CN117395248A

CN117395248A - Method, device, equipment and medium for scheduling application arrangement based on computing power network

Info

Publication number: CN117395248A
Application number: CN202210782074.XA
Authority: CN
Inventors: 史露强; 钟储建; 潘宇虹; 陈航; 郭岳; 陈健飞; 陈琪
Original assignee: China Mobile Communications Group Co Ltd; China Mobile Group Zhejiang Co Ltd
Current assignee: China Mobile Communications Group Co Ltd; China Mobile Group Zhejiang Co Ltd
Priority date: 2022-07-04
Filing date: 2022-07-04
Publication date: 2024-01-12

Abstract

The embodiment of the invention relates to the technical field of computers, and discloses an application programming and scheduling method, device and equipment based on a computing power network and a computer readable storage medium, wherein the method comprises the following steps: the first-level application scheduler sends a query request to the computational network arrangement layer according to the computational power request, acquires computational power arrangement information and network routing information sent by the computational network arrangement layer, and determines a target data center corresponding to the computational power request; an application controller of a target data center receives a first scheduling instruction sent by a primary application scheduler, pulls an application mirror image from a mirror image warehouse of the target data center, and synchronizes the application mirror image from a mirror image warehouse of an original data center to a mirror image warehouse of the target data center; the application controller of the target data center creates an application instance according to the application image. The embodiment of the invention solves the problem of ubiquitous power-on service application in the power-on network in the prior art, and the application instance of the service application can be smoothly scheduled among different data centers.

Description

Method, device, equipment and medium for scheduling application arrangement based on computing power network

Technical Field

The embodiment of the invention relates to the technical field of computers, in particular to an application programming and scheduling method, device and equipment based on a computing power network and a computer readable storage medium.

Background

The calculation network takes calculation force as a center and takes the network as a root, so that a user can enjoy calculation force connection at any time and any place. Through innovatively constructing a ubiquitous fused computing power network, the fusion development of computing power in a physical space, a logical space and a heterogeneous space is promoted, so that the computing power is a social service like water and electricity, which can be accessed at one point and used at the same time, and the network is free from places, the computing power is free from places and the intelligence is free from places. The current research of the computational power network is mainly focused on exploration in aspects of computational power measurement, computational power perception, computational power routing, computational power management and the like, and a novel computational power routing measurement mechanism of 'optimal path plus optimal computational power' is supported by means of a unified computational power measurement system and capability template.

However, for the computational power network, how the application realizes smooth scheduling based on computational power and network routing, so that the application can run on ubiquitous computational power, which is still relatively lack of research.

Disclosure of Invention

In view of the above problems, embodiments of the present invention provide an application scheduling method, apparatus, device, and computer readable storage medium based on a computing power network, which are used to solve the problem of how to enable service applications by ubiquitous computing power in the computing power network in the prior art, and an application instance of a service application can be smoothly scheduled between different data centers.

According to an aspect of an embodiment of the present invention, there is provided an application orchestration scheduling method based on a computing network, the method including:

the first-level application scheduler sends a query request to the computational network arrangement layer according to the computational power request, acquires computational power arrangement information and network routing information sent by the computational network arrangement layer, and determines a target data center corresponding to the computational power request;

an application controller of a target data center receives a first scheduling instruction sent by a primary application scheduler, pulls an application mirror image from a mirror image warehouse of the target data center, and synchronizes the application mirror image from a mirror image warehouse of an original data center to a mirror image warehouse of the target data center;

the application controller of the target data center creates an application instance according to the application image.

In an alternative manner, before the application controller of the target data center receives the first scheduling instruction sent by the primary application scheduler and pulls the application image from the image bin library of the target data center, the method further includes:

the first-level application scheduler determines the application type of the calculation force request;

under the condition that the application type is a stateful application, the primary application scheduler decides the granularity of the data slice according to the user service level protocol and the algorithm information;

The primary data manager receives a second scheduling instruction of the primary application scheduler and sends data slice granularity and target data center information to a data controller of the original data center.

The data controller of the original data center receives the data slicing granularity and the target data center information, slices the application data corresponding to the calculation force request, and migrates the sliced data to the target data center.

In an optional manner, after the application controller of the target data center receives the first scheduling instruction sent by the primary application scheduler and pulls the application image from the image bin library of the target data center, the method further includes:

the application controller of the original data center receives a third scheduling instruction sent by the primary application scheduler, and destroys the original application instance corresponding to the calculation force request.

In an alternative manner, after the application controller of the target data center creates the application instance from the application image, the method further includes:

the application controller of the target data center controls the application embodiment to consume the corresponding service data to obtain a processing result;

the application controller of the target data center sends the processing result to a primary application scheduler;

And the first-stage application scheduler receives the processing results and performs aggregation processing on the processing results.

the mirror image warehouse of each data center synchronizes the application mirror images of the data centers according to the scheduling conditions, and synchronizes the application mirror images to the primary mirror image warehouse;

and the primary mirror image warehouse backs up the synchronous application mirror images.

In an alternative manner, before the primary application scheduler sends the query request to the algorithm orchestration layer according to the computational power request, the method further comprises:

the computing network sensing layer acquires real-time sensing information of each data center, registers the real-time sensing information to the computing network arrangement layer, and the real-time sensing information comprises computing power information and network information;

after the primary application scheduler sends the query request to the computing network orchestration layer according to the computing power request, the method further comprises:

and the computing network arrangement layer performs computing power arrangement and network route planning according to the computing power information and the network information to obtain computing power arrangement information and network route information, and sends the computing power arrangement information and the network route information to the primary application dispatcher.

According to another aspect of the embodiments of the present invention, there is provided an application orchestration scheduling device based on a computing network, including:

the first-level application scheduler is used for sending a query request to the computational network arrangement layer according to the computational power request, acquiring computational power arrangement information and network routing information sent by the computational network arrangement layer, and determining a target data center corresponding to the computational power request.

The application controller of the target data center is used for receiving the first scheduling instruction sent by the first-stage application scheduler, pulling an application mirror image from a mirror image warehouse of the target data center, and synchronizing the application mirror image from a mirror image warehouse of the original data center to a mirror image warehouse of the target data center; an application instance is created from the application image.

In an alternative, the apparatus further comprises:

the first-level application scheduler is further used for determining the application type of the computing force request before the application controller of the target data center receives the first scheduling instruction sent by the first-level application scheduler and pulls the application mirror image from the mirror image warehouse of the target data center; in the case of a stateful application, the data slice granularity is determined according to the user service level protocol and the gateway information.

And the primary data manager is used for receiving a second scheduling instruction of the primary application scheduler and sending the data slice granularity and the target data center information to the data controller of the original data center.

And the data controller of the original data center is used for receiving the data slicing granularity and the target data center information, slicing the application data corresponding to the calculation force request, and transferring the sliced data to the target data center.

In an alternative, the apparatus further comprises:

and the application controller of the primary data center is used for receiving the first scheduling instruction sent by the primary application scheduler from the application controller of the target data center, pulling the application image from the image warehouse of the target data center, receiving the third scheduling instruction sent by the primary application scheduler, and destroying the primary application instance corresponding to the calculation force request.

In an optional manner, the application controller of the target data center is further configured to control the application embodiment to consume the corresponding service data after the application controller of the target data center creates the application instance according to the application image, so as to obtain a processing result; and sending the processing result to the primary application scheduler.

The first-level application dispatcher is also used for receiving the processing results and carrying out aggregation processing on the processing results.

In an alternative, the apparatus further comprises:

and the mirror image warehouse of the data center is used for synchronizing the application mirror images of the data center according to the scheduling condition and synchronizing the application mirror images to the primary mirror image warehouse.

And the primary mirror image warehouse is used for backing up the synchronous application mirror images.

In an alternative, the apparatus further comprises:

the computing network perception layer is used for acquiring real-time perception information of each data center and registering the real-time perception information to the computing network arrangement layer, wherein the real-time perception information comprises computing power information and network information.

And the computing network arrangement layer is used for carrying out computing power arrangement and network route planning according to the computing power information and the network information to obtain computing power arrangement information and network route information, and sending the computing power arrangement information and the network route information to the primary application scheduler.

According to another aspect of the embodiment of the present invention, there is provided an application orchestration scheduling device, including:

the device comprises a processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface complete communication with each other through the communication bus;

the memory is configured to hold at least one executable instruction that causes the processor to:

In one alternative, the executable instructions cause the processor to:

before an application controller of a target data center receives a first scheduling instruction sent by a primary application scheduler, and pulls an application image from an image warehouse of the target data center, the primary application scheduler determines the application type of a calculation force request;

In one alternative, the executable instructions cause the processor to:

after an application controller of a target data center receives a first scheduling instruction sent by a first-stage application scheduler and pulls an application image from an image warehouse of the target data center, an application controller of an original data center receives a third scheduling instruction sent by the first-stage application scheduler, and an original application instance corresponding to the computing force request is destroyed.

In one alternative, the executable instructions cause the processor to:

after an application controller of the target data center creates an application instance according to the application image, the application controller of the target data center controls the application embodiment to consume corresponding service data to obtain a processing result;

the application controller of the target data center sends a processing result to the primary application dispatcher;

In one alternative, the executable instructions cause the processor to:

before an application controller of a target data center receives a first scheduling instruction sent by a primary application scheduler and pulls an application image from an image warehouse of the target data center, an image warehouse of each data center synchronizes the application image of the data center according to scheduling conditions and synchronizes the application image to the primary image warehouse;

In one alternative, the executable instructions cause the processor to:

before a primary application scheduler sends a query request to a computational network arrangement layer according to a computational power request, a computational network perception layer acquires real-time perception information of each data center and registers the real-time perception information to the computational network arrangement layer, wherein the real-time perception information comprises computational power information and network information;

after the first-level application scheduler sends a query request to the calculation network arrangement layer according to the calculation power request, the calculation network arrangement layer performs calculation power arrangement and network route planning according to the calculation power information and the network information to obtain calculation power arrangement information and network route information, and sends the calculation power arrangement information and the network route information to the first-level application scheduler.

According to yet another aspect of embodiments of the present invention, there is provided a computer-readable storage medium having stored therein at least one executable instruction for causing an application orchestration scheduling device to:

In an alternative, the executable instructions may be specifically configured to cause an application orchestration scheduling device to:

According to the embodiment of the invention, the primary application scheduler sends the query request to the computational network arrangement layer according to the computational power request, and the computational power arrangement information and the network routing information sent by the computational network arrangement layer are obtained, so that the current computational power resource is determined to be better, the computing task is suitable to be executed, and the target data center for external service is provided. The application controller of the target data center receives the first scheduling instruction sent by the first-level application scheduler, pulls an application image from an image warehouse of the target data center, synchronizes the application image from an image warehouse of the original data center to an image warehouse of the target data center, creates an application instance according to the application image, can build a hierarchical control mechanism according to application scheduling, can realize smooth scheduling deployment of the application instance of the service application among different data centers, pulls up the application embodiment and provides services to the outside, and solves the problem of how to enable the service application by ubiquitous computing power in a computing network.

The foregoing description is only an overview of the technical solutions of the embodiments of the present invention, and may be implemented according to the content of the specification, so that the technical means of the embodiments of the present invention can be more clearly understood, and the following specific embodiments of the present invention are given for clarity and understanding.

Drawings

The drawings are only for purposes of illustrating embodiments and are not to be construed as limiting the invention. Also, like reference numerals are used to designate like parts throughout the figures. In the drawings:

FIG. 1 is a schematic view of a scenario of a computational power process according to an embodiment of the present invention;

fig. 2 is a schematic flow chart of an application scheduling method based on a computing power network according to an embodiment of the present invention;

fig. 3 is a schematic flow chart of an application calling part processing method in an application scheduling method according to an embodiment of the present invention;

fig. 4 is a schematic flow chart of a processing method of an algorithm network aware scheduling part in an application scheduling method according to an embodiment of the present invention;

fig. 5 shows a schematic structural diagram of an application scheduling device based on a computing network according to an embodiment of the present invention;

Fig. 6 shows a schematic structural diagram of an application scheduling device according to an embodiment of the present invention.

Detailed Description

Exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. While exemplary embodiments of the present invention are shown in the drawings, it should be understood that the present invention may be embodied in various forms and should not be limited to the embodiments set forth herein.

First, technical terms related to the embodiments of the present invention will be explained.

Computing power is the computing power, and specifically refers to data collection, transmission, computation, and storage capabilities. The magnitude of the calculation force indicates the strength of the digital information processing capability.

A computing power network may refer to a network that connects computing power. All the power supplies in the network can be shared by the demand side of the access network through the power network.

The image may be a lightweight, executable, stand-alone software package that packages software running environments and software developed based on the running environments, and that contains all of the content needed to run a piece of software, including code, runtime, libraries, environment variables, and configuration files.

The computational effort network is used for solving the problem of routing the computational effort network based on ubiquitous computing, but the following problems exist for actually using the computational effort for the running business application:

1. Smooth scheduling problem for business applications: the computational power network perceives and arranges ubiquitous computing resources and optimizes network routing, business applications are difficult to smoothly migrate in different data centers according to the information, and the computational power resources are difficult to truly pull up application instances and provide services for the outside.

2. Data remote access latency problem for stateful services: for stateful services, simple migration application embodiments can introduce latency problems that can be introduced across data centers to access data, and business application usage awareness is degraded.

In view of this, the embodiments of the present invention provide a method, apparatus, device, and computer readable storage medium for scheduling application arrangement based on a computing power network, which can solve the problem of how to enable service applications by ubiquitous computing power in the computing power network through a hierarchical control mechanism, and enable application instances of the service applications to be scheduled smoothly between different data centers. And the problem of time delay caused by access of stateful services across data centers in the power network is solved, and the data are distributed and consumed through data slicing and automatic migration, so that the data consumption efficiency is improved.

Fig. 1 shows a schematic view of a scenario of a computing force process according to an embodiment of the present invention. As shown in fig. 1, a user terminal 11 sends a computational power request to a computational power network-based application orchestration scheduler 12. Application orchestration scheduling device 12 may perform the computational power network-based application orchestration scheduling method according to any of the embodiments of the present invention for computational power requests.

Here, the application orchestration scheduler 12 may include a computing network aware orchestration portion and an application scheduling portion. The computing net perception arrangement part comprises a computing net perception layer and a computing net arrangement layer. The computing network arrangement layer is used for computing power arrangement and network routing planning and can comprise a computing power arrangement module and a network routing module. The computing network perception layer is used for acquiring computing power information such as computing power residual information of each data center, network information such as data center network topology and residual bandwidth in real time.

The application scheduling portion may include a data plane and a control plane. The control plane can adopt a distributed architecture and can comprise an application scheduling center, a data center application controller and a data center data controller, wherein the data center application controller and the data center data controller are deployed in each data center. The application scheduling center may include a primary application scheduler and a primary data manager.

The first-Level application scheduler is used for inquiring the calculation network information according to the calculation power request and the calculation network arrangement layer and determining whether to carry out application migration and data migration according to a Service-Level agent (SLA) and the calculation network information; aggregation of computation results of the same service in different data centers.

And the primary data manager is used for receiving the scheduling instruction of the primary application scheduler and scheduling the data center data controller of the data center.

The data center application controller is used for receiving the scheduling instruction of the first-level application controller, and pulling or destroying the application instance in the data center by pulling the application mirror image in the mirror image warehouse; and reporting the service processing result of the data center to a primary application scheduler.

Here, the data center application controller may control processes including, but not limited to, container and virtual machine processes.

And the data center data controller is used for receiving the scheduling instruction of the primary data manager, slicing the data of the service system in the data center and migrating the data to the data memory of the target data center.

The data plane comprises application instances, data storage, a mirror warehouse and a primary mirror warehouse which are deployed in each data center.

The application instance may be an application process running in a data center, including but not limited to a container or virtual machine process.

Data storage, means for storing business system consumption data, including but not limited to databases, block stores, and network attached stores (Network Attached Storage, NAS).

The mirror image warehouse is deployed in the data center and can be used for storing application mirror images applied by the data center.

The first-level mirror image warehouse is used for storing the full-network application mirror images and plays roles of mirror image backup and centralized management.

In the embodiment, smooth scheduling deployment of the application instance of the service application among different data centers is realized through the hierarchical control mechanism. And by performing fragment migration on the data of the stateful application, the time delay of the stateful application service when the stateful application service is accessed across the data center is reduced. In addition, the computing power network and the application hierarchical scheduling structure supplement the application deployment connotation of the computing power network, and the computing power can be obtained and used really through application hierarchical smooth deployment.

Fig. 2 is a flowchart of an application scheduling method based on a computing power network according to an embodiment of the present invention, where the method is performed by the application scheduling device 12 based on a computing power network shown in fig. 1. As shown in fig. 2, the method comprises the steps of:

step 21: the first-level application scheduler sends a query request to the computational network arrangement layer according to the computational power request, acquires computational power arrangement information and network routing information sent by the computational network arrangement layer, and determines a target data center corresponding to the computational power request.

Here, the calculation request may be a calculation request of a target service application sent by the user side, and may include a calculation requirement of the target service. The query request may be to query currently optimal computational orchestration information and network routing information. The target data center may be one or more data centers in the computing power network where computing resources and storage resources are preferred.

Step 22: and the application controller of the target data center receives the first scheduling instruction sent by the first-stage application scheduler, and pulls the application mirror image from the mirror image warehouse of the target data center.

The application image can be synchronized from the image warehouse of the original data center to the image warehouse of the target data center.

Here, the original data center and the target data center are different data centers. The first scheduling instruction may be for instructing an application controller of the target data center to pull an application image from an image library of the target data center. The application image may be an image of a business application corresponding to the computing force request.

In the embodiment of the invention, the mirror image warehouse of the original data center synchronizes the application mirror image to the first-level mirror image warehouse, and the first-level mirror image warehouse synchronizes the application mirror image to the mirror image warehouse of the target data center. In this way, application mirroring can be smoothed across scheduling deployments between different data centers.

Step 23: the application controller of the target data center creates an application instance according to the application image.

In the above embodiment, the first-level application scheduler sends the query request to the computing network arranging layer according to the computing power request, and obtains the computing power arranging information and the network routing information sent by the computing network arranging layer, so as to determine that the current computing power resource is better, and the computing resource is suitable for executing the computing task, and provide the target data center for external service. The application controller of the target data center receives the first scheduling instruction sent by the first-level application scheduler, pulls an application image from an image warehouse of the target data center, synchronizes the application image from an image warehouse of the original data center to an image warehouse of the target data center, creates an application instance according to the application image, can build a hierarchical control mechanism according to application scheduling, can realize smooth scheduling deployment of the application instance of the service application among different data centers, pulls up the application embodiment and provides services to the outside, and solves the problem of how to enable the service application by ubiquitous computing power in a computing network.

In an optional manner, before the application controller of the target data center receives the first scheduling instruction sent by the primary application scheduler, and pulls the application image from the image bin library of the target data center, the method further includes:

the first-level data manager receives a second scheduling instruction of the first-level application scheduler and sends data slice granularity and target data center information to a data controller of the original data center;

Here, the application types may include stateful applications and stateless applications. A stateful application may be a data center where the application's computational force requests must be submitted to save its relevant state information. The stateless application may be a computational force request of an application that may include all of the information needed to respond to the request.

The second scheduling instructions may be used to instruct the primary data manager to issue data slice granularity and target data center information to the primary data center. The application data may be data stored in a data center by a stateful application. The slice data may be data obtained by slicing application data.

In the embodiment of the invention, a data controller of an original data center receives data slicing granularity and target data center information, executes a data slicing instruction, slices application data corresponding to a calculation force request, and migrates the sliced data to the target data center.

In the embodiment, the data of the stateful application is fragmented and migrated according to the queried computing network information and the application type of the service system, and the problem of time delay caused by accessing the stateful service across the data center in the computing network can be solved by matching with the application instance scheduling, so that the data is consumed in a distributed manner, and the data consumption efficiency is further improved.

Here, the third scheduling instruction may be used to instruct to destroy the original application instance of the original data center. The original application instance corresponding to the computing force request may be an application process corresponding to the computing force request in the original data center.

In the above embodiment, by destroying the original application embodiment of the original data center, the occupation of the computing resources of the data center can be relieved, so that the application scheduling can be participated in later.

In an alternative manner, after the application controller of the target data center creates the application instance from the application image, the method further comprises:

Here, the processing result may be a calculation processing result. In the embodiment of the invention, the target data center can be one or more data centers, and the application controller of the target data center controls the respective application embodiment to consume the corresponding service data to obtain respective processing results and sends the respective processing results to the primary application dispatcher. The first-level application scheduler gathers one or more processing results and performs aggregation processing on the processing results.

It should be noted that, the plurality of processing results are processing results of the same service applied to different data centers.

In the above embodiment, the processing efficiency may be improved by processing the service data in a distributed manner and aggregating the processing results.

In order to fully understand an application scheduling method based on a computing network, fig. 3 shows a flow diagram of an application calling part processing method in the application scheduling method according to an embodiment of the present invention, and as shown in fig. 3, the application calling part processing method includes the following steps:

step 31, a primary application scheduler of an application scheduling center receives a calculation force request of a target service application sent by a user side;

step 32, the first-level application dispatcher initiates a query operation (namely, sends a query request) to the calculation network arrangement layer according to the calculation power request to acquire calculation network information;

the computing network information may include currently optimal computing power arrangement information and network routing information, among other things.

In step 33, the primary application scheduler detects whether the computational power request is a stateful application. If the calculation force request is a stateful application, the process proceeds to step 34. In the case that the calculation force request is a stateless application, the flow proceeds to step 37.

At step 34, the primary application dispatcher dispatches application controllers deployed within the target data center.

And 35, pulling the application mirror image from the mirror image warehouse by the application controller of the target data center.

At step 36, the application controller of the target data center deploys the application embodiment at the target data center.

Here, the application controller of the target data center creates an application instance according to the application image, the application instance is pulled up in the data center, and the application controller of the original data center destroys the original application instance.

Step 37, the primary application scheduler decides the data slicing granularity according to the user SLA and the calculation network information.

Step 38, the primary application scheduler schedules the primary data manager and issues data slice granularity information and target data center information.

At step 39, the target data center application controller pulls the application image.

In step 310, the original data center data controller executes the data slicing instruction and migrates the sliced data to the target data center.

In step 311, the application controller of each data center deploys and destroys the application embodiments according to the instructions.

Here, the application controller of the target data center pulls up the application instance in the data center, and the application controller of the original data center destroys the original application instance.

In step 312, the application instance in each data center consumes the corresponding service data, and gathers the processing results to the primary application scheduler.

In step 313, the primary application scheduler aggregates the processing results of each data center.

In an alternative manner, before the primary application scheduler sends the query request to the algorithm orchestration layer according to the algorithm power request, the method further comprises:

the computing network sensing layer acquires real-time sensing information of each data center and registers the real-time sensing information to the computing network arrangement layer.

Here, the real-time awareness information may include, but is not limited to, computing power information and network information. The computing force information may include, but is not limited to, computing force remaining information of the data center. The network information may include, but is not limited to, network topology information, residual bandwidth information, network device information, and the like of the data center.

In an alternative manner, after the primary application scheduler sends the query request to the computing network orchestration layer according to the computing power request, the method further comprises:

the computing network arrangement layer performs computing power arrangement and network route planning according to the computing power information and the network information to obtain computing power arrangement information and network route information, and sends the computing power arrangement information and the network route information to the primary application scheduler.

Here, the calculation force arrangement information may include a calculation force arrangement result. The network routing information may include routing path information to the target data center.

In order to fully understand the application programming and scheduling method based on the computing network, fig. 4 shows a flow chart of a processing method of a computing network aware-programming part in the application programming and scheduling method according to an embodiment of the present invention, and as shown in fig. 4, the processing method of the computing network aware-programming part may include the following steps.

Step 41, computing power and network information awareness.

Here, the algorithm-aware layer deployed within the data center perceives the data center residual computing power, network device information, and network bandwidth residual information in real time (i.e., the algorithm-aware layer perceives real-time perceived information).

And step 42, reporting the calculation power and the network information.

Here, the computing network awareness layer registers real-time awareness information to the computing network orchestration layer.

Step 43, calculation effort arrangement and network route planning.

After receiving the inquiry request, the computing network arrangement layer performs computing power arrangement and network route planning according to the real-time sensing information to obtain computing power arrangement information and network route information, and synchronizes the computing power arrangement result and the network route information to the application scheduling center.

Here, the scheduling case may be a case of scheduling application images. The primary mirror repository may be used to manage mirror information for the mirror repository for each data center.

In the embodiment of the application, the mirror image warehouse of each data center synchronizes the scheduled application mirror image to the mirror image warehouse of the target data center according to the condition that the primary application scheduler schedules the application mirror image. And uploading the synchronized application images to the primary image warehouse by the image warehouse of each center data so that the primary image warehouse backs up the synchronized application images.

Fig. 5 shows a schematic structural diagram of an application scheduling device based on a computing network according to an embodiment of the present invention. As shown in fig. 5, the apparatus 50 includes: a primary application scheduler 51, an application controller 52 of the target data center.

The primary application scheduler 51 is configured to send a query request to the computing network arrangement layer according to the computing power request, obtain computing power arrangement information and network routing information sent by the computing network arrangement layer, and determine a target data center corresponding to the computing power request.

An application controller 52 of the target data center, configured to receive a first scheduling instruction sent by the primary application scheduler, pull an application image from an image warehouse of the target data center, and synchronize the application image from an image warehouse of the original data center to an image warehouse of the target data center; an application instance is created from the application image.

In an alternative, the apparatus 50 further includes a primary data manager and a data controller of the primary data center.

The primary application scheduler 51 is further configured to determine an application type of the computing power request before the application controller of the target data center receives the first scheduling instruction sent by the primary application scheduler and pulls the application image from the image bin library of the target data center; in the case of a stateful application, the data slice granularity is determined according to the user service level protocol and the gateway information.

In an alternative, the apparatus 50 also includes an application controller of the original data center.

In an alternative, the apparatus 50 further includes a mirror warehouse and a primary mirror warehouse of the data center.

In an alternative, the apparatus 50 further includes a calculation network awareness layer and a calculation network orchestration layer.

Fig. 6 shows a schematic structural diagram of an application scheduling device according to an embodiment of the present invention, and the specific embodiment of the present invention is not limited to a specific implementation of the application scheduling device.

As shown in fig. 6, the application orchestration scheduling device may include: a processor 602, a communication interface (Communications Interface), a memory 606, and a communication bus 608.

Wherein: processor 602, communication interface 604, and memory 606 perform communication with each other via communication bus 608. Communication interface 604 is used to communicate with network elements of other devices, such as clients or other servers. The processor 602 is configured to execute the program 610, and may specifically perform the relevant steps in the above-described embodiment of the method for scheduling application orchestration based on a computing network.

In particular, program 610 may include program code comprising computer-executable instructions.

The processor 602 may be a central processing unit CPU or a specific integrated circuit ASIC (Application Specific Integrated Circuit) or one or more integrated circuits configured to implement embodiments of the present invention. The one or more processors comprised by the application orchestration scheduling device may be the same type of processor, such as one or more CPUs; but may also be different types of processors such as one or more CPUs and one or more ASICs.

A memory 606 for storing a program 610. The memory 606 may comprise high-speed RAM memory or may further comprise non-volatile memory (non-volatile memory), such as at least one disk memory.

Program 610 may be specifically invoked by processor 602 to cause an application orchestration scheduling device to:

In an alternative, program 610 may be specifically invoked by processor 602 to cause an application orchestration scheduling device to:

The embodiment of the invention provides a computer readable storage medium, wherein the storage medium stores at least one executable instruction, and the executable instruction enables an application scheduling device to execute the application scheduling method based on the computing power network in any method embodiment when the executable instruction runs on the application scheduling device.

The executable instructions may be specifically for causing an application orchestration scheduling device to:

Embodiments of the present invention provide a computer program that is callable by a processor to cause an application orchestration scheduling device to perform the method of the power network-based application orchestration scheduling method of any of the method embodiments described above.

An embodiment of the present invention provides a computer program product comprising a computer program stored on a computer readable storage medium, the computer program comprising program instructions which, when run on a computer, cause the computer to perform the method for scheduling application orchestration based on a computer power network in any of the method embodiments described above.

The algorithms or displays presented herein are not inherently related to any particular computer, virtual system, or other apparatus. Various general-purpose systems may also be used with the teachings herein. The required structure for a construction of such a system is apparent from the description above. In addition, embodiments of the present invention are not directed to any particular programming language. It will be appreciated that the teachings of the present invention described herein may be implemented in a variety of programming languages, and the above description of specific languages is provided for disclosure of enablement and best mode of the present invention.

In the description provided herein, numerous specific details are set forth. However, it is understood that embodiments of the invention may be practiced without these specific details. In some instances, well-known methods, structures and techniques have not been shown in detail in order not to obscure an understanding of this description.

Similarly, it should be appreciated that in the above description of exemplary embodiments of the invention, various features of the embodiments of the invention are sometimes grouped together in a single embodiment, figure, or description thereof for the purpose of streamlining the disclosure and aiding in the understanding of one or more of the various inventive aspects. However, the disclosed method should not be construed as reflecting the intention that: i.e., the claimed invention requires more features than are expressly recited in each claim.

Those skilled in the art will appreciate that the modules in the apparatus of the embodiments may be adaptively changed and disposed in one or more apparatuses different from the embodiments. The modules or units or components of the embodiments may be combined into one module or unit or component, and they may be divided into a plurality of sub-modules or sub-units or sub-components. Any combination of all features disclosed in this specification (including any accompanying claims, abstract and drawings), and all of the processes or units of any method or apparatus so disclosed, may be used in combination, except insofar as at least some of such features and/or processes or units are mutually exclusive. Each feature disclosed in this specification (including any accompanying claims, abstract and drawings), may be replaced by alternative features serving the same, equivalent or similar purpose, unless expressly stated otherwise.

It should be noted that the above-mentioned embodiments illustrate rather than limit the invention, and that those skilled in the art will be able to design alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses shall not be construed as limiting the claim. The word "comprising" does not exclude the presence of elements or steps not listed in a claim. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The invention may be implemented by means of hardware comprising several distinct elements, and by means of a suitably programmed computer. In the unit claims enumerating several means, several of these means may be embodied by one and the same item of hardware. The use of the words first, second, third, etc. do not denote any order. These words may be interpreted as names. The steps in the above embodiments should not be construed as limiting the order of execution unless specifically stated.

Claims

1. An application orchestration scheduling method based on a computing network, the method comprising:

an application controller of a target data center receives a first scheduling instruction sent by the primary application scheduler, pulls an application image from an image warehouse of the target data center, and synchronizes the application image from an image warehouse of an original data center to an image warehouse of the target data center;

and the application controller of the target data center creates an application instance according to the application image.

2. The method of claim 1, wherein before the application controller of the target data center receives the first scheduling instruction sent by the primary application scheduler, the method further comprises, before pulling an application image from a mirror library of the target data center:

the primary application scheduler determines the application type of the computing force request;

in the case that the application type is a stateful application, the primary application scheduler decides a data slice granularity according to a user service level protocol and a computing network information;

And the primary data manager receives the second scheduling instruction of the primary application scheduler and sends the data slice granularity and the target data center information to the data controller of the original data center.

And the data controller of the original data center receives the data slicing granularity and the target data center information, slices the application data corresponding to the computing force request, and migrates the sliced data to the target data center.

3. The method of claim 1, wherein after the application controller of the target data center receives the first scheduling instruction sent by the primary application scheduler, the method further comprises, after pulling the application image from the mirror library of the target data center:

and the application controller of the original data center receives a third scheduling instruction sent by the primary application scheduler and destroys the original application instance corresponding to the computing force request.

4. The method of claim 2, wherein after the application controller of the target data center creates an application instance from the application image, the method further comprises:

the application controller of the target data center controls the application embodiment to consume corresponding service data to obtain a processing result;

The application controller of the target data center sends the processing result to the primary application scheduler;

and the primary application scheduler receives the processing result and performs aggregation processing on the processing result.

5. The method of any of claims 1 to 4, wherein before the application controller of the target data center receives the first scheduling instruction sent by the primary application scheduler, the method further comprises, before pulling an application image from a mirror library of the target data center:

6. The method of claim 1, wherein before the primary application scheduler sends a query request to the algorithm orchestration layer according to the computational power request, the method further comprises:

the computing network sensing layer acquires real-time sensing information of each data center and registers the real-time sensing information to the computing network arranging layer, wherein the real-time sensing information comprises computing power information and network information;

After the primary application scheduler sends a query request to the computing network orchestration layer according to the computing power request, the method further comprises:

and the computing network arrangement layer performs computing force arrangement and network route planning according to the computing force information and the network information to obtain computing force arrangement information and network route information, and sends the computing force arrangement information and the network route information to the primary application dispatcher.

7. An application orchestration scheduling device based on a computing network, the device comprising:

the first-level application scheduler is used for sending a query request to the computational network arrangement layer according to the computational power request, acquiring computational power arrangement information and network routing information sent by the computational network arrangement layer, and determining a target data center corresponding to the computational power request;

an application controller of a target data center is used for receiving a first scheduling instruction sent by the primary application scheduler, pulling an application mirror image from a mirror image warehouse of the target data center, and synchronizing the application mirror image from a mirror image warehouse of an original data center to a mirror image warehouse of the target data center; and creating an application instance according to the application image.

8. The apparatus of claim 7, wherein the apparatus further comprises:

The computing network perception layer is used for acquiring real-time perception information of each data center and registering the real-time perception information to the computing network arrangement layer, wherein the real-time perception information comprises computing power information and network information;

the computing network arrangement layer is used for carrying out computing power arrangement and network route planning according to the computing power information and the network information after the primary application scheduler sends a query request to the computing network arrangement layer according to the computing power request, obtaining computing power arrangement information and network route information, and sending the computing power arrangement information and the network route information to the primary application scheduler.

9. An application orchestration scheduling device, comprising: the device comprises a processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface complete communication with each other through the communication bus;

the memory is configured to store at least one executable instruction that causes the processor to perform the operations of the power network-based application orchestration scheduling method according to any one of claims 1-6.

10. A computer readable storage medium, wherein at least one executable instruction is stored in the storage medium, which when run on an application orchestration scheduling device, causes the application orchestration scheduling device to perform the operations of the computing-network-based application orchestration scheduling method according to any one of claims 1-6.