memory usage amount is 160M, is 15% using the actual CPU usages of D2 in the process of running, actual memory Usage amount is 350M.Detection means by calculate using A1, using B1, using C1, using C2 and application D1 performance data, It can determine that resource node 1 currently there remains 15% idle cpu resource, 800M memory source.

S204, detection means send the idling-resource amount of each resource node calculated to dispatching device.

It is understood that each apply estimated resource requirement during deployment and the application in the process of running The resource of actual use may be not fully equal.It is likely larger than and was disposing using the resource actually used in the process of running The resource requirement estimated in journey, it is also possible to less than the resource requirement estimated during deployment.In embodiments of the present invention, detect Device calculates the current idling-resource amount of each resource node, sent by monitoring the Monitoring Data of each resource node in real time To dispatching device, to enable effective idling-resource amount of the dispatching device according to each resource node, application deployment are treated in deployment, So as to improve the resource utilization of resource node.

Further, based on the method described in Fig. 9, after above-mentioned S202, the present invention also provides a kind of dynamic adjustment side Method, as shown in Figure 10, the dynamic adjusting method can include：

S301, detection means according to Monitoring Data corresponding with each resource node and it is default with it is each disposed should With corresponding SLA conditions, at least one application to be adjusted in all applications for determining to have disposed on multiple resource nodes be present, treat The performance data of adjustment application is unsatisfactory for SLA conditions corresponding to the application to be adjusted.

In embodiments of the present invention, when detection means receives the monitoring for the corresponding resource node that each agent apparatus is sent After data, the application disposed on corresponding resource node can be monitored according to Monitoring Data.

Wherein, Monitoring Data includes the performance data of each application disposed on corresponding resource node, the performance data The application disposed including this actual CPU usage operationally, actual memory utilization rate, in addition to the application disposed The error rate that business datum receives, and per second establish call number (Call Attempts Per Second, CAPS) etc..

SLA conditions corresponding to the application disposed can include the SLA conditions of the application disposed itself, such as industry The upper limit, the CAPS upper limit etc. of the error rate of business data receiver, the SLA conditions can also include the application place resource disposed The SLA conditions of node, for example, on the resource node CPU usage of every kind of application the upper limit, the memory usage of every kind of application The upper limit etc..

Exemplary, it is assumed that include using SLA conditions corresponding to A1：Using A1 business datum receive error rate it is upper 5% is limited to, the upper limit using A1 and application B1 total CPU usage is 20%.Include using A1 performance data：Business datum The error rate 10% of reception, using the actual CPU usages of A1 in the process of running be 20%, i.e., using A1 performance data not Satisfaction applies SLA conditions corresponding to A1, then detection means can determine that using A1 be an application to be adjusted.

S302, detection means send scheduling decision message to dispatching device, and the scheduling decision message includes at least one The mark and adjust instruction of application to be adjusted, the adjust instruction are at least one to be adjusted for indicating that dispatching device redeploys this Using.

S303, for each application to be adjusted at least one application to be adjusted, dispatching device according to adjust instruction and Deployment conditions, the application to be adjusted is redeployed and meets another of the resource requirement of the application to be adjusted in idling-resource amount On resource node, wherein, the SLA parameters for another resource node disposed after the application to be adjusted meet the pre- of the resource node If SLA conditions.

In inventive embodiments, dispatching device redeploys to each application to be adjusted, can include：Dispatching device Will be out of service and delete on the resource node where application to be adjusted from the application originally to be adjusted, then according to deployment bar Part, a new example deployment of the application to be adjusted is met in idling-resource amount the application to be adjusted resource requirement it is another On one resource node.

Exemplary, it is assumed that detection means determines to be unsatisfactory for applying SLA conditions corresponding to A1 using A1 performance data, from And detection means sends application A1 mark and adjust instruction to dispatching device by scheduling decision message, dispatching device according to Adjust instruction indicates, it is determined that to being redeployed using A1.It is an example using A using A1, dispatching device is to application A1 is redeployed as the example for redeploying an application A, and deletes the exemplary application A1 disposed.Specifically, adjust Degree device is out of service by the application A1 being deployed on resource node 1, and is deleted from the resource node 1 and apply A1.Then exist Application A example is redeployed on another resource node in addition to resource node 1, for example, on the top of resource node 8 Using an A new exemplary application A4, the idling-resource amount of the resource node 8 meets the resource requirement using A, and disposes for administration The SLA parameters of resource node 8 after application A4 meet the SLA conditions of the resource node 8.

It is understood that the concrete mode that dispatching device is redeployed to each application to be adjusted, may refer to As Fig. 3 to Fig. 8 it is any shown in, dispatching device treats the method that application deployment are disposed, and here is omitted.

The method adjusted by dynamic provided in an embodiment of the present invention, detection means can monitor each disposed in real time The running status of application, and it is determined that performance data in the application disposed be present is unsatisfactory for the to be adjusted of corresponding SLA conditions Using when, indicate that dispatching device redeploys to the application to be adjusted in time, improve application in the process of running reliable Property, optimize the performance of system.

What deserves to be explained is scheme provided in an embodiment of the present invention, can be applied various large-scale in cloud computing In application cluster software deployment scene.

Exemplary, as shown in figure 11, a kind of Apache (a kind of Web server) container for implementing to provide for the present invention should With the network architecture with a kind of MySQL (Relational DBMS) container application, Apache container applications and MySQL hold Device applies the exploitation and deployment for being used for completing Web applications., can be to Apache containers using scheme provided in an embodiment of the present invention Disposed using with MySQL container applications.

During deployment Apache container applications and MySQL container applications, dispatching device can be preferentially according to resource Each group of Apache container application and MySQL container applications are deployed on same resource node by the SLA conditions of node, so that The data of MySQL container applications can be accessed directly in the machine by obtaining Apache container applications, to ensure that user can experience height The service effectiveness of performance low time delay.For Apache container applications and the cluster of MySQL containers, dispatching device can preferential basis The SLA conditions of resource node, different groups of Apache container applications and MySQL containers are deployed on different resource nodes, So as to avoid causing whole system not run due to the failure of a resource node.In Apache container applications and MySQL In the cluster of container, when dispatching device can not enter each group of Apache container application and MySQL containers according to affine sexual intercourse During row deployment, dispatching device can be preferentially by the part Apache containers in the cluster of Apache container applications and MySQL containers Disposed using with MySQL containers according to affine sexual intercourse, remaining Apache container applications and MySQL containers can be full In the case of the SLA conditions of sufficient resource node, deployment is completed, to ensure in the cluster of Apache container applications and MySQL containers Container application is deployed to power.

It is above-mentioned that mainly scheme provided in an embodiment of the present invention is described from the angle of interaction between each network element.Can With understanding, each network element, such as dispatching device, detection means, deployment managing device, agent apparatus etc. are above-mentioned in order to realize Function, it comprises perform the corresponding hardware configuration of each function and/or software module.Those skilled in the art should be easy to , it is realized that the unit and algorithm steps of each example described with reference to the embodiments described herein, the present invention can be with hardware Or the combining form of hardware and computer software is realized.Some function drives hardware with hardware or computer software actually Mode performs, application-specific and design constraint depending on technical scheme.Professional and technical personnel can be to each specific Application realize described function using distinct methods, but this realization is it is not considered that beyond the scope of this invention.

The embodiment of the present invention can be according to above method example to dispatching device, detection means, deployment managing device, agency Device etc. carries out the division of functional module, for example, can correspond to each function divides each functional module, can also by two or More than two functions are integrated in a processing module.Above-mentioned integrated module can both be realized in the form of hardware, It can be realized in the form of software function module.It should be noted that the division in the embodiment of the present invention to module is signal Property, only a kind of division of logic function, there can be other dividing mode when actually realizing.

In the case where dividing each functional module using corresponding each function, Figure 12 A show institute in above-described embodiment A kind of possible structural representation for the dispatching device being related to, dispatching device include：Receiving unit, acquiring unit and deployment are single Member.Receiving unit is used to support dispatching device to perform process S204 and figure in process S101, Fig. 9 in Fig. 3 as shown in Figure 3 Process S302 in 10；Acquiring unit is used to support dispatching device to perform the process S102 in Fig. 3；Deployment unit is used to support to adjust Degree device performs the process S303 in process S103 and Figure 10 in Fig. 3.Wherein, each step that above method embodiment is related to All related contents can quote corresponding function module function description, will not be repeated here.

In the case of using integrated unit, Figure 12 B show the one of dispatching device involved in above-described embodiment The possible structural representation of kind.Dispatching device includes：Processing module 1200 and communication module 1201.Processing module 1200 be used for pair The action of dispatching device is controlled management, for example, processing module 1200 is used to support dispatching device to perform the process in Fig. 3 The process S302 and S303 in process S204 and Figure 10 in S101-S103, Fig. 9, and/or for techniques described herein Other processes.Communication module 1201 is used for the communication for supporting dispatching device and other network entities, for example, with the work(shown in Fig. 1 Communication between energy module or network entity.Dispatching device can also include memory module 1202, for storing dispatching device Program code and data.

Wherein, processing module 1200 can be processor or controller, such as can be central processing unit (Central Processing Unit, CPU), general processor, digital signal processor (Digital Signal Processor, DSP), Application specific integrated circuit (Application-Specific Integrated Circuit, ASIC), field programmable gate array It is (Field Programmable Gate Array, FPGA) or other PLDs, transistor logic, hard Part part or its any combination.What it can realize or perform with reference to described by the disclosure of invention various exemplary patrols Collect square frame, module and circuit.The processor can also be the combination for realizing computing function, such as include one or more micro- places Manage device combination, combination of DSP and microprocessor etc..Communication module 1201 can be transceiver, transmission circuit or communication interface Deng.Memory module 1202 can be memory.

When processing module 1200 is processor, communication module 1201 is communication interface, when memory module 1202 is memory, Dispatching device involved by the embodiment of the present invention can be the dispatching device shown in Figure 12 C.

Refering to shown in Figure 12 C, the dispatching device includes：Processor 1210, communication interface 1211, memory 1212 and total Line 1213.Wherein, communication interface 1211, processor 1210 and memory 1212 are connected with each other by bus 1213；Bus 1213 can be Peripheral Component Interconnect standard (Peripheral Component Interconnect, PCI) bus or extension work Industry normal structure (Extended Industry Standard Architecture, EISA) bus etc..The bus can divide For address bus, data/address bus, controlling bus etc..For ease of representing, only represented in Figure 12 C with a thick line, it is not intended that Only a bus or a type of bus.

In the case where dividing each functional module using corresponding each function, Figure 13 A show institute in above-described embodiment A kind of possible structural representation for the detection means being related to, detection means include：Acquiring unit, computing unit, transmitting element And determining unit.Acquiring unit is used to support detection means to perform such as the process S202 in Fig. 9；Computing unit is used to support to examine Device is surveyed to perform such as the process S203 in Fig. 9；Transmitting element is used to support detection means to perform process S204 and figure in Fig. 9 Process S302 in 10；Determining unit is used to support detection means to perform such as the process S301 in Figure 10.Wherein, the above method is real The function description of corresponding function module can be quoted by applying all related contents for each step that example is related to, no longer superfluous herein State.

In the case of using integrated unit, Figure 13 B show the one of detection means involved in above-described embodiment The possible structural representation of kind.Detection means includes：Processing module 1300 and communication module 1301.Processing module 1300 be used for pair The action of detection means is controlled management, for example, processing module 1300 is used to support detection means to perform such as the process in Fig. 9 Process S301 and S302 in S202-S204 and Figure 10, and/or other processes for techniques described herein.Communication Module 1301 is used to support detection means and the communication of other network entities, such as real with the functional module shown in Fig. 1 or network Communication between body.Detection means can also include memory module 1302, for storing the program code and data of detection means.

Wherein, processing module 1300 can be processor or controller, such as can be CPU, general processor, DSP, ASIC, FPGA either other PLDs, transistor logic, hardware component or its any combination.It can be with Realize or perform the various exemplary logic blocks with reference to described by the disclosure of invention, module and circuit.The processing Device can also be the combination for realizing computing function, such as be combined comprising one or more microprocessors, the group of DSP and microprocessor Close etc..Communication module 1301 can be communication interface, transmission circuit or communication interface etc..Memory module 1302 can be storage Device.

When processing module 1300 is processor, communication module 1301 is communication interface, when memory module 1302 is memory, Detection means involved by the embodiment of the present invention can be the detection means shown in Figure 13 C.

Refering to shown in Figure 13 C, the detection means includes：Processor 1310, communication interface 1311, memory 1312 and total Line 1313.Wherein, communication interface 1311, processor 1310 and memory 1312 are connected with each other by bus 1313；Bus 1313 can be pci bus or eisa bus etc..The bus can be divided into address bus, data/address bus, controlling bus etc..For It is easy to represent, is only represented in Figure 13 C with a thick line, it is not intended that an only bus or a type of bus.

In the case where dividing each functional module using corresponding each function, Figure 14 A show institute in above-described embodiment A kind of possible structural representation for the deployment managing device being related to, deployment managing device include：Receiving unit and transmitting element. Receiving unit is used for the set of applications and deployment conditions for receiving user's input；Transmitting element is used to support deployment managing device to hold Process S101 in row such as Fig. 3.Wherein, all related contents for each step that above method embodiment is related to can be quoted The function description of corresponding function module, will not be repeated here.

In the case of using integrated unit, Figure 14 B show deployment managing device involved in above-described embodiment A kind of possible structural representation.Deployment managing device includes：Processing module 1400 and communication module 1401.Processing module 1400 are used to be controlled management to the action for disposing managing device, for example, processing module 1400 is used to support deployment management dress Put and perform such as the process S101 in Fig. 3 and/or other processes for techniques described herein.Communication module 1401 is used for branch Deployment managing device and the communication of other network entities are held, for example, it is logical between the functional module shown in Fig. 1 or network entity Letter.Deployment managing device can also include memory module 1402, for storing the program code and data of deployment managing device.

Wherein, processing module 1400 can be processor or controller, such as can be CPU, general processor, DSP, ASIC, FPGA either other PLDs, transistor logic, hardware component or its any combination.It can be with Realize or perform the various exemplary logic blocks with reference to described by the disclosure of invention, module and circuit.The processing Device can also be the combination for realizing computing function, such as be combined comprising one or more microprocessors, the group of DSP and microprocessor Close etc..Communication module 1401 can be communication interface, transmission circuit or communication interface etc..Memory module 1402 can be storage Device.

When processing module 1400 is processor, communication module 1401 is communication interface, when memory module 1402 is memory, Deployment managing device involved by the embodiment of the present invention can be the deployment managing device shown in Figure 14 C.

Refering to shown in Figure 14 C, the deployment managing device includes：Processor 1410, communication interface 1411, memory 1412 with And bus 1413.Wherein, communication interface 1411, processor 1410 and memory 1412 are connected with each other by bus 1413；Always Line 1413 can be pci bus or eisa bus etc..The bus can be divided into address bus, data/address bus, controlling bus etc.. For ease of representing, only represented in Figure 14 C with a thick line, it is not intended that an only bus or a type of bus.

The step of method or algorithm with reference to described by the disclosure of invention, can be realized in a manner of hardware, also may be used By be by computing device software instruction in a manner of realize.Software instruction can be made up of corresponding software module, software mould Block can be stored on random access memory (Random Access Memory, RAM), flash memory, read-only storage (Read Only Memory, ROM), Erasable Programmable Read Only Memory EPROM (Erasable Programmable ROM, EPROM), electricity can EPROM (Electrically EPROM, EEPROM), register, hard disk, mobile hard disk, read-only optical disc (CD-ROM) or in the storage medium of any other form well known in the art.A kind of exemplary storage medium is coupled to place Device is managed, so as to enable a processor to from the read information, and information can be write to the storage medium.Certainly, store Medium can also be the part of processor.Processor and storage medium can be located in ASIC.In addition, the ASIC can position In core network interface equipment.Certainly, processor and storage medium can also be present in core network interface as discrete assembly and set In standby.

Those skilled in the art are it will be appreciated that in said one or multiple examples, work(described in the invention It is able to can be realized with hardware, software, firmware or their any combination.When implemented in software, can be by these functions It is stored in computer-readable medium or is transmitted as one or more instructions on computer-readable medium or code. Computer-readable medium includes computer-readable storage medium and communication media, and wherein communication media includes being easy to from a place to another Any medium of one place transmission computer program.It is any that storage medium can be that universal or special computer can access Usable medium.

Above-described embodiment, the purpose of the present invention, technical scheme and beneficial effect are carried out further Describe in detail, should be understood that the embodiment that the foregoing is only the present invention, be not intended to limit the present invention Protection domain, all any modification, equivalent substitution and improvements on the basis of technical scheme, done etc., all should It is included within protection scope of the present invention.

Claims (18)

1. a kind of application scheduling method thereof, it is characterised in that methods described includes：

Dispatching device receives set of applications from deployment managing device, and the set of applications includes each treating in the set of applications Corresponding relation between application deployment and the resource requirement for treating application deployment；

The dispatching device obtains the idling-resource amount of each resource node and default deployment conditions, institute in multiple resource nodes Stating deployment conditions includes the preset service level protocol SLA conditions of each resource node；

Application deployment are each treated for described, the dispatching device treats that application deployment are disposed according to the deployment conditions by described On the resource node of resource requirement for treating application deployment described in meeting in idling-resource amount, wherein, application deployment are treated described in deployment The SLA parameters of the resource node afterwards meet the default SLA conditions of the resource node.

2. according to the method for claim 1, it is characterised in that the dispatching device obtains each to be provided in multiple resource nodes The idling-resource amount of source node, including：

The dispatching device receives the idling-resource amount of each resource node from detection means.

3. method according to claim 1 or 2, it is characterised in that the set of applications also includes in the set of applications Treat the affine sexual intercourse between application deployment, the affine sexual intercourse includes treating between application deployment in the set of applications Binding relationship or resource contention relation, have binding relationship in the set of applications treats that application deployment treat that deployment should for one group With,

Application deployment are each treated for described, the dispatching device treats that application deployment are disposed according to the deployment conditions by described On the resource node of resource requirement for treating application deployment described in meeting in idling-resource amount, wherein, application deployment are treated described in deployment The SLA parameters of the resource node afterwards meet the default SLA conditions of the resource node, including：

Application deployment are treated for each group in the set of applications, the dispatching device is according to the deployment conditions, by described in One group is treated that application deployment are deployed in idling-resource amount and met on the resource node of the described one group total resources demand for treating application deployment, Wherein, the SLA parameters for disposing one group of resource node after application deployment meet the default SLA of the resource node Condition.

4. according to the method for claim 3, it is characterised in that if the dispatching device according to the deployment conditions, by institute State one group and treat that application deployment are deployed in the resource node that idling-resource amount meets the described one group total resources demand for treating application deployment There is the application disposed before upper, on the multiple resource node,

The dispatching device treats that application deployment are deployed in described in idling-resource amount satisfaction according to the deployment conditions by described one group On the resource node of one group of total resources demand for treating application deployment, including：

The dispatching device treats that application deployment are deployed in described in idling-resource amount satisfaction according to the deployment conditions by described one group One group of total resources demand for treating application deployment, and with described one group to treat that application deployment do not possess the resource competing for the application disposed Strive on the resource node of relation.

5. method according to claim 1 or 2, it is characterised in that the set of applications also includes in the set of applications The affine sexual intercourse of application deployment is treated, the affine sexual intercourse includes treating that the resource between application deployment is competing in the set of applications Strive relation,

If the dispatching device according to the deployment conditions, by it is described treat application deployment be deployed in idling-resource amount meet described in treat There is the application disposed before on the resource node of the resource requirement of application deployment, on the multiple resource node,

The dispatching device according to the deployment conditions, by it is described treat application deployment be deployed in idling-resource amount meet described in treat portion On the resource node for affixing one's name to the resource requirement of application, including：

The dispatching device according to the deployment conditions, by it is described treat application deployment be deployed in idling-resource amount meet described in treat portion Affix one's name to the resource requirement of application, and the application disposed and the resource section treated application deployment and do not possess the resource contention relation Point on.

6. according to the method described in claim any one of 1-5, it is characterised in that the dispatching device is according to the deployment bar Part, by it is described treat application deployment be deployed in idling-resource amount meet described in treat application deployment resource requirement resource node on it Before, methods described also includes：

The dispatching device determines to be not present in the multiple resource node described in deployment after application deployment, and SLA parameters meet The resource node of corresponding default SLA conditions；

The dispatching device redeploys to the certain applications disposed on the multiple resource node, described more to cause The idling-resource amount increase of part resource node in individual resource node；

The dispatching device according to the deployment conditions, by it is described treat application deployment be deployed in idling-resource amount meet described in treat portion On the resource node for affixing one's name to the resource requirement of application, including：

The dispatching device treats that application deployment are deployed in the free time in the part resource node according to the deployment conditions by described On the resource node of resource requirement that application deployment are treated described in stock number satisfaction.

7. according to the method described in claim any one of 2-6, it is characterised in that methods described also includes：

The dispatching device receives the scheduling decision message that the detection means is sent, and the scheduling decision message includes adjustment Instruction and the mark of at least one application to be adjusted, the adjust instruction be used to indicating the dispatching device redeploy it is described extremely At least one application to be adjusted of the mark instruction of a few application to be adjusted；

For each application to be adjusted at least one application to be adjusted, the dispatching device is according to the adjust instruction With the deployment conditions, the application to be adjusted is redeployed to the resource need for meeting the application to be adjusted in idling-resource amount On another resource node asked, wherein, the SLA parameters for another resource node disposed after the application to be adjusted expire The default SLA conditions of foot another resource node.

8. a kind of application scheduling method thereof, it is characterised in that methods described includes：

Detection means obtains Monitoring Data corresponding with each resource node in multiple resource nodes, and the Monitoring Data includes The performance data for the application disposed on the resource node；

The detection means calculates the idling-resource amount of the resource node according to the Monitoring Data；

The detection means sends the idling-resource amount of each resource node calculated, each resource to dispatching device The idling-resource amount of node is used for dispatching device deployment and treats application deployment.

9. according to the method for claim 8, it is characterised in that methods described also includes：

The detection means according to Monitoring Data corresponding with each resource node and it is default with it is each disposed should With corresponding service-level agreement SLA conditions, exist at least in all applications for determining to have disposed on the multiple resource node One application to be adjusted, the performance data of the application to be adjusted are unsatisfactory for SLA conditions corresponding to the application to be adjusted；

The detection means to the dispatching device send scheduling decision message, the scheduling decision message include described at least The mark and adjust instruction of one application to be adjusted, the adjust instruction be used to indicating the dispatching device redeploy it is described extremely A few application to be adjusted.

A kind of 10. dispatching device, it is characterised in that including：

Receiving unit, for receiving set of applications from deployment managing device, the set of applications is included in the set of applications Each corresponding relation treated between application deployment and the resource requirement for treating application deployment；

Acquiring unit, for obtaining the idling-resource amount of each resource node and default deployment conditions in multiple resource nodes, The deployment conditions include the preset service level protocol SLA conditions of each resource node；

Deployment unit, for each treating application deployment for described, according to the deployment conditions, treat that application deployment are disposed by described On the resource node of resource requirement for treating application deployment described in meeting in idling-resource amount, wherein, application deployment are treated described in deployment The SLA parameters of the resource node afterwards meet the default SLA conditions of the resource node.

11. dispatching device according to claim 10, it is characterised in that

When the acquiring unit obtains the idling-resource amount of each resource node in multiple resource nodes, specifically include：It is described to obtain Unit is taken to receive the idling-resource amount of each resource node from detection means.

12. the dispatching device according to claim 10 or 11, it is characterised in that the receiving unit receive it is described should Also included treating the affine sexual intercourse between application deployment in the set of applications with set, the affine sexual intercourse includes described answer With the binding relationship treated between application deployment or resource contention relation in set, there is binding relationship in the set of applications Treat that application deployment treat application deployment for one group,

The deployment unit, specifically for treating application deployment for each group in the set of applications, portion is treated by described one group Administration's application is deployed in idling-resource amount and met on the resource node of the described one group total resources demand for treating application deployment, wherein, portion The SLA parameters for affixing one's name to the described one group resource node after application deployment meet the default SLA conditions of the resource node.

13. dispatching device according to claim 12, it is characterised in that if the deployment unit is according to the deployment bar Part, treat that application deployment are deployed in the money that idling-resource amount meets the described one group total resources demand for treating application deployment by described one group There is the application disposed before on source node, on the multiple resource node,

The deployment unit treats that application deployment are deployed in described in idling-resource amount satisfaction according to the deployment conditions by described one group On the resource node of one group of total resources demand for treating application deployment, specifically include：

The deployment unit treats that application deployment are deployed in described in idling-resource amount satisfaction according to the deployment conditions by described one group One group of total resources demand for treating application deployment, and with described one group to treat that application deployment do not possess the resource competing for the application disposed Strive on the resource node of relation.

14. the dispatching device according to claim 10 or 11, it is characterised in that the receiving unit receive it is described should Also include the affine sexual intercourse that application deployment are treated in the set of applications with set, the affine sexual intercourse includes the application collection The resource contention relation between application deployment is treated in conjunction,

If the deployment unit according to the deployment conditions, by it is described treat application deployment be deployed in idling-resource amount meet described in treat There is the application disposed before on the resource node of the resource requirement of application deployment, on the multiple resource node,

The deployment unit according to the deployment conditions, by it is described treat application deployment be deployed in idling-resource amount meet described in treat portion On the resource node for affixing one's name to the resource requirement of application, specifically include：

The deployment unit according to the deployment conditions, by it is described treat application deployment be deployed in idling-resource amount meet described in treat portion Affix one's name to the resource requirement of application, and the application disposed and the resource section treated application deployment and do not possess the resource contention relation Point on.

15. according to the dispatching device described in claim any one of 10-14, it is characterised in that

The deployment unit, it is additionally operable to, according to the deployment conditions, treat that application deployment are deployed in idling-resource amount and expired by described Before treating described in foot on the resource node of the resource requirement of application deployment, determine that deployment institute is not present in the multiple resource node State after application deployment, the resource node of SLA conditions is preset corresponding to SLA parameters satisfaction；And on the multiple resource node The certain applications disposed are redeployed, to cause the idling-resource of the part resource node in the multiple resource node Amount increase；

The deployment unit according to the deployment conditions, by it is described treat application deployment be deployed in idling-resource amount meet described in treat portion On the resource node for affixing one's name to the resource requirement of application, specifically include：

The deployment unit treats that application deployment are deployed in the free time in the part resource node according to the deployment conditions by described On the resource node of resource requirement that application deployment are treated described in stock number satisfaction.

16. according to the dispatching device described in claim any one of 11-15, it is characterised in that

The receiving unit, is additionally operable to receive the scheduling decision message that the detection means is sent, in the scheduling decision message Mark including adjust instruction and at least one application to be adjusted, the adjust instruction are used to indicate deployment unit portion again Affix one's name at least one application to be adjusted of the mark instruction of at least one application to be adjusted；

The deployment unit, it is additionally operable to for each application to be adjusted at least one application to be adjusted, according to described Adjust instruction and the deployment conditions, the application to be adjusted is redeployed and meets the application to be adjusted in idling-resource amount Resource requirement another resource node on, wherein, another resource node disposed after the application to be adjusted SLA parameters meet the default SLA conditions of another resource node.

A kind of 17. detection means, it is characterised in that including：

Acquiring unit, for obtaining Monitoring Data corresponding with each resource node in multiple resource nodes, the monitoring number According to the performance data of the application including having been disposed on the resource node；

Computing unit, the Monitoring Data for being obtained according to the acquiring unit calculate the idling-resource of the resource node Amount；

Transmitting element, for sending the idling-resource for each resource node that the computing unit calculates to dispatching device Amount, the idling-resource amount of each resource node are used for dispatching device deployment and treat application deployment.

18. detection means according to claim 17, it is characterised in that the detection means also includes determining unit,

The determining unit, for obtained according to the acquiring unit Monitoring Data corresponding with each resource node and Default service-level agreement SLA conditions corresponding with each application disposed, determine on the multiple resource node portion At least one application to be adjusted be present in all applications of administration, the performance data of the application to be adjusted is unsatisfactory for described to be adjusted Using corresponding SLA conditions；

The transmitting element, it is additionally operable to send scheduling decision message to the dispatching device, the scheduling decision message includes The mark and adjust instruction of at least one application to be adjusted, the adjust instruction are used to indicate the dispatching device again Dispose at least one application to be adjusted.