WO2013082742A1

WO2013082742A1 - Resource scheduling method, device and system

Info

Publication number: WO2013082742A1
Application number: PCT/CN2011/083472
Authority: WO
Inventors: 张妮
Original assignee: 华为技术有限公司
Priority date: 2011-12-05
Filing date: 2011-12-05
Publication date: 2013-06-13
Also published as: CN103503404A

Abstract

The present invention relates to a resource scheduling method, device and system. The method includes: acquiring the traffic of each service currently running in an application, the application being deployed in a virtual machine cluster of an internal cloud platform; determining the number of resources to be extended by the virtual machine cluster according to a pre-acquired process extension policy of the application when the traffic of each service currently running in the application meets a preset condition; sending a virtual machine application request to an external cloud platform according to the determined number of resources when it is determined to apply for resource scheduling to the external cloud platform; and after receiving virtual machine information returned by the external cloud platform, deploying a process of the application for an external virtual machine to which the virtual machine information points. The present invention ensures the quality of service of an application by scheduling external cloud resources to complement the internal cloud resources.

Description

Resource scheduling method, device and system

Embodiments of the present invention relate to network communication technologies, and in particular, to a resource scheduling method, apparatus, and system. Background technique

The cloud is an emerging network-based distributed technology that encapsulates specific technical details through the cloud and then provides services by providing users with a relatively simple interface. Many enterprises currently deploy private cloud platforms to migrate applications to their own private cloud platforms. Such as China Telecom, mobile and other enterprises, their applications include Enterprise Resource Planning (ERP), billing, online business hall, billing statistics analysis, marketing analysis, portals, etc., they will deploy their respective applications. Go to the company's own private cloud platform.

Typically, the enterprise calculates the amount of resources required based on the traffic of a specific application, thereby deploying resources for the application within the enterprise private cloud platform. The existing technology for deploying resources for an application in the enterprise's private cloud platform is to calculate the virtual machine cluster of the internal cloud platform by calculating the resources required for each time zone application and then applying the maximum required resources according to each time zone application. Perform static configuration of resources.

However, the business volume of applications in most enterprises is not stable in different time zones. For example: The business volume of applications in most enterprises has surged at the end of the year, but it has been stable in other time zones. If the enterprise deploys the application based on the maximum number of resources calculated in each time zone, in the most time zone where the traffic is low, more resources in the configured resources are idle, resulting in waste of resources. If the enterprise does not deploy resources based on the maximum amount of traffic calculated by the application in each time zone, when the application peaks in traffic, the pre-deployed resources will be insufficient to provide the corresponding services. Summary of the invention

The embodiment of the invention provides a resource scheduling method, device and system for supplementing internal cloud resources by scheduling external cloud resources, thereby ensuring service quality of the application.

The embodiment of the invention provides a resource scheduling method, including:

Obtaining the traffic of each process currently running in the application, where the application is deployed in a virtual machine cluster of the internal cloud platform; When the traffic of each process currently running in the application meets the preset condition, determine the number of resources that the virtual machine cluster needs to expand according to the pre-obtained process scaling policy of the application; and determine to apply to the external cloud platform. When scheduling resources, sending a virtual machine application request to the external cloud platform according to the determined number of resources;

After receiving the virtual machine information returned by the external cloud platform, the process of the application is deployed to an external virtual machine pointed to by the virtual machine information.

The embodiment of the invention further provides a resource scheduling device, including:

The information is obtained by using the ear block to obtain the traffic of each process currently running in the application, and the application is deployed in the virtual machine cluster of the internal cloud platform;

a resource number determining module, configured to determine, according to the pre-obtained process scaling policy of the application, the number of resources that the virtual machine cluster needs to expand, when the traffic of each process that is currently running in the application meets a preset condition;

a resource scheduling application module, configured to send a virtual machine application request to the external cloud platform according to the determined number of resources when determining to apply for scheduling resources to the external cloud platform;

The application process migration module is configured to deploy the process of the application to an external virtual machine pointed to by the virtual machine information after receiving the virtual machine information returned by the external cloud platform.

The embodiment of the invention further provides a resource scheduling system, including:

Internal cloud platform and external cloud platform;

The resource scheduling device is connected to the internal cloud platform and the external cloud platform.

The resource scheduling method, device, and system provided by the embodiments of the present invention may perform dynamic resource scheduling according to the process of the process for the service volume of each application of the deployed virtual machine cluster. In the dynamic scheduling process of the resource, if the idle resource of the internal cloud platform cannot meet the service growth requirement of the application, the external cloud platform may apply for resources for resource supplementation, and the process of the external virtual machine is deployed, thereby ensuring The quality of service applied. DRAWINGS

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the embodiments or the description of the prior art will be briefly described below. Obviously, the drawings in the following description are only It is some embodiments of the present invention, and those of ordinary skill in the art, Other drawings may also be obtained from these drawings without the inventive labor. FIG. 1 is a flowchart of a resource scheduling method according to Embodiment 1 of the present invention;

2 is a schematic structural diagram of a resource scheduling system according to Embodiment 2 of the present invention;

FIG. 3 is an example of an internal structure of the internal cloud platform in FIG. 2;

4 is a signaling interaction diagram of a resource scheduling method according to Embodiment 3 of the present invention;

FIG. 5 is a signaling interaction diagram of a resource scheduling method according to Embodiment 4 of the present invention;

FIG. 6 is a schematic structural diagram of a resource scheduling apparatus according to Embodiment 5 of the present invention. detailed description

The technical solutions in the embodiments of the present invention are clearly and completely described in the following with reference to the accompanying drawings in the embodiments of the present invention. It is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments obtained by a person of ordinary skill in the art based on the embodiments of the present invention without the creative work are all within the scope of the present invention.

The serial numbers of the following embodiments of the present invention are merely for the description, and do not represent the advantages and disadvantages of the embodiments.

FIG. 1 is a flowchart of a resource scheduling method according to Embodiment 1 of the present invention. The method shown in Figure 1 includes:

Step 11: Obtain the traffic of each process currently running in the application, where the application is deployed in a virtual machine cluster of the internal cloud platform.

The internal cloud platform is pre-deployed with multiple virtual machine clusters; one virtual machine cluster includes multiple virtual machines; one virtual machine cluster can deploy one application or multiple applications; one application can include one or more processes; one process can be deployed at On one or more virtual machines.

After the application deployment of the internal cloud platform is completed, the real-time business load of the deployed applications in the virtual machine cluster in the internal cloud platform can be obtained. The real-time service load of an application may include: information about the currently running process in the application, the traffic volume of each process, and the like; the traffic of a certain process is information such as the number of user requests received by the process.

Step 12: When the traffic of each process currently running in the application meets the preset condition, determine the number of resources that the virtual machine cluster needs to expand according to the pre-obtained process scaling policy of the application.

The preset conditions described in this step can be set according to actual needs. For example, the preset condition may be: the sum of the traffic of each process of the application is greater than the first preset threshold, and the idle resources of the virtual machine cluster cannot be The traffic of the virtual machine cluster cannot be satisfied. These processes have increased business volume requirements. The idle resources of the virtual machine cluster may be acquired in advance.

If the traffic of each process currently running in the application meets the preset condition, the number of resources that the virtual machine cluster needs to expand is determined according to the pre-obtained process scaling policy of the application.

The application's process scaling policy can be pre-fetched, such as obtaining from an application deployment request provided by the user. The process scaling policy may include: increasing the number of processes of the application; in this case, the number of resources that the virtual machine cluster needs to expand is: increasing the number of resources required by the preset number of processes. Alternatively, the process scaling policy may include: increasing the number of resources of one or more processes currently running in the application; in this case, the number of resources that the virtual machine cluster needs to expand is: for each process currently running in the application. The total number of resources to be added.

Step 13: When determining to apply for scheduling resources to the external cloud platform, send a virtual machine application request to the external cloud platform according to the determined number of the resources.

This step may decide whether to apply for scheduling resources to the external cloud platform according to a predetermined policy.

Optionally, this step may be based on the idle resource status of the internal cloud platform, and the decision is to apply for scheduling resources to the external cloud platform or the internal cloud platform. For example, if the resources of the internal cloud platform are sufficient, and the idle resources of the internal cloud platform can provide the determined scheduling of the number of the resources, the determined resources may be scheduled for the virtual machine cluster in the idle resources of the internal cloud platform. number. If the resources of the internal cloud platform are insufficient, and the idle resources of the internal cloud platform cannot provide the determined scheduling of the number of the resources, the virtual machine application request may be sent to the external cloud platform according to the determined number of the resources, and used to the external cloud platform. Apply for scheduling resources.

Optionally, the step may further determine whether to apply for scheduling resources to the external cloud platform according to the traffic volume change trend of each process currently running in the application. If the traffic growth of each process currently running in the application is a short-term behavior, it is determined to apply for scheduling resources to the external cloud platform; if the traffic growth of each process currently running in the application is a long-term behavior, then determining to the external cloud platform Apply for scheduling resources.

Step 14: After receiving the virtual machine information returned by the external cloud platform, deploy the process of the application to an external virtual machine pointed to by the virtual machine information.

For ease of description, in the embodiment of the present invention: virtual environment deployment based on internal cloud platform The virtual machine is called an internal virtual machine; the virtual machine deployed based on the physical resources of the external cloud platform is called an external virtual machine.

After receiving the virtual machine application request, the external cloud platform deploys the external virtual machine according to the virtual machine application request and returns the virtual machine information; the virtual machine information includes: information such as the access address of the external virtual machine. After receiving the virtual machine information returned by the external cloud platform, the process of the application may be deployed to the external virtual machine according to the preset process migration policy.

The preset process migration policy is as follows: a process in which the frequency of interaction with other processes is lower than a preset threshold, that is, the first process, the internal virtual machine of the virtual machine cluster of the application is deployed from the internal cloud platform, and the internal virtual machine is migrated to the external virtual machine. on. After the first process is migrated to the external virtual machine, the internal cloud platform deploys the resources occupied by the first process released by the virtual machine cluster of the application, and can be used by other processes provided to the application to meet the service quality of the application. Claim. In addition, since the first process interacts with other processes less frequently, the migration of the first process to the external virtual machine does not increase the interaction between the processes, thereby having less impact on the performance and quality of the application. .

Alternatively, the preset process migration policy may also be: deploying all processes currently running in the application on the external virtual machine, and migrating part of the application's business load from the internal virtual machine of the virtual machine cluster to the external virtual machine. After this deployment, the application can load balance through internal virtual machines and external virtual machines, thus ensuring the service quality of the application.

The above two process migration strategies can be applied to deploy a single application in a virtual machine cluster, and can also be applied to deploy multiple applications in a virtual machine cluster.

In addition, if a plurality of applications are deployed in a virtual machine cluster, the preset process migration policy may also be, for example, one of a plurality of applications deployed in the virtual machine cluster, that is, the first application and the currently running in the application. All processes migrate from the internal virtual machines of the virtual machine cluster to the external virtual machines. After the deployment, the resources occupied by the first application released by the virtual machine cluster can be used to provide services for other applications deployed in the virtual machine cluster to ensure the service quality of other applications, and the service and quality of service of the first application may be External virtual machine is provided.

The resource scheduling method provided in this embodiment may dynamically schedule resources according to the process of the process for the service volume of each process of the application deployed in the virtual machine cluster. In the dynamic scheduling process of the resource, if the idle resource of the internal cloud platform cannot meet the service growth requirement of the application, the external cloud platform may apply for resources for resource supplementation, and the process of the external virtual machine is deployed, thereby ensuring The quality of service applied. FIG. 2 is a schematic structural diagram of a resource scheduling system according to Embodiment 2 of the present invention. The resource scheduling system shown in FIG. 2 includes: an internal cloud platform 21, a resource scheduling device 22, and an external cloud platform 23; the resource scheduling device 22 is communicatively coupled to the internal cloud platform 21 and the external cloud platform 23, respectively.

The internal structure of the internal cloud platform 21 can be as shown in FIG. 3. Specifically, the internal cloud platform can deploy multiple virtual machine clusters based on physical resources of the internal cloud platform, and each virtual machine cluster includes one or more virtual machines, one virtual A cluster consists of multiple virtual machines; a virtual machine cluster can deploy one application or multiple applications; one application can include multiple processes; and one process can be deployed on one or more virtual machines. The internal cloud platform 21 can be a private cloud platform; or the internal cloud platform 21 can be a public cloud platform, such as the Amazon Elastic Compute Cloud (Amazon EC2) platform.

The resource scheduling device 22 is responsible for requesting resources such as external virtual machines from the external cloud platform when the service load of an application deployed on the internal cloud platform meets the preset conditions and the idle resources of the internal cloud platform cannot meet the application service quality.

The external cloud platform 23 may be a private cloud platform; or the external cloud platform 23 may be a public cloud platform, such as Amazon EC2, IBM Capacity on Demand (IBM COD), HP Enterprise service, etc. cloud platform.

FIG. 4 is a signaling interaction diagram of a resource scheduling method according to Embodiment 3 of the present invention. This embodiment is described by taking the resource scheduling system shown in FIG. 2 and FIG. 3 as an example. In this embodiment, a virtual machine cluster is taken as an example, and an application is deployed in the virtual machine cluster. As shown in Figure 2-4, the resource scheduling method provided in this embodiment includes:

Step 41: The resource scheduling device collects information about the virtual machine cluster information and the application service quality. The VM cluster information includes: the identifier of the deployed VM cluster, the application of each VM cluster deployment, and the resource usage of each VM cluster.

The application service quality related information includes: application corresponding virtual machine cluster, application load status, and the like; application load conditions include: information about the currently running process in the application, and the traffic volume of each process.

Step 42: The resource scheduling device determines, according to the application service quality related information of the application, whether the traffic volume of each process currently running in the application meets the preset condition. If yes, go to step 43; otherwise, go to step 41.

The preset conditions described in this step can be set according to actual needs. For example, the preset condition can be: The sum of the traffic of each process is greater than the first preset threshold, and the idle resources of the virtual machine cluster cannot meet the traffic growth requirement of each process of the application; or, the preset condition may also be: The traffic of each of the multiple processes is greater than the second preset threshold, and the idle resources of the virtual machine cluster cannot meet the traffic growth requirements of the processes. The idle resources of the virtual machine cluster may be acquired in advance.

Step 43: The resource scheduling device acquires a process scaling policy of the application, and determines, according to the process scaling policy, the number of resources that the virtual machine cluster needs to expand.

Step 44: The resource scheduling device may apply for the resource to the internal cloud platform or the external cloud platform according to the application service quality related information of the application, and if it is determined to apply for the resource to the internal cloud platform, perform step 45; if it is determined to the external cloud platform To apply for resources, go to Step 49.

The resource scheduling device may analyze the traffic volume change trend of each application process according to the application service quality related information of the application, and determine whether to apply for scheduling resources to the internal cloud platform or the external cloud platform according to the analysis result. If the analysis results indicate that the business load of each process of the application is a long-term phenomenon, resources can be applied to the internal cloud platform. If the analysis result indicates that the service load of each process of the application is a short-term phenomenon, resources can be applied to the external cloud platform, thereby avoiding the problem of low resource utilization caused by the blind expansion of the physical resources of the internal cloud platform.

Step 45: The resource scheduling device sends a resource request request to the internal cloud platform for requesting the determined number of resources for the virtual machine cluster scheduling.

The resource application request may include: determining the number of the resources, the identifier of the virtual machine cluster, and the like. Step 46: The internal cloud platform receives the resource application request, determines whether the determined scheduling of the number of resources can be provided, and if yes, performs step 47; otherwise, performs step 49.

The internal cloud platform determines the number of idle resources and the number of resources requested by the resource application request. Is it possible to provide this resource scheduling?

If the resources of the internal cloud platform are sufficient, for example, the idle resources of the internal cloud platform are greater than or equal to the determined number of the resources, the internal cloud platform may provide the determined scheduling of the number of resources.

If the resources of the internal cloud platform are insufficient, for example, the idle resources of the internal cloud platform are smaller than the determined number of the resources, the internal cloud platform cannot provide the determined scheduling of the number of resources.

Step 47: The internal cloud platform allocates physical resources matching the determined number of the resources in the idle resource, and sends the information of the allocated physical resources to the resource scheduling device.

If the internal cloud platform determines a schedule that can provide the determined number of the resources, allocate, in an idle resource of the internal cloud platform, a physical resource that matches the determined number of the resources, and send the information of the allocated physical resource to the resource. Dispatching device.

Step 48: The resource scheduling device performs process scaling processing according to the foregoing process scaling policy based on the physical resources allocated by the internal cloud platform.

The resource scheduling device process scaling processing, for example, the resource scheduling device adds a preset number of processes to the virtual machine cluster, or the resource scheduling device increases the memory resources available to the one or more processes that the application needs to increase resources.

Step 49: The resource scheduling apparatus sends a virtual machine application request to the external cloud platform according to the determined number of the resources.

The virtual machine application request includes: the specification and quantity of the virtual machine, the image identifier that the virtual machine needs to be installed, the virtual machine cluster information corresponding to the virtual machine, and the external cloud platform identifier.

The file corresponding to the image ID that the virtual machine needs to be installed can be uploaded to the external cloud platform through the interface of the external cloud platform.

The external cloud platform can be a private cloud platform; or the external cloud platform can be a public cloud platform, such as

Cloud platforms such as Amazon EC2, IBM COD, and HP Enterprise Business.

The resource scheduling device specifically requests from which external cloud platform to apply for resources, which can be pre-configured by the administrator. Alternatively, the resource scheduling device may automatically decide which external cloud platform to apply for the resource according to the predetermined condition, for example, according to the specification information of the virtual machine that can be provided by the external cloud platform acquired in advance, the same specification of the virtual machine specification required for the current resource scheduling can be provided. Or an external cloud platform close to the specification, as a target external cloud platform.

Step 410: The external cloud platform receives the virtual machine application request, and deploys the external virtual machine according to the virtual machine application request in the physical resource of the external cloud platform. The virtual machine information is returned to the resource scheduling device. Step 411: The external cloud platform returns virtual machine information to the resource scheduling device.

The virtual machine information may include: an access address of an external virtual machine, and the like.

Step 412: The resource scheduling device stores related information of the external virtual machine.

The resource scheduling device may include: an access address of the external virtual machine, a virtual machine cluster information corresponding to the external virtual machine, a resource attribute of the external virtual machine, and the like; wherein the resource attribute of the external virtual machine may be: an external cloud Resources.

Step 413: The resource scheduling device performs process scaling processing according to the foregoing process scaling policy, and deploys the applied process according to the preset process migration policy on the external virtual machine pointed to by the received virtual machine information.

In this embodiment, after the resource scheduling apparatus completes the current resource scheduling, the foregoing steps are repeatedly performed, and the dynamic scheduling of the resources may be performed according to the real-time service load of the deployed application of the virtual machine cluster. In the dynamic scheduling process of resources, it is possible to determine whether to allocate resources to the internal cloud platform or to the external cloud platform according to the service load trend of the application, thereby improving the flexibility of resource scheduling. In addition, when the idle resources of the internal cloud platform cannot meet the service growth requirements of the application, the external cloud platform can be applied for resources for resource supplementation, and the flexible process migration strategy is used to deploy the external virtual machine process, thereby ensuring the protection. The quality of service applied. FIG. 5 is a signaling interaction diagram of a resource scheduling method according to Embodiment 4 of the present invention. This embodiment is described by taking the resource scheduling system shown in FIG. 2 and FIG. 3 as an example. In this embodiment, a virtual machine cluster is taken as an example, and the virtual machine cluster is deployed with multiple applications. As shown in FIG. 2, FIG. 3 and FIG. 5, the resource scheduling method provided in this embodiment includes:

Step 51: The resource scheduling device collects information about the application service quality of each application deployed by the virtual machine cluster information and the virtual machine cluster.

Step 52: The resource scheduling device determines, according to the application service quality related information of each application in the virtual machine cluster, whether the traffic volume of each process currently running in each application meets the preset condition. If yes, go to step 53; otherwise, perform the step. 51.

Step 53 - Step 512: The application involved in the step 53 - 512 is an application in which the traffic of each process currently running meets the preset condition in the application deployed in the virtual machine cluster, and other technical solutions and steps 43 - 412 Similar, it will not be repeated here.

Step 513: The resource scheduling device deploys the process of the application according to the preset process migration policy on the external virtual machine pointed to by the received virtual machine information.

Or, one of the multiple applications deployed in the virtual machine cluster, such as the first application and all processes currently running in the application, all migrate from the internal virtual machine of the virtual machine cluster to the external virtual Planned. After the deployment, the resources occupied by the first application released by the virtual machine cluster can be used to provide services for other applications deployed in the virtual machine cluster to ensure the service quality of other applications, and the service and quality of service of the first application may be External virtual machine is provided.

In this embodiment, the real-time service load of each deployed application of the virtual machine cluster is dynamically scheduled in a scenario where multiple applications are deployed in a virtual machine cluster. In the dynamic scheduling process of resources, it is possible to determine whether to allocate resources to the internal cloud platform or to the external cloud platform according to the traffic load trend of an application, thereby improving the flexibility of resource scheduling. In addition, when the idle resources of the internal cloud platform cannot meet the service growth requirements of the application, the external cloud platform can be applied for resources for resource supplementation, and the flexible process migration strategy is used to deploy the external virtual machine process, thereby ensuring the protection. Quality of service for multiple applications deployed by this virtual machine cluster.

FIG. 6 is a schematic structural diagram of a resource scheduling apparatus according to Embodiment 5 of the present invention. The resource scheduling apparatus shown in FIG. 6 includes: an information obtaining module 61, a resource number determining module 62, a resource scheduling request module 63, and an application process migration module 64.

The information acquisition block 61 can be used to obtain the traffic of each process currently running in the application, and the application is deployed in the virtual machine cluster of the internal cloud platform;

The resource number determining module 62 is configured to determine, according to the pre-obtained process scaling policy of the application, the number of resources that the virtual machine cluster needs to expand, when the traffic of each process that is currently running in the application meets the preset condition;

The resource scheduling application module 63 may be configured to send a virtual machine application request to the external cloud platform according to the determined number of resources when determining to apply for scheduling resources to the external cloud platform;

The application process migration module 64 is configured to deploy the process of the application to an external virtual machine pointed to by the virtual machine information after receiving the virtual machine information returned by the external cloud platform.

Optionally, the process scaling policy includes: increasing the number of processes of the application; the number of resources that the virtual machine cluster needs to expand is: increasing the number of resources required by the preset number of processes. Or, optionally, the process scaling policy includes: increasing a number of resources of one or more processes currently running in the application; the number of resources that the virtual machine cluster needs to expand is: being currently running in the application The total amount of resources required for each process.

The application process migration module can perform process migration deployment according to the preset process migration policy. For example, if a virtual machine cluster is deployed with one or more applications, the application process migration module 64 is specifically configured to determine the first process according to the inter-process topology relationship of the application, and the first process is The internal virtual machine of the virtual machine cluster is migrated to the external virtual machine; the first process is a process that is currently running in the application and whose frequency of interaction with other processes is lower than a preset frequency.

Or,

If a virtual machine cluster is deployed with one or more applications, the application process migration module 64 is specifically configured to deploy each process currently running in the application on the external virtual machine, and load part of the service load of the application. The internal virtual machine of the virtual machine cluster is migrated to the external virtual machine.

If a virtual machine cluster is deployed with multiple applications, the application process migration module 64 is specifically configured to use the first application of the plurality of applications and the processes currently running in the first application from the internal of the virtual machine cluster. The virtual machine is migrated to the external virtual machine.

Further, the resource scheduling apparatus may further include: a scheduling decision module 65.

The scheduling decision module 65 may be configured to determine to apply for scheduling resources to the external cloud platform when the idle resources of the internal cloud platform cannot provide the determined scheduling of the number of resources. Alternatively, the scheduling decision module 65 may be configured to analyze a trend of the traffic volume of each process currently running in the application, and determine to apply for scheduling resources to the external cloud platform according to the analysis result.

In this embodiment, the resource scheduling apparatus can dynamically schedule resources according to the real-time service load of the deployed application of the virtual machine cluster. In the dynamic scheduling process of resources, it is possible to determine whether to allocate resources to the internal cloud platform or to the external cloud platform according to the service load trend of the application, thereby improving the flexibility of resource scheduling. In addition, when the idle resources of the internal cloud platform cannot meet the service growth requirements of the application, the external cloud platform can be applied for resources for resource supplementation, and the flexible process migration strategy is used to deploy the external virtual machine process, thereby ensuring the protection. The quality of service applied. For the working mechanism of the resource scheduling apparatus in this embodiment, refer to the corresponding descriptions of the corresponding embodiments in FIG. 1 and FIG. 4-5. The structure of the resource scheduling system in the hybrid cloud mode of the resource scheduling apparatus can be referred to the corresponding embodiment in FIG. 2-3. Record; no longer repeat here.

It will be understood by those of ordinary skill in the art that the drawings are only a schematic representation of one embodiment, and the modules or processes in the drawings are not necessarily required to practice the invention.

It will be understood by those skilled in the art that the modules in the apparatus in the embodiments may be distributed in the apparatus of the embodiment according to the embodiment, or may be correspondingly changed in one or more apparatuses different from the embodiment. The modules of the above embodiments may be combined into one module, or may be further split into a plurality of sub-modules.

One of ordinary skill in the art can understand that all or part of the steps of the above method embodiments are implemented. The foregoing program may be stored in a computer readable storage medium, and when executed, the program includes the steps of the foregoing method embodiment; and the foregoing storage medium includes: ROM, RAM , a variety of media that can store program code, such as a disk or an optical disk.

It should be noted that the above embodiments are only for explaining the technical solutions of the present invention, and are not intended to be limiting; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those skilled in the art that: The technical solutions described in the foregoing embodiments are modified, or the equivalents of the technical features are replaced by the equivalents of the technical solutions of the embodiments of the present invention.

Claims

Claim

A resource scheduling method, comprising:

Obtaining the traffic of each process currently running in the application, where the application is deployed in a virtual machine cluster of the internal cloud platform;

When the traffic of each process currently running in the application meets the preset condition, determine the number of resources that the virtual machine cluster needs to expand according to the pre-obtained process scaling policy of the application; and determine to apply to the external cloud platform. When scheduling resources, sending a virtual machine application request to the external cloud platform according to the determined number of resources;

2. The method of claim 1 wherein

The process scaling policy includes: increasing the number of processes of the application; the number of resources that the virtual machine cluster needs to expand is: increasing the number of resources required by the preset number of processes;

Or,

The process scaling policy includes: increasing the number of resources of one or more processes currently running in the application; the number of resources that the virtual machine cluster needs to expand is: required for each process currently running in the application The total number of resources.

The method according to claim 1 or 2, wherein the process of deploying the application to an external virtual machine pointed to by the virtual machine information comprises:

Determining a first process according to an inter-process topology relationship of the application, and migrating the first process from an internal virtual machine of the virtual machine cluster to the external virtual machine; the first process is in the application The currently running process that interacts with other processes less frequently than the preset frequency;

Or,

Deploying each process currently running in the application on the external virtual machine, and migrating part of the service load of the application from the internal virtual machine of the virtual machine cluster to the external virtual machine.

The method according to claim 1 or 2, wherein if the virtual machine cluster is deployed with a plurality of the applications, the process of the application is deployed to an external virtual machine pointed to by the virtual machine information, Includes:

The first application of the plurality of applications and the processes currently running in the first application are migrated from an internal virtual machine of the virtual machine cluster to the external virtual machine.

The method according to any one of claims 1-4, wherein the determining to apply for scheduling resources to the external cloud platform comprises:

When the idle resource of the internal cloud platform cannot provide the determined scheduling of the number of resources, the external cloud platform is requested to apply for scheduling resources;

Or,

The trend of the traffic volume of each process currently running in the application is analyzed, and the scheduling resource is requested to be applied to the external cloud platform according to the analysis result.

6. A resource scheduling device, comprising:

7. Apparatus according to claim 6 wherein:

Or,

8. Apparatus according to claim 6 or claim 7 wherein:

The application process migration module is specifically configured to determine a first process according to an inter-process topology relationship of the application, and migrate the first process from an internal virtual machine of the virtual machine cluster to the external virtual machine; The first process is a process that is currently running in the application and that interacts with other processes less frequently than a preset frequency;

or, The application process migration module is specifically configured to deploy each process currently running in the application on the external virtual machine, and migrate part of the service load of the application from the internal virtual machine of the virtual machine cluster to the Describe the external virtual machine.

The device according to claim 6 or 7, wherein the virtual machine cluster is deployed with a plurality of the applications;

The application process migration module is configured to migrate the first application of the plurality of applications and each process currently running in the first application from an internal virtual machine of the virtual machine cluster to the external virtual machine. .

The device according to any one of claims 6-9, wherein the device further comprises: a scheduling decision module;

The scheduling decision module is configured to: when the idle resource of the internal cloud platform cannot provide the determined scheduling of the number of resources, determine to apply for scheduling resources to the external cloud platform;

Or,

The scheduling decision module is configured to analyze a trend of the traffic volume of each process currently running in the application, and determine to apply for scheduling resources to the external cloud platform according to the analysis result.

11. A resource scheduling system, comprising:

Internal cloud platform and external cloud platform;

The resource scheduling apparatus according to any one of claims 6 to 10; wherein the resource scheduling apparatus is in communication connection with the internal cloud platform and the external cloud platform.