CN110825494A

CN110825494A - Physical machine scheduling method and device and computer storage medium

Info

Publication number: CN110825494A
Application number: CN201911060134.1A
Authority: CN
Inventors: 白石
Original assignee: Beijing Jingdong Century Trading Co Ltd; Beijing Jingdong Shangke Information Technology Co Ltd
Current assignee: Beijing Jingdong Century Trading Co Ltd; Beijing Jingdong Shangke Information Technology Co Ltd
Priority date: 2019-11-01
Filing date: 2019-11-01
Publication date: 2020-02-21

Abstract

The disclosure relates to a physical machine scheduling method and device and a computer storage medium, and relates to the field of cloud computing. The physical machine scheduling method comprises the following steps: dividing physical machine clusters according to initial service requirements of various service systems to obtain various sub-clusters, wherein each sub-cluster comprises a plurality of physical machines, and different sub-clusters are used for building different types of cloud platforms and deploying different types of service systems; acquiring the number of physical machines required by each type of service system at fixed time; for each type of service system, under the condition that the number of the existing physical machines of the target sub-cluster corresponding to each type of service system is smaller than the required number of the physical machines, scheduling the physical machines of other sub-clusters except the target sub-cluster so as to meet the required number of the physical machines of each type of service system. According to the method and the device, the efficiency of deploying the service system is improved, and the capacity expansion cost is reduced.

Description

Physical machine scheduling method and device and computer storage medium

Technical Field

The disclosure relates to the field of cloud computing, and in particular, to a physical machine scheduling method and apparatus, and a computer-readable storage medium.

Background

At present, a private cloud platform converts physical resources into virtual resources mainly through a virtualization technology, so as to provide services for users. The core resources such as the CPU and the memory of the physical machine of the virtual machine cluster are virtualized, so that the physical machine can bear a plurality of virtual machines. A user can obtain the virtual machine resources through the management platform and construct a service system on the virtual machine.

With the development of technology, a container cloud platform appears, and a platform capable of uniformly scheduling and managing containers, such as a kubernets container cluster management system, is constructed on the system through a container cluster management system at the bottom layer. The container cloud can be built on a physical machine cluster and also can be built on a virtual machine cluster. And the user constructs a service system on the container by applying for the container resource.

With the continuous expansion of data, big data technology is more and more important. The cluster processing big data has higher requirements on performances such as stability and the like. Therefore, it is more advantageous to process big data in a physical machine cluster because it is more stable and less lossy than a virtual machine cluster.

In the related technology, different types of clusters of different types of service systems are individually expanded in a manner of re-purchasing a physical machine so as to adapt to the requirement change of various service systems.

Disclosure of Invention

The inventor thinks that: in the related technology, the efficiency of deploying the service system is low, and the capacity expansion cost is high.

In view of the above technical problems, the present disclosure provides a solution, which improves the efficiency of deploying a service system and reduces the capacity expansion cost.

According to a first aspect of the present disclosure, there is provided a physical machine scheduling method, including: dividing physical machine clusters according to initial service requirements of various service systems to obtain various sub-clusters, wherein each sub-cluster comprises a plurality of physical machines, and different sub-clusters are used for building different types of cloud platforms and deploying different types of service systems; acquiring the number of physical machines required by each type of service system at fixed time; for each type of service system, under the condition that the number of the existing physical machines of the target sub-cluster corresponding to each type of service system is smaller than the required number of the physical machines, scheduling the physical machines of other sub-clusters except the target sub-cluster so as to meet the required number of the physical machines of each type of service system.

In some embodiments, scheduling the physical machines of the sub-clusters other than the target sub-cluster to satisfy the required number of physical machines of each type of business system comprises: and dividing the specified number of physical machines of other sub-clusters except the target sub-cluster into the target sub-clusters, wherein the specified number is greater than or equal to the difference value between the required number of physical machines and the number of the existing physical machines.

In some embodiments, partitioning a specified number of physical machines of other sub-clusters than the target sub-cluster into the target sub-cluster comprises: acquiring the physical machines in idle states of the other sub-clusters to obtain at least one idle physical machine; and under the condition that the number of the at least one idle physical machine is greater than or equal to the difference value between the required physical machine number and the existing physical machine number, dividing a specified number of idle physical machines into the target sub-cluster.

In some embodiments, partitioning a specified number of free physical machines into the target sub-cluster comprises: backing up the resources on the idle physical machines with the specified number to specified storage equipment or transferring the resources on the idle physical machines with the specified number to specified physical machines of other sub-clusters; clearing resources on the specified number of idle physical machines; and building a cloud platform corresponding to each type of service system on the specified number of idle physical machines.

In some embodiments, partitioning a specified number of free physical machines into the target sub-cluster comprises: after the idle state of the idle physical machine lasts for a first preset time, switching the state of the idle physical machine into a state to be processed; after the state to be processed of the idle physical machine lasts for a second preset time, switching the state of the idle physical machine into a recovery state; and dividing a specified number of idle physical machines in a recovery state into the target sub-clusters.

In some embodiments, the multi-class sub-cluster includes at least two of a first sub-cluster, a second sub-cluster, and a third sub-cluster, the first sub-cluster is used for building a container cloud platform, the second sub-cluster is used for building a virtual machine based virtual machine cloud platform, and the third sub-cluster is used for building a physical cloud platform.

In some embodiments, scheduling the physical machines of the sub-clusters other than the target sub-cluster to satisfy the required number of physical machines of each type of business system comprises: and migrating the each type of business system to other sub-clusters except the target sub-cluster.

In some embodiments, the target sub-cluster is a first sub-cluster, the other sub-clusters are second sub-clusters, and migrating the each type of business system to other sub-clusters except the target sub-cluster comprises: building a container cloud platform on the other sub-clusters; acquiring a container mirror image corresponding to each type of service system; creating a container instance using the container mirror; and migrating the container instance to the container cloud platform.

In some embodiments, the target sub-cluster is a third sub-cluster, the other sub-clusters are second sub-clusters, and migrating the each type of service system to other sub-clusters except the target sub-cluster comprises: acquiring a service system mirror image corresponding to each type of service system; the service system mirror image is migrated to a virtual machine cloud platform corresponding to the other sub-clusters; and creating a virtual machine instance on the virtual machine cloud platform corresponding to the other sub-clusters by using the service system mirror image.

In some embodiments, the target sub-cluster is a second sub-cluster, the other sub-clusters are third sub-clusters, and migrating the each type of service system to other sub-clusters except the target sub-cluster comprises: building a virtual machine cloud platform on the other sub-clusters; acquiring a service system mirror image corresponding to each type of service system; creating a virtual machine instance by using the service system mirror image; and migrating the virtual machine instance to the virtual machine cloud platform.

In some embodiments, the physical machine cluster includes a plurality of racks, each rack includes a plurality of physical machines, and dividing the physical machine cluster according to the initial service requirements of the various service systems to obtain the multi-class sub-clusters includes: for each type of service system, under the condition that the initial service requirement comprises a first service type, all physical machines of at least one rack are selected to form a type of sub-cluster, wherein the first service type comprises a low-reliability service type.

In some embodiments, the physical machine cluster includes a plurality of racks, each rack includes a plurality of physical machines, the physical machine cluster is divided according to the initial service requirements of the various service systems, and the obtaining of the multi-class sub-cluster further includes: for each type of service system, under the condition that the initial service requirement is a second service type, selecting part of physical machines from each rack of different racks respectively to form a type of sub-cluster, wherein the second service type comprises a high-security service type.

According to a second aspect of the present disclosure, there is provided a scheduling apparatus for a physical machine, comprising: the system comprises a dividing module, a storage module and a processing module, wherein the dividing module is configured to divide a physical machine cluster according to initial service requirements of various service systems to obtain various sub-clusters, each sub-cluster comprises a plurality of physical machines, and different sub-clusters are used for building different types of cloud platforms and deploying different types of service systems; the acquisition module is configured to acquire the number of the required physical machines of each type of service system at regular time; and the scheduling module is configured to schedule the physical machines of other sub-clusters except the target sub-cluster to meet the required number of the physical machines of each type of service system under the condition that the existing number of the physical machines of the target sub-cluster corresponding to each type of service system is less than the required number of the physical machines.

According to a third aspect of the present disclosure, there is provided a block chain-based scheduling apparatus for a physical machine, including: a memory; and a processor coupled to the memory, the processor configured to perform the physical machine scheduling method of any of the above embodiments based on instructions stored in the memory.

According to a fourth aspect of the present disclosure, a physical machine scheduling system comprises: the physical machine scheduling apparatus and the physical machine cluster according to any of the embodiments above, wherein the physical machine scheduling apparatus is configured to execute the physical machine scheduling method according to any of the embodiments above on the physical machine cluster.

According to a fifth aspect of the present disclosure, a computer-storable medium has stored thereon computer program instructions which, when executed by a processor, implement the physical machine scheduling method of any one of the above embodiments.

In the embodiment, the efficiency of deploying the service system is improved, and the capacity expansion cost is reduced.

Drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments of the disclosure and together with the description, serve to explain the principles of the disclosure.

The present disclosure may be more clearly understood from the following detailed description, taken with reference to the accompanying drawings, in which:

FIG. 1 illustrates a flow diagram of a physical machine scheduling method according to some embodiments of the present disclosure;

FIG. 2A illustrates a schematic diagram of selecting all physical machines of at least one rack to form a class of sub-clusters according to some embodiments of the present disclosure;

FIG. 2B is a schematic diagram illustrating the selection of a portion of physical machines from each of the different racks to form a class of sub-clusters, respectively, according to some embodiments of the present disclosure;

FIG. 3 illustrates an architecture diagram of a multi-cloud platform system, according to some embodiments of the present disclosure;

FIG. 4 illustrates a schematic diagram of building a multi-cloud platform system, according to some embodiments of the present disclosure;

FIG. 5 is a schematic diagram of a network management layer of a multi-element cloud according to some embodiments of the present disclosure;

FIG. 6 illustrates a block diagram of a physical machine scheduling apparatus, according to some embodiments of the present disclosure;

FIG. 7 illustrates a block diagram of a physical machine scheduling apparatus, in accordance with some embodiments of the present disclosure;

FIG. 8 illustrates a block diagram of a computer system for implementing some embodiments of the present disclosure.

Detailed Description

Various exemplary embodiments of the present disclosure will now be described in detail with reference to the accompanying drawings. It should be noted that: the relative arrangement of the components and steps, the numerical expressions, and numerical values set forth in these embodiments do not limit the scope of the present disclosure unless specifically stated otherwise.

Meanwhile, it should be understood that the sizes of the respective portions shown in the drawings are not drawn in an actual proportional relationship for the convenience of description.

The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the disclosure, its application, or uses.

Techniques, methods, and apparatus known to those of ordinary skill in the relevant art may not be discussed in detail but are intended to be part of the specification where appropriate.

In all examples shown and discussed herein, any particular value should be construed as merely illustrative, and not limiting. Thus, other examples of the exemplary embodiments may have different values.

It should be noted that: like reference numbers and letters refer to like items in the following figures, and thus, once an item is defined in one figure, further discussion thereof is not required in subsequent figures.

Fig. 1 illustrates a flow diagram of a physical machine scheduling method according to some embodiments of the present disclosure.

As shown in fig. 1, the scheduling method of the physical machine includes steps S110 to S130. In some embodiments, the physical machine scheduling method is performed by a management cluster comprised of a plurality of physical machines.

In step S110, the physical machine cluster is divided according to the initial service requirements of each service system, so as to obtain a plurality of types of sub-clusters. Each type of sub-cluster comprises a plurality of physical machines, and different sub-clusters are used for building different types of cloud platforms and deploying different types of service systems. For example, the initial business requirements include an initial number of physical machines required to deploy different types of business systems in an initial state.

In some embodiments, the multi-class sub-cluster includes at least two of the first sub-cluster, the second sub-cluster, and the third sub-cluster. The first sub-cluster is used for building a container cloud platform. The second sub-cluster is used for building a virtual machine cloud platform based on the virtual machine. The third sub-cluster is used for building a physical cloud platform. The management cluster is responsible for controlling and managing the various sub-clusters.

In some embodiments, the physical machine cluster includes a plurality of racks, each rack including a plurality of physical machines. The method comprises the following steps of dividing physical machine clusters according to initial service requirements of various service systems to obtain various sub-clusters.

For each class of service system, under the condition that the initial service requirement comprises a first service type, all physical machines of at least one rack are selected to form a class of sub-clusters. Wherein the first traffic type comprises a low reliability traffic type.

The step of selecting all physical machines of at least one rack to form a type of sub-cluster in some embodiments will be described in detail below with reference to fig. 2A.

Fig. 2A illustrates a schematic diagram of selecting all physical machines of at least one rack to form a class of sub-clusters according to some embodiments of the present disclosure.

As shown in fig. 2A, the racks A, B, C, D, E, F in the physical machine cluster include 5 physical machines per rack. In some embodiments, the technician designates all 3 physical machines of rack G to form a management cluster. And the management cluster divides the physical machine cluster to obtain a first sub-cluster, a second sub-cluster and a third sub-cluster. For example, the division is performed in units of racks, all physical machines of the rack a and all physical machines of the rack B are divided into a first sub-cluster, all physical machines of the rack C and all physical machines of the rack D are divided into a second sub-cluster, and all physical machines of the rack E and all physical machines of the rack F are divided into a third sub-cluster.

The method for dividing the physical machine clusters by taking the rack as a unit is suitable for scenes with small change of a service system, main operation of intranet communication in the same type of cloud platform or general availability requirements. In the present disclosure, this type of division is referred to as vertical division, and the sub-clusters obtained by vertical division are relatively independent and clearly divided.

For example, a physical machine cluster includes a plurality of racks, each rack including a plurality of physical machines. The method for obtaining the multi-class sub-clusters by dividing the physical machine clusters according to the initial service requirements of various service systems further comprises the following steps.

For each type of service system, under the condition that the initial service requirement is a second service type, selecting part of physical machines from each rack of different racks respectively to form a type of sub-cluster. Wherein the second traffic type comprises a high security traffic type.

The following will describe in detail, with reference to fig. 2B, a step of selecting a part of physical machines from each of different racks to form a type of sub-cluster in some embodiments.

FIG. 2B illustrates a schematic diagram of a selection of a portion of physical machines from each of different racks to form a class of sub-clusters, respectively, according to some embodiments of the present disclosure.

As shown in fig. 2B, the physical machine cluster includes physical machines 2 to 5 of rack A, B, C, D, E, F and physical machines 1 to 5 of rack G. In some embodiments, physical machines 1 of rack A, B, C, D, E, F are selected by a technician to form a management cluster. And the management cluster divides the physical machine cluster to obtain a first sub-cluster, a second sub-cluster and a third sub-cluster. For example, the physical machines 3, 4 of rack A, B, C, D, E, F are divided into a first sub-cluster. Physical machine 2 of rack A, B, C, D and physical machines 1 and 2 of rack E, F are divided into a second sub-cluster. Physical machines 5 of rack A, B, C, D, E, F are divided into a third sub-cluster. It should be understood that some physical machines in different racks may be selected randomly, or some physical machines in different racks may be selected according to the physical machine identification sequence, and the sequential selection manner is easier to manage.

In the mode of selecting a physical machine as a type of sub-cluster across a plurality of racks, the type of sub-cluster is distributed on different racks, so that resource instances of a service system in the type of cluster can be distributed on different racks and also can be distributed in different fault domains, and the service system can obtain better capacity of resisting the fault of the racks. The horizontal division mode can adapt to the development change of a service system, the implementation of intranet communication in various cloud platforms or scenes with higher availability requirements.

In some embodiments, in the case that a user needs to have a certain security requirement level higher than a preset level, it indicates that isolation protection needs to be performed on the service system, and therefore, isolation area resources need to be provided to meet the isolation requirement of the service. In this case, for example, all the physical machines of the rack G of the physical machine cluster are divided into isolation zones.

In some embodiments, the two ways of partitioning the physical machine cluster described above may exist simultaneously. I.e. the division is performed in a mixed manner.

Returning to fig. 1, in step S120, the required number of physical machines of each type of service system is obtained periodically. It should be understood that, as time changes, the number of the required physical machines of each type of service system at different times changes continuously, that is, the present disclosure implements, for each type of service system, dynamic adjustment of the correspondence among the service system, the physical machines, and the sub-clusters, and implements reasonable allocation of resources.

In step S130, for each class of service system, in the case that the existing physical machine number of the target sub-cluster corresponding to each class of service system is less than the required physical machine number, scheduling the physical machines of other sub-clusters except the target sub-cluster to satisfy the required physical machine number of each class of service system. And under the condition that the number of the existing physical machines of the target sub-cluster corresponding to each type of service system is less than the required number of the physical machines, indicating that the service system needs to be expanded and increasing the number of the physical machines, thereby providing support for the operation of the service system.

Scheduling the physical machines of the sub-clusters other than the target sub-cluster to meet the required number of physical machines of each type of business system is achieved, for example, as follows.

A specified number of physical machines of the other sub-clusters except the target sub-cluster are divided into the target sub-clusters. The specified number is greater than or equal to a difference between the number of required physical machines and the number of existing physical machines. And designating the part of which the quantity is greater than the difference value between the required physical machine quantity and the existing physical machine quantity for the emergency of the business system.

In some embodiments, partitioning a specified number of physical machines of the other sub-clusters to the target sub-cluster includes the following steps.

Firstly, acquiring the physical machines in idle states of other sub-clusters to obtain at least one idle physical machine.

And then, under the condition that the number of at least one idle physical machine is greater than or equal to the difference value between the required physical machine number and the existing physical machine number, dividing the specified number of idle physical machines into the target sub-clusters. In some embodiments, the target sub-cluster is augmented by purchasing a new physical machine in the event that the number of at least one idle physical machine is less than the difference between the number of required physical machines and the number of existing physical machines.

In some embodiments, the partitioning of the specified number of free physical machines to the target sub-cluster is accomplished as follows.

Firstly, backing up the resources on the specified number of idle physical machines to the specified storage device or migrating the resources on the specified number of idle physical machines to the specified physical machine of other sub-cluster. Then, the resources on the specified number of idle physical machines are emptied. And finally, building a cloud platform corresponding to each type of service system on a specified number of idle physical machines.

In other embodiments, the partitioning of the specified number of free physical machines to the target sub-cluster is accomplished as follows.

Firstly, after the idle state of the idle physical machine lasts for a first preset time, the state of the idle physical machine is switched to a to-be-processed state. And then, after the state to be processed of the idle physical machine lasts for a second preset time, switching the state of the idle physical machine into a recovery state. And finally, dividing the specified number of idle physical machines in the recovery state into target sub-clusters.

And migrating each type of business system to other sub-clusters except the target sub-cluster.

In some embodiments, the target sub-cluster is a first sub-cluster and the other sub-clusters are second sub-clusters. Migrating each type of business system to a subset group other than the target subset group includes the following steps.

First, a container cloud platform is built on other sub-clusters. For example, a certain number of virtual machines are created on at least one physical machine of the second sub-cluster, and a virtual machine-based container cloud platform is built on the virtual machines.

And secondly, acquiring a container mirror image corresponding to each type of service system.

Then, using the container mirror, a container instance is created. In some embodiments, the storage space of the container image is shared between the virtual machine-based container cloud platform and the original container cloud platform on the target sub-cluster, so that the scheduling system can create container instances on the virtual machine-based container cloud platform and the original container cloud platform by using the container image.

And finally, migrating the container instance to a container cloud platform. In some embodiments, the container cloud platform based on the virtual machine and the original container cloud platform are subjected to unified scheduling management, so that container instances on the original container cloud platform can be migrated and scheduled to the container cloud platform based on the virtual machine cluster.

In some embodiments, the target sub-cluster is a third sub-cluster and the other sub-clusters are second sub-clusters. Migrating each type of business system to a subset group other than the target subset group includes the following steps.

Firstly, a service system mirror image corresponding to each type of service system is obtained. For example, an application analysis tool is utilized to scan system resource dependencies of a business system on a physical cloud platform. System resource dependencies include, for example, system libraries, underlying software, dependent data files, operating environments, and the like. After the resource dependency relationship is recorded and stored, a P2V (Physical to virtual) tool is used to package the system on the Physical machine of the third sub-cluster into a business system image that can be used in the virtual machine environment.

And then, mirroring the service system and migrating the service system to a virtual machine cloud platform corresponding to other sub-clusters. For example, the business system image is migrated to the image management system of the virtual machine cloud platform corresponding to the other sub-cluster by using the data migration tool.

And finally, creating a virtual machine instance on virtual machine cloud platforms corresponding to other sub-clusters by using the service system mirror image.

For example, migrating each type of business system to a sub-cluster other than the target sub-cluster further comprises: and scanning whether the resource dependence of the service system in the virtual machine cloud platform is consistent with the original environment by using an application analysis tool, and verifying whether the service system can normally operate. And when detecting that the key system component is missing, re-executing the step of migrating each type of service system to other sub-clusters except the target sub-cluster, or migrating the relevant files and programs of the key component to the virtual machine cloud platform independently.

In some embodiments, in a case that the service system on the physical machine of the target sub-cluster is already associated with the load balancing instance, the service systems on the virtual machine cloud platforms of other sub-clusters are also associated with the original load balancing instance. And under the condition that the service system on the physical machine of the target sub-cluster is not associated with the load balancing example, creating a new load balancing example, and simultaneously associating the service system on the physical machine of the target sub-cluster and the service systems on the virtual machine cloud platforms of other sub-clusters to the new load balancing example. And the load balancing examples adopt the IP of the physical machine of the target sub-cluster to provide access service to the outside.

In some embodiments, the target sub-cluster is a second sub-cluster and the other sub-clusters are third sub-clusters. Migrating each type of business system to a subset group other than the target subset group includes the following steps.

Firstly, a virtual machine cloud platform is built on other sub-clusters. For example, the virtual machine cloud platform is installed on idle physical machines of other sub-clusters.

And secondly, acquiring a service system mirror image corresponding to each type of service system.

Then, a virtual machine instance is created using the business system image.

And finally, migrating the virtual machine instance to the virtual machine cloud platform.

In some embodiments, in a case that the service system on the physical machine of the target sub-cluster is already associated with the load balancing instance, the service systems on the virtual machine cloud platforms of other sub-clusters are also associated with the original load balancing instance. And under the condition that the service system on the physical machine of the target sub-cluster is not associated with the load balancing example, creating a new load balancing example, and simultaneously associating the service system on the physical machine of the target sub-cluster and the service systems on the virtual machine cloud platforms of other sub-clusters to the new load balancing example. The load balancing examples provide access service to the outside by adopting the IP of the virtual machine corresponding to the physical machine of the target sub-cluster.

The scheduling method of the physical machine realizes the scheduling of the physical machine among the different types of sub-clusters so as to meet the service requirement of the service system, does not need to purchase the physical machine again, improves the efficiency of deploying the service system and reduces the capacity expansion cost.

In some embodiments, the method is applied to a large-scale private cloud scene, and the overall cloud computing environment meeting the large-scale business system is built by utilizing the multi-element cloud. The multi-element cloud is a cluster of divided physical machines to obtain a plurality of types of sub-clusters, and each type of sub-cluster bears one type of cloud. For example, the multivariate cloud includes at least two of a physical cloud, a virtual cloud, and a container cloud.

For example, a data staging system is built on a physical cloud. Because a data center system needs large-scale high-performance data processing capacity, data processing service is relatively stable, and the requirement on stability is high, the data center system is suitable for carrying by adopting physical cloud.

And building an application system background on the container cloud. The container cloud is used for supporting high concurrent processing capacity, and the container cloud is used for bearing a background service program of the application system, so that the whole application system can be well supported.

And building a system needing elastic resource change on the virtual machine cloud. By utilizing the convenient elastic expansion characteristic of the virtual machine cloud, the application system with large resource demand change at the service peak and the service valley can be deployed on the virtual machine cloud, so that the better resource cost control is realized. In the private cloud, the virtual machines can be concentrated on part of the physical hosts, and the idle physical hosts can run with low power consumption or sleep, so that the running cost is reduced.

According to the method and the device, the characteristics of different cloud platforms are utilized, and the integral application system of the large-scale business can be better supported.

Fig. 3 illustrates an architecture diagram of a multi-cloud platform system according to some embodiments of the present disclosure.

As shown in fig. 3, the multi-cloud platform system 3 includes a physical machine cluster 31 and a physical machine scheduling device 32. The physical machine scheduling device 32 is configured to perform the physical machine scheduling method according to any of the embodiments of the present disclosure, for example, perform steps S110 to S130 shown in fig. 1. It should be understood that the multi-cloud platform system herein is a multi-cloud platform system in a state corresponding to a certain time.

The physical machine cluster 31 includes a virtual machine sub-cluster 311, a container sub-cluster 312, and a physical sub-cluster 313.

The virtual machine sub-cluster 311 is configured to build a virtual machine cloud platform based on virtual machine technology, and provide virtual machine computing resources to users.

The container sub-cluster 312 is configured to build a container cloud platform, providing container computing resources to users, based on container orchestration and cluster management techniques.

The physical sub-clusters 313 are configured to build physical cloud platforms providing users with computing resources without performance loss.

The physical machine scheduling device 32 includes a virtual machine sub-cluster management platform 321, a container sub-cluster management platform 322, a physical sub-cluster management platform 323, a network support layer 324, and a unified management platform 325.

The virtual machine sub-cluster management platform 321 is configured to manage and control the virtual machine sub-cluster 311, and provide functions of resource deployment, scheduling, recovery, migration, security management, monitoring, and the like, so that the virtual machine sub-cluster 311 can operate efficiently.

The container sub-cluster management platform 322 is configured to manage and control the container sub-cluster 312, provide functions such as resource deployment, scheduling, recovery, migration, security management, monitoring, and the like, and enable the container sub-cluster 312 to operate efficiently.

The physical sub-cluster management platform 323 is configured to manage and control the physical sub-cluster 313, provide functions such as resource deployment, scheduling, recovery, migration, security management, monitoring, and the like, and enable the physical sub-cluster 313 to operate efficiently.

The network support layer 324 is configured to provide core functions such as network communication, network isolation, network monitoring, and the like, and implement network communication and data traffic control of the overall multi-element cloud platform system, so that each module of a sub-cluster based on different cloud resource types can communicate with each other at a high speed, and mutual isolation between sub-clusters of different cloud resource types is ensured.

The unified management platform 325 is configured to uniformly manage resource deployment of the three seed clusters, perform unified scheduling on physical server resources, and recover redundant resources. The unified management platform is also configured to provide global user authority management, tenant management, and the like. The users comprise tenants of the cloud platform and technical personnel for monitoring and managing the multi-element cloud platform system. The unified management platform is further configured to perform unified storage and management on data, manage migration of computing resources, perform unified security management, control operation of the multi-element cloud platform system, and the like.

Fig. 4 illustrates a schematic diagram of building a multi-cloud platform system, according to some embodiments of the present disclosure.

As shown in FIG. 4, the OpenStack Ironics platform 41 is utilized to rapidly deploy a host operating system on each physical machine of each type of sub-cluster of a physical machine cluster. The host operating system is the CentOS operating system 430. Ironic is a child of the OpenStack community, dedicated to providing bare metal services.

A plurality of virtual machines 432A are built on the host operating systems of the respective physical machines of the virtual machine sub-cluster 43A using the OpenStack framework 431A. The system of virtual machines includes, for example, any one of KVM, Xen, and QEMU. A business system 433A of the user can be deployed on the plurality of virtual machines 432A. In some embodiments, the virtual machine-based container 434A can be built on multiple virtual machines using kubernets technology. Kubernets is an open source for managing containerized applications on multiple hosts in a cloud platform.

The Container orchestration and management system 431B is built on the host operating systems of the physical machines of the Container sub-cluster 43B using kubernets technology, and provides the Container computing resources for the user by supporting Pod432B and Container (Container) 433B. Business systems 434B of the users can be deployed on the multiple containers. Pod is the smallest deployable unit that can create and manage kubernets computations.

Basic software such as Driver 431C, static library 432C, and database 433C, which are necessary for running an application, is installed on the host operating system of each physical machine of the physical machine sub-cluster 43C, and deployment of the business system 434C is realized.

The bottom layer of the multi-platform system performs unified management on computing, storage and network resources through the multi-cloud management platform 42, so as to realize unified scheduling of hardware resources. The multi-cloud management platform includes a block storage 421, an OpenStack base network 422, and a unified management module 423, and is configured to perform the physical machine scheduling method according to any some embodiments of the present disclosure.

It should be understood that the multi-cloud platform system constructed here is a multi-cloud platform system in a state corresponding to a certain time.

Fig. 5 is a schematic diagram of a network management layer of a multi-element cloud according to some embodiments of the present disclosure.

As shown in fig. 5, the network management layer 5 of the multivariate cloud includes an underlay layer 51 and an Overlay layer 52. In some embodiments, the network management layer 5 belongs to the network support layer 324 as shown in fig. 3.

The underslay layer 51 includes a Border layer 511, a Spine layer 512, and a ServerLeaf layer 513. The network device and the physical machine device are divided into three layers through a Border layer 511, a Spine layer 512 and a Serverleaf layer 513, and the network device and the physical machine device are managed by using a management network and an out-of-band management network.

Overlay layer 52 includes SDN controller 521, virtual machine cloud network 522, container cloud network 523, and physical cloud network 524. The SDN controller 521 implements network interworking and security isolation of resources in the virtual machine cloud network 522, the container cloud network 523, and the physical cloud network 524 using SDN technology.

Fig. 6 illustrates a block diagram of a physical machine scheduling apparatus, according to some embodiments of the present disclosure.

As shown in fig. 6, the physical machine scheduling apparatus 6 includes a dividing module 61, an obtaining module 62, and a scheduling module 63.

The dividing module 61 is configured to divide the physical machine cluster according to the initial service requirements of various service systems, so as to obtain multiple types of sub-clusters, for example, execute step S110 shown in fig. 1.

The obtaining module 62 is configured to obtain the required number of physical machines of each type of service system at regular time, for example, execute step S120 shown in fig. 1.

The scheduling module 63 is configured to, for each class of service system, in a case that the existing physical machine number of the target sub-cluster corresponding to each class of service system is less than the required physical machine number, schedule the physical machines of the sub-clusters other than the target sub-cluster to meet the required physical machine number of each class of service system, for example, execute step S130 shown in fig. 1.

Fig. 7 illustrates a block diagram of a physical machine scheduling apparatus, according to some embodiments of the present disclosure.

As shown in fig. 7, the physical machine scheduling device 7 includes a memory 71; and a processor 72 coupled to the memory 71. The memory 71 is used for storing instructions for executing the corresponding embodiments of the physical machine scheduling method. The processor 72 is configured to perform the physical machine scheduling method in any of the embodiments of the present disclosure based on instructions stored in the memory 71.

In some embodiments, a physical machine scheduling system includes a physical machine scheduling apparatus and a physical machine cluster according to any embodiment of the present disclosure. The physical machine scheduling apparatus is configured to perform the physical machine scheduling method in any of the embodiments of the present disclosure on the cluster of physical machines.

As shown in FIG. 8, computer system 80 may take the form of a general purpose computing device. Computer system 80 includes a memory 810, a processor 820, and a bus 800 that connects the various system components.

The memory 810 may include, for example, system memory, non-volatile storage media, and the like. The system memory stores, for example, an operating system, an application program, a Boot Loader (Boot Loader), and other programs. The system memory may include volatile storage media such as Random Access Memory (RAM) and/or cache memory. The non-volatile storage medium, for example, stores instructions to perform corresponding embodiments of at least one of the physical machine scheduling methods. Non-volatile storage media include, but are not limited to, magnetic disk storage, optical storage, flash memory, and the like.

The processor 820 may be implemented as discrete hardware components, such as a general purpose processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA) or other programmable logic device, discrete gates or transistors, or the like. Accordingly, each of the modules, such as the judging module and the determining module, may be implemented by a Central Processing Unit (CPU) executing instructions in a memory for performing the corresponding step, or may be implemented by a dedicated circuit for performing the corresponding step.

The bus 800 may use any of a variety of bus architectures. For example, bus structures include, but are not limited to, Industry Standard Architecture (ISA) bus, Micro Channel Architecture (MCA) bus, and Peripheral Component Interconnect (PCI) bus.

The computer system 80 may also include an input-output interface 830, a network interface 840, a storage interface 850, and the like. These

interfaces

830, 840, 850 and the connection between the memory 88 and the processor 820 may be via a bus 800. The input/output interface 830 may provide a connection interface for input/output devices such as a display, a mouse, and a keyboard. The network interface 840 provides a connection interface for various networking devices. The storage interface 850 provides a connection interface for external storage devices such as a floppy disk, a usb disk, and an SD card.

Various aspects of the present disclosure are described herein with reference to flowchart illustrations and/or block diagrams of methods, apparatus and computer program products according to embodiments of the disclosure. It will be understood that each block of the flowchart illustrations and/or block diagrams, and combinations of blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer-readable program instructions.

These computer-readable program instructions may be provided to a processor of a general purpose computer, special purpose computer, or other programmable apparatus to produce a machine, such that the execution of the instructions by the processor results in an apparatus that implements the functions specified in the flowchart and/or block diagram block or blocks.

These computer-readable program instructions may also be stored in a computer-readable memory that can direct a computer to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instructions which implement the function specified in the flowchart and/or block diagram block or blocks.

The present disclosure may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects.

By the physical machine scheduling method and device and the computer storage medium in the embodiment, the efficiency of deploying the service system is improved, and the capacity expansion cost is reduced.

So far, the physical machine scheduling method and apparatus, and the computer-readable storage medium according to the present disclosure have been described in detail. Some details that are well known in the art have not been described in order to avoid obscuring the concepts of the present disclosure. It will be fully apparent to those skilled in the art from the foregoing description how to practice the presently disclosed embodiments.

Claims

1. A method of scheduling a physical machine, comprising:

dividing physical machine clusters according to initial service requirements of various service systems to obtain various sub-clusters, wherein each sub-cluster comprises a plurality of physical machines, and different sub-clusters are used for building different types of cloud platforms and deploying different types of service systems;

acquiring the number of physical machines required by each type of service system at fixed time;

for each type of service system, under the condition that the number of the existing physical machines of the target sub-cluster corresponding to each type of service system is smaller than the required number of the physical machines, scheduling the physical machines of other sub-clusters except the target sub-cluster so as to meet the required number of the physical machines of each type of service system.

2. The physical machine scheduling method of claim 1, wherein scheduling the physical machines of the other sub-clusters except the target sub-cluster to meet the required number of physical machines of each class of business system comprises:

and dividing the specified number of physical machines of other sub-clusters except the target sub-cluster into the target sub-clusters, wherein the specified number is greater than or equal to the difference value between the required number of physical machines and the number of the existing physical machines.

3. The physical machine scheduling method of claim 2, wherein dividing the specified number of physical machines of the other sub-clusters to the target sub-cluster comprises:

acquiring the physical machines in idle states of the other sub-clusters to obtain at least one idle physical machine;

and under the condition that the number of the at least one idle physical machine is greater than or equal to the difference value between the required physical machine number and the existing physical machine number, dividing a specified number of idle physical machines into the target sub-cluster.

4. The physical machine scheduling method of claim 3, wherein dividing a specified number of idle physical machines into the target sub-clusters comprises:

backing up the resources on the idle physical machines with the specified number to specified storage equipment or transferring the resources on the idle physical machines with the specified number to specified physical machines of other sub-clusters;

clearing resources on the specified number of idle physical machines;

and building a cloud platform corresponding to each type of service system on the specified number of idle physical machines.

5. The physical machine scheduling method of claim 3, wherein dividing a specified number of idle physical machines into the target sub-clusters comprises:

after the idle state of the idle physical machine lasts for a first preset time, switching the state of the idle physical machine into a state to be processed;

after the state to be processed of the idle physical machine lasts for a second preset time, switching the state of the idle physical machine into a recovery state;

and dividing a specified number of idle physical machines in a recovery state into the target sub-clusters.

6. The physical machine scheduling method according to claim 1, wherein the plurality of types of sub-clusters include at least two of a first sub-cluster, a second sub-cluster and a third sub-cluster, the first sub-cluster is used for building a container cloud platform, the second sub-cluster is used for building a virtual machine cloud platform based on a virtual machine, and the third sub-cluster is used for building a physical cloud platform.

7. The physical machine scheduling method of claim 6, wherein scheduling the physical machines of the other sub-clusters except the target sub-cluster to meet the required number of physical machines of each class of business system comprises:

and migrating the each type of business system to other sub-clusters except the target sub-cluster.

8. The physical machine scheduling method of claim 7, wherein the target sub-cluster is a first sub-cluster, the other sub-clusters are second sub-clusters, and migrating the each type of business system to the other sub-clusters except the target sub-cluster comprises:

building a container cloud platform on the other sub-clusters;

acquiring a container mirror image corresponding to each type of service system;

creating a container instance using the container mirror;

and migrating the container instance to the container cloud platform.

9. The physical machine scheduling method of claim 7, wherein the target sub-cluster is a third sub-cluster, the other sub-clusters are second sub-clusters, and migrating the each type of business system to the other sub-clusters except the target sub-cluster comprises:

acquiring a service system mirror image corresponding to each type of service system;

the service system mirror image is migrated to a virtual machine cloud platform corresponding to the other sub-clusters;

and creating a virtual machine instance on the virtual machine cloud platform corresponding to the other sub-clusters by using the service system mirror image.

10. The physical machine scheduling method of claim 7, wherein the target sub-cluster is a second sub-cluster, the other sub-clusters are third sub-clusters, and migrating the each type of business system to other sub-clusters except the target sub-cluster comprises:

building a virtual machine cloud platform on the other sub-clusters;

creating a virtual machine instance by using the service system mirror image;

and migrating the virtual machine instance to the virtual machine cloud platform.

11. The physical machine scheduling method of claim 1, wherein the physical machine cluster comprises a plurality of racks, each rack comprises a plurality of physical machines, and dividing the physical machine cluster according to initial service requirements of the various service systems to obtain the multi-class sub-clusters comprises:

for each type of service system, under the condition that the initial service requirement comprises a first service type, all physical machines of at least one rack are selected to form a type of sub-cluster, wherein the first service type comprises a low-reliability service type.

12. The physical machine scheduling method of claim 1, wherein the physical machine cluster comprises a plurality of racks, each rack comprises a plurality of physical machines, and the dividing the physical machine cluster according to the initial service requirements of the various service systems to obtain the various types of sub-clusters further comprises:

for each type of service system, under the condition that the initial service requirement is a second service type, selecting part of physical machines from each rack of different racks respectively to form a type of sub-cluster, wherein the second service type comprises a high-security service type.

13. A physical machine scheduling apparatus comprising:

the system comprises a dividing module, a storage module and a processing module, wherein the dividing module is configured to divide a physical machine cluster according to initial service requirements of various service systems to obtain various sub-clusters, each sub-cluster comprises a plurality of physical machines, and different sub-clusters are used for building different types of cloud platforms and deploying different types of service systems;

the acquisition module is configured to acquire the number of the required physical machines of each type of service system at regular time;

and the scheduling module is configured to schedule the physical machines of other sub-clusters except the target sub-cluster to meet the required number of the physical machines of each type of service system under the condition that the existing number of the physical machines of the target sub-cluster corresponding to each type of service system is less than the required number of the physical machines.

14. A physical machine scheduling apparatus comprising:

a memory; and

a processor coupled to the memory, the processor configured to perform the physical machine scheduling method of any of claims 1-12 based on instructions stored in the memory.

15. A physical machine scheduling system comprising:

the physical machine scheduling apparatus and physical machine cluster of claim 13 or 14, wherein the physical machine scheduling apparatus is configured to perform the physical machine scheduling method of any one of claims 1 to 12 on the physical machine cluster.

16. A computer-storable medium having stored thereon computer program instructions which, when executed by a processor, implement the physical machine scheduling method of any one of claims 1 to 12.