CN112269647A

CN112269647A - Node scheduling, switching and coordinating method and corresponding device, equipment and medium thereof

Info

Publication number: CN112269647A
Application number: CN202011157270.5A
Authority: CN
Inventors: 曾仰鑫; 陈建文; 叶键晖
Original assignee: Guangzhou Huaduo Network Technology Co Ltd
Current assignee: Guangzhou Huaduo Network Technology Co Ltd
Priority date: 2020-10-26
Filing date: 2020-10-26
Publication date: 2021-01-26

Abstract

The application discloses a node scheduling, switching and coordinating method and corresponding devices, equipment and media thereof, wherein the scheduling method comprises the following steps: responding to a node query request, acquiring a specified task type, querying an access entry of a corresponding service node from a node database as feedback of the request, wherein each service node is used for scheduling and executing nodes for tasks of the type corresponding to the service node; during the period that the service node corresponding to the task type cannot be inquired in the node database, responding to a role switching instruction of an execution node, storing an access entry of the execution node and the task type as a data record of the node database, and upgrading the execution node to be the service node; and monitoring the connection state of all service nodes, and clearing the data record of the service node from the node database after the service node is disconnected. The multi-center distributed task scheduling system constructed according to the technical scheme of the application can schedule various types of tasks more efficiently and orderly.

Description

Node scheduling, switching and coordinating method and corresponding device, equipment and medium thereof

Technical Field

The present application relates to the field of computer networks, and in particular, to a method for scheduling, switching, and coordinating nodes, and to an apparatus, a device, and a nonvolatile storage medium corresponding to the methods.

Background

The system scale of each large internet company is gradually developed, the number of tasks related to services is gradually increased and becomes more and more complex, and particularly, in a distributed environment, when a plurality of service systems exist, each service system has a task processing requirement, so that a task platform is required to uniformly manage and schedule the tasks of each service system, namely, a distributed task scheduling system, so as to respond to a large number of task processing requests of each type of service system in time and meet the task processing requirements of corresponding services.

Currently, the execution logic of most distributed task scheduling systems is roughly: firstly, a unique scheduling node is configured through a scheduling center, and tasks are distributed to execution nodes by the scheduling node so that the corresponding execution nodes can execute the distributed tasks; and secondly, for the execution of the task, periodically scanning task configuration records by a plurality of task execution nodes to preempt unexecuted tasks and execute the tasks. According to this execution logic, there are at least the following drawbacks:

first, in the manner of configuring scheduling nodes by a scheduling center in the prior art, even if the scheduling center adopts a high availability scheme, because the scheduling center has a single structure of the scheduling center, that is, only one scheduling center node providing scheduling service exists, if the scheduling center node has a fault such as downtime, the system may be unavailable for a period of time and may need to be restored to an available state for a long time, so that a large number of tasks cannot be executed, and various types of services cannot be provided for users.

Secondly, in the processing method related to task execution in the prior art, because the availability of the execution nodes is not considered, the same task is directly and unconditionally preempted by the multiple execution nodes, so that task conflict is easily caused, the multiple execution nodes repeatedly execute the same task, and because the task loads of different execution nodes are unbalanced, the greater processing efficiency cannot be realized reasonably and efficiently in practice.

In view of the above problems generally faced by distributed task scheduling systems, it is imperative to provide some effective solutions.

Disclosure of Invention

The first objective of the present application is to provide a node scheduling method and a corresponding device, electronic device, and non-volatile storage medium.

A second object of the present application, which is adapted to the first object, is to provide a node switching method and a corresponding apparatus, electronic device, and non-volatile storage medium.

A third object of the present application, which is adapted to the first object, is to provide a node coordination method and a corresponding apparatus, electronic device, and non-volatile storage medium.

In order to meet various purposes of the application, the following technical scheme is adopted in the application:

a node scheduling method adapted to the primary object of the present application, includes the following steps:

responding to a node query request, acquiring a specified task type, querying and determining an access entry of a service node corresponding to the task type from a node database, and serving as feedback to the request, wherein each service node is used for scheduling a specific execution node for a task of the corresponding type;

when the service node corresponding to the task type is not inquired in the node database, responding to the role switching instruction of any execution node, and storing the access entry of the execution node and the task type in a correlation manner as the data record of the node database so as to upgrade the execution node to the service node;

and monitoring the connection state of any service node, and clearing the data record corresponding to the service node from the node database after the connection of any service node is disconnected.

In a further embodiment, the node query request is originated from the executing node, and the executing node correspondingly issues the role switching instruction when there is no service node feedback.

In a preferred embodiment, the node query request originates from a task agent node, and the fed back service node schedules a task operation request of the task for the agent node.

In some embodiments, the node database includes data records of a plurality of different service nodes, each service node corresponds to one task type, and only one service node corresponds to one task type at the same time.

In a further embodiment, the steps of the method are performed concurrently.

A node switching method applicable to the second object of the present application includes the following steps:

initiating a node query request, querying a service node matched with a task type preset by a current node from a node database, and acquiring a corresponding feedback result;

when the feedback result indicates that no data record of the corresponding service node exists in the node database, initiating a role switching instruction to store an access entry of the current node and a preset task type in the node database in a correlation manner for query, so as to upgrade the current node into the service node;

after the current node is upgraded to a service node, receiving a task operation request corresponding to the task type, and adding the task operation request into a task pool for the execution node to seize and process;

and when the feedback result contains the access entrance of the matched service node, keeping the current node as the execution node for processing the task operation request of the preset task type.

In a further embodiment, the current node performs the following steps in the role of the serving node:

receiving a task operation request which is sent by an agent node and accords with a preset task type, and adding the task operation request into a task pool;

selecting part of execution nodes for processing the task operation request according to load balancing information reported by the execution nodes matched with the preset task type;

sending a task message informing processing of the task operation request to the partial execution node;

preempting processing of the task operation request by a first responder of the partial execution nodes.

continuously uploading self load balancing information to a service node matched with the self preset task type;

when receiving a task message which is sent by the service node and used for notifying the processing of task operation requests, preempting at least one task operation request from the task pool;

and performing task processing on the successfully preempted task operation request.

A node coordination method applied to the third object of the present application includes the following steps:

the method comprises the steps that a plurality of task types are adapted, a plurality of service nodes of each task type compete for scheduling rights of tasks of the type, and a service node which is used for determining to obtain the scheduling rights is the service node of the type and is used for scheduling specific execution nodes for the tasks of the type;

each determined service node selects part of execution nodes for the tasks requested to be processed by the agent node according to the load balancing information reported by the execution nodes related to the service node on the task type;

the determined service node sends a task message about the task requesting processing to the partial execution node;

and the determined service node schedules the task requesting processing to a winner competing for the execution right of the task requesting processing in the partial execution nodes to implement the task processing.

A node scheduling apparatus adapted to the first objective of the present application, comprising:

the request response unit is used for responding to the node query request, acquiring the specified task type, querying and determining an access entrance of a service node corresponding to the task type from a node database as feedback to the request, and each service node is used for scheduling a specific execution node for the task of the corresponding type;

the instruction response unit is used for responding to the role switching instruction of any execution node during the period that the service node corresponding to the task type is not inquired in the node database, and storing the access entry of the execution node and the task type in a correlation manner as a data record of the node database so as to upgrade the execution node to the service node;

and the node monitoring unit is used for monitoring the connection state of any service node, and clearing the data record corresponding to the service node from the node database after the connection of any service node is disconnected.

A node switching apparatus adapted to the second object of the present application, comprising:

the request initiating unit is used for initiating a node query request, querying a service node matched with a task type preset by a current node from a node database, and acquiring a corresponding feedback result;

an instruction initiating unit, configured to, when the feedback result indicates an access entry of a service node of a node, maintain a current node as the execution node for processing that no data record of a corresponding service node exists in the preset point database, initiate a role switching instruction, so as to store the access entry of the current node and a preset task type in the node database in association for query, so as to upgrade the current node to the service node;

the request adding unit is used for receiving a task operation request corresponding to the task type after the current node is upgraded to a service node, and adding the task operation request into a task pool for the execution node to seize and process;

and the feedback determining unit is used for keeping the current node as the execution node to process the task operation request of the preset task type when the feedback result contains the access entrance of the matched service node.

A node coordination apparatus adapted to the third object of the present application includes:

the node determining unit is used for adapting to a plurality of task types, a plurality of service nodes of each task type compete for the scheduling weight of the task serving the task type, and a service node used for scheduling a specific execution node for the task of the corresponding type is correspondingly determined;

the node screening unit is used for optimizing a part of execution nodes for the task requested to be processed by the agent node according to the load balancing information reported by the execution nodes related to the service nodes on the task type by each determined service node;

the task initiating unit is used for sending a task message about the task requesting processing to the partial execution node by the determined service node;

and the task scheduling unit is used for scheduling the task requesting processing to a winner competing for the execution right of the task requesting processing in the partial execution nodes by the determined service node to implement task processing.

A further object of the present application is to provide an electronic device, which includes a central processing unit and a memory, wherein the central processing unit is used to invoke and run a computer program stored in the memory to execute the steps of the node scheduling method, the node switching method, or the node coordination method.

A non-volatile storage medium storing a computer program according to the node scheduling method, the node switching method, or the node coordinating method, wherein the computer program executes the steps included in the corresponding method when the computer program is called by a computer.

Compared with the prior art, the application has the following advantages:

firstly, the application allows tasks to be classified and processed according to different types, a plurality of service nodes are equipped to adapt to different task types, each service node plays the role of a dispatching center and is responsible for dispatching corresponding execution nodes for processing tasks of corresponding types, wherein, the service node and the execution node are provided with a uniform operation logic, when the logic is operated in a certain server, the server has the capability of writing self information into the node database firstly and preempting the scheduling center which becomes a certain type of task, i.e. becomes said service node, and when the service node of the corresponding task type, i.e. the scheduling center, already exists in the node database, the (node) server can make itself work with the executing node, and is responsible for receiving the scheduling of the service node corresponding to the type and matching with the task of the corresponding type. Therefore, by applying the principle, a plurality of servers are deployed, and unified management is performed by virtue of the node database, so that the distributed task scheduling system of a multi-scheduling center can be realized, and compared with the distributed task scheduling system of a single center which is not divided into task types, the distributed task scheduling system has stronger adaptability, even if a service node serving a certain task type goes down, the corresponding task scheduling of the service node is influenced in a short time, other servers can be quickly switched from the role of an execution node to the role of the service node for replacement, the current fault is quickly repaired, the disaster tolerance capability is strong, and the breakdown of the whole scheduling system cannot be caused.

Secondly, the method realizes a role automatic switching mechanism by utilizing the operation of the same set of operation logic on the same server, when a node database responds to a node query request of an execution node and a service node corresponding to the node query request cannot be queried in the database, a role switching instruction sent by the execution node is responded so as to switch the execution node into the service node, and by the principle, the same server has two working modes, namely, one node server can be responsible for the work of the execution node and the service node, so that the node server has mobility, can flexibly and automatically switch the working modes, and can improve the deployment and maintenance efficiency of a scheduling system.

In addition, when the service nodes of each task type are determined, the service nodes of each task type are recorded by the node database, any node server is allowed to write a role switching instruction which is willing to work on the service node into the node database, and then the node server which is responded firstly is allowed to become a corresponding dispatching center according to the first-come-first-serve principle, therefore, a competition mechanism is realized, the competition mechanism is applicable, each node server under the same task type can be converted into a service node when needed, i.e., becomes the dispatch center, and thus, for the entire distributed task scheduling system deployed in the solution of the present application, because each node server is given fair competition opportunity, the node server with urgent requirement can obtain the scheduling right more easily, and the time efficiency of the scheduling system when self-coordination is carried out in the scheduling system is also improved.

Drawings

The foregoing and/or additional aspects and advantages of the present application will become apparent and readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings of which:

FIG. 1 is a schematic representation of a deployment of the subject technology in a network environment;

FIG. 2 is a schematic diagram of a system principle of a multi-center distributed task scheduling system according to an embodiment of the present disclosure;

fig. 3 is a flowchart illustrating an exemplary embodiment of a node scheduling method according to the present application;

FIG. 4 is a functional block diagram of an exemplary embodiment of a node scheduling apparatus of the present application;

fig. 5 is a flowchart illustrating an exemplary embodiment of a node switching method according to the present application;

fig. 6 is a flowchart illustrating another embodiment of a node switching method according to the present application;

fig. 7 is a flowchart illustrating a node switching method according to another embodiment of the present application;

FIG. 8 is a functional block diagram of an exemplary embodiment of a node switching apparatus of the present application;

FIG. 9 is a schematic flow chart diagram illustrating an exemplary embodiment of a node coordination method of the present application;

fig. 10 is a schematic block diagram of an exemplary embodiment of a node coordination apparatus according to the present application.

Detailed Description

Reference will now be made in detail to embodiments of the present application, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to the same or similar elements or elements having the same or similar function throughout. The embodiments described below with reference to the drawings are exemplary only for the purpose of explaining the present application and are not to be construed as limiting the present application.

As used herein, the singular forms "a", "an", "the" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms "comprises" and/or "comprising," when used in this specification, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof. It will be understood that when an element is referred to as being "connected" or "coupled" to another element, it can be directly connected or coupled to the other element or intervening elements may also be present. Further, "connected" or "coupled" as used herein may include wirelessly connected or wirelessly coupled. As used herein, the term "and/or" includes all or any element and all combinations of one or more of the associated listed items.

It will be understood by those within the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the prior art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.

As will be appreciated by those skilled in the art, "client," "terminal," and "terminal device" as used herein include both devices that are wireless signal receivers, which are devices having only wireless signal receivers without transmit capability, and devices that are receive and transmit hardware, which have receive and transmit hardware capable of two-way communication over a two-way communication link. Such a device may include: cellular or other communication devices such as personal computers, tablets, etc. having single or multi-line displays or cellular or other communication devices without multi-line displays; PCS (Personal Communications Service), which may combine voice, data processing, facsimile and/or data communication capabilities; a PDA (Personal Digital Assistant), which may include a radio frequency receiver, a pager, internet/intranet access, a web browser, a notepad, a calendar and/or a GPS (Global Positioning System) receiver; a conventional laptop and/or palmtop computer or other device having and/or including a radio frequency receiver. As used herein, a "client," "terminal device" can be portable, transportable, installed in a vehicle (aeronautical, maritime, and/or land-based), or situated and/or configured to operate locally and/or in a distributed fashion at any other location(s) on earth and/or in space. The "client", "terminal Device" used herein may also be a communication terminal, a web terminal, a music/video playing terminal, such as a PDA, an MID (Mobile Internet Device) and/or a Mobile phone with music/video playing function, and may also be a smart tv, a set-top box, and the like.

The hardware referred to by the names "server", "client", "service node", etc. is essentially an electronic device with the performance of a personal computer, and is a hardware device having necessary components disclosed by the von neumann principle such as a central processing unit (including an arithmetic unit and a controller), a memory, an input device, an output device, etc., a computer program is stored in the memory, and the central processing unit calls a program stored in an external memory into the internal memory to run, executes instructions in the program, and interacts with the input and output devices, thereby completing a specific function.

It should be noted that the concept of "server" as referred to in this application can be extended to the case of a server cluster. According to the network deployment principle understood by those skilled in the art, the servers should be logically divided, and in physical space, the servers may be independent from each other but can be called through an interface, or may be integrated into one physical computer or a set of computer clusters. Those skilled in the art will appreciate this variation and should not be so limited as to restrict the implementation of the network deployment of the present application.

Referring to FIG. 1, the hardware basis required for implementing the embodiments of the present invention may be deployed with reference to the architecture shown in the drawings. The service node 80, the agent node 81 and the execution node 82 are all deployed at the cloud, wherein the service node 80 is dedicated to receiving task operation requests sent by the agent node and receiving load balancing information reported by the execution node and other interactions. Each node 80, 81 and 82 may act as an application server and further connect to other servers such as storage servers, backup servers, to interact with the node database to form a logically associated server cluster. The service node 80, the agent node 81 and the enforcement node 82 all have access to the internet by means of well-known network access methods, and data communication links can be established among each other so as to realize various data exchanges.

Referring to fig. 2, a system structure of a multi-center distributed task scheduling system constructed by implementing the related technical solution of the present application is shown in the figure, and the system is composed of a node database a, an executing node B, a service node C, and a proxy node D. The node database A is a first layer of the system, the execution node B, the service node C and the agent node are positioned on a second layer of the system, and the terminal equipment D is positioned on a third layer of the system. The node database in the first layer provides corresponding services for all devices in the second layer, such as providing a service of querying a service node for the executing node B and the proxy node D, providing a role switching service for the executing node B, and monitoring the connection state of the service node C in real time to update the data records in the database in real time. The respective devices located at the second layer establish data communication links with each other to provide respective services for each other, for example, the executing node B may send its own load balancing information to the serving node C so that the serving node C sends the task operation request sent by the proxy node D thereto, and the executing node B may feed back a respective result to the proxy node D after completing the task operation request. And the terminal equipment E on the third layer and the proxy node D establish a data communication link so as to upload the tasks to be processed to the multi-center distributed task scheduling system for processing through the proxy node D, and the proxy node D can feed back the execution result of the tasks to the terminal equipment E. It should be noted that when the executing node B switches to the serving node through the role switch command, the executing logic thereof may refer to the executing logic of the serving node C, which interacts with other devices substantially identically. The schematic diagram only shows the single existence of each device, when the system is actually deployed, a plurality of execution nodes B, service nodes C, agent nodes D, and terminal devices E in fig. 2 exist, and the node database a does not necessarily exist in a single server, and may also be stored in a distributed manner, or multiple terminals are mutually backed up and synchronized in real time, as long as it can provide services for corresponding devices, and those skilled in the art can design flexibly, as long as the system can operate according to the method implemented in the present application.

The terminal device of the application is provided with a related operating system, such as an IOS (operating system), an HMS (grand Mongolian language), an Android and other operating systems providing equivalent functions, under the support of the operating system, data communication can be performed between the server and the terminal device through a communication link agreed with each other, application programs developed for the operating systems can also run normally in the terminal device, and human-computer interaction and remote interaction are realized through the terminal device.

The application program of the present application is generally referred to as an application program running on an electronic device such as a server or a terminal device, and the application program implements the related technical solution of the present application in a programming manner, and a program code of the application program can be saved in a non-volatile storage medium that can be identified by a computer in the form of a computer executable instruction, and is called into a memory by a central processing unit to run, and the related apparatus of the present application is constructed by running the application program on the computer.

The person skilled in the art will know this: although the various methods of the present application are described based on the same concept so as to be common to each other, they may be independently performed unless otherwise specified. In the same way, for each embodiment disclosed in the present application, it is proposed based on the same inventive concept, and therefore, concepts of the same expression and concepts of which expressions are different but are appropriately changed only for convenience should be equally understood.

Referring to fig. 3, a node scheduling method of the present application is executed by an application program of a server carrying a node database, and in an exemplary embodiment, the method includes the following steps:

step S11, in response to the node query request, acquiring the specified task type, querying from the node database to determine the access entry of the service node corresponding to the task type, as a feedback to the request, where each service node is configured to schedule a specific execution node for the task of the type corresponding to the service node:

the node database is a database storing data records related to information of service nodes, so as to respond to the node query request, acquire a task type specified in the request, query the service node corresponding to the task type, send an access entry of the service node to a corresponding execution node or proxy node as feedback of the node query request, and complete the operation of scheduling the corresponding service node for the execution node or proxy node.

The node query request is sent to the node database by the execution node or the proxy node, so that the node database queries the service node corresponding to the task type according to the task type specified by the node query request, and if the service node corresponding to the task type exists, the node database correspondingly sends the access entry of the service node to the execution node or the proxy node, so that the execution node or the proxy node accesses the service node to perform task scheduling. If the data record of the service node corresponding to the task type does not exist in the node database, the node database sends a corresponding feedback result to the execution node, so that the execution node sends a role switching instruction to the execution node according to the feedback result.

The node database is a database used for storing data records related to a plurality of different service nodes, is deployed in a corresponding server and provides data services for determining the corresponding relation between tasks of various task types and the service nodes thereof so that an execution node or an agent node can inquire the corresponding service node in the node database through a node inquiry request; each service node stored in the node database can only schedule the corresponding task type, and only one service node is corresponding to one task type at the same time, so that the feedback from the control node database to the proxy node only comprises an access entry of the corresponding service node, and the proxy node is prevented from sending task operation requests to a plurality of service nodes, and tasks required to be processed by the proxy node are repeatedly processed.

The data record refers to data of the service node information stored in the node database, and the data includes configuration information of the service node itself, such as an access entry of the service node, a scheduling task type and the like, so that the node database determines the service node scheduled by the corresponding task type through the stored data record, and sends the information of the access entry of the service node and the like to the corresponding execution node.

When the node query request comes from an executing node, the executing node continuously uploads the server address and the load balancing information to the service node through the access entry, so that the service node adds the server address and the load balancing information to a corresponding database, and selects at least one executing node for processing a task corresponding to the task operation request by screening the executing nodes existing in the database.

When the executing node is in a state of not full of load, a node query request of a corresponding task is sent to the node database at regular time, a corresponding service node is queried, a corresponding task is obtained for processing, and the computing resources of the executing node are reasonably and efficiently utilized.

The proxy node generally refers to a server that receives a task to be processed sent by a client, and when the client performs a corresponding task operation, the proxy node sends the task to be processed to the corresponding proxy server, so that the proxy node sends the node query request to a node database, acquires an access entry of a corresponding service node from the node database, and sends a corresponding task operation request to the service node through the access entry, so that the service node sends the task operation request to a corresponding partial execution node to complete scheduling of the task.

After the executing node sends the node query request to the node database, when the received feedback result does not include the access entry of the corresponding service node, a role switching instruction of the corresponding task type is sent to the node database so as to switch the executing node to the service node corresponding to the task type, and the task of the type is scheduled, and for the implementation mode of the role switching, reference is made to the description of the subsequent steps.

Step S12, when the service node corresponding to the task type is not found in the node database, in response to the role switching instruction of any execution node, storing the access entry of the execution node and the task type in association as the data record of the node database, so that the execution node is upgraded to the service node:

when the node database responds to the node query request sent by the execution node and does not query the service node corresponding to the task type corresponding to the node query request, the feedback result sent to the execution node does not include the access entry of the service node, and the execution node sends a role switching instruction to the node database, so that the node database responds to the role switching instruction and stores the information related to the access entry of the execution node and the task type in the data record of the node database, and the execution node is switched to the service node of the task type corresponding to the node query request.

When the node database receives a plurality of role switching instructions at the same time, the node database responds to the role switching instruction received firstly, and stores the access entry and the task type information of the execution node to which the role switching instruction belongs as data records so as to determine the task type of the execution node and send the access entry of the execution node to the corresponding execution node, thereby completing the upgrading of the execution node into the service node of the corresponding task type.

And after the executing node is switched to the service node, continuously receiving load balancing information of the executing node of the task type corresponding to the node query request, receiving a task operation request corresponding to the task type sent by the proxy node, and adding the task operation request into a task pool so as to schedule at least one corresponding executing node for the task operation request, so that the task corresponding to the task operation request is processed by the executing node.

The role switching module constructed in the step efficiently utilizes the computing resources of the executing node, and simultaneously can provide a solution for a certain task type when the corresponding service node is in a down state, when the node database responds to the node query request sent by the executing node and queries that the corresponding service node is in the down state, the node database cannot feed back the executing node, so that the executing node sends a role switching instruction to the node database, the node database responds to the computing switching instruction and switches the computing switching instruction into the service node of the corresponding task type, task scheduling of the task type is carried out, and task scheduling service of the task type is repaired.

Step S13, monitoring the connection status of any service node, and when any service node is disconnected, clearing the data record corresponding to the service node from the node database:

in order to monitor the current running state of each service node and immediately discover and process the service nodes in the downtime state, the node database can monitor the connection state of the service nodes stored in the node database, when the disconnection of the service nodes reaches a preset time or the service nodes do not respond to related requests for many times, the node database indicates that the service nodes are in the downtime state, the node database can clear the data records of the service nodes so as to send the execution nodes of the node query requests corresponding to the service nodes, and the task scheduling of processing the tasks for the new service nodes is upgraded by sending role switching instructions.

All the steps realized by the method are executed concurrently, namely the node database can concurrently execute the response of the node query request, the response of the role switching instruction and the connection state of the service nodes in the monitoring node database, and the data records of the node database in the corresponding server are updated in real time through a network link, so that a multi-center distributed task scheduling system with a multi-thread working mode is constructed.

Further, a node scheduling apparatus of the present application may be constructed by functionalizing the steps in the methods disclosed in the above embodiments, and according to this idea, please refer to fig. 4, wherein in an exemplary embodiment, the apparatus includes:

a request response unit 11, configured to respond to a node query request, obtain a task type specified in the node query request, query and determine an access entry of a service node corresponding to the task type from a node database, where each service node is used to schedule a specific execution node for a task of a type corresponding to the service node as a feedback to the request;

the instruction response unit 12 is configured to respond to a role switching instruction of any execution node during a period that a service node corresponding to the task type is not queried in the node database, store an access entry of the execution node and the task type in association as a data record of the node database, so that the execution node is upgraded to the service node;

and the node monitoring unit 13 is configured to monitor a connection state of any service node, and when any service node is disconnected, clear a data record corresponding to the service node from the node database.

Referring to fig. 5, a node switching method according to the present application is implemented by an application program on a node server side, and in an exemplary embodiment, the method includes the following steps:

step S21, initiating a node query request, querying a node database for a service node matching the task type preset by the current node, and obtaining a corresponding feedback result:

the executing node sends the node query request of the corresponding task type to the node database at regular time, so that the node database responds to the node query request, queries the service node corresponding to the task type, and sends the feedback information of the response to the executing node.

When the feedback information contains the access entry of the service node corresponding to the node query request, the execution node continuously uploads the responsible balance information of the execution node to the corresponding service node through the access entry so as to participate in task operation request scheduling activity of the service node and preempt the execution right of the task type corresponding to the corresponding task operation request.

When the feedback result does not contain the access entry of the service node corresponding to the node query request, the execution node sends a role switching instruction to the node database so as to switch the execution node to the service node of the task type corresponding to the node query request and schedule the corresponding execution node for the task operation request of the task type.

The processing or scheduling task type corresponding to each execution node or service node may be generally set or modified at the right time by a technician during deployment, so as to determine that the execution node sends a corresponding node query request to the node database to query the corresponding service node, and the service node receives a corresponding task operation request.

Step S22, when the feedback result indicates that there is no data record of the corresponding service node in the node database, initiating a role switching instruction to store the access entry of the current node and the preset task type in the node database in association for query, so as to upgrade the current node to the service node:

as described above, when the node database responds to the node query request sent by the execution node and does not query the service node corresponding to the node query request, a response feedback result is sent to the execution node, so that the execution node sends the role switching instruction and sends the node configuration information such as the access entry, the scheduling task type, and the like of the role switching instruction to the node database, and the node database responds to the role switching instruction, stores the node information of the execution node as a data record, and completes the operation of upgrading the execution node to the service node.

The executing node sends the role switching instruction to the node database to complete the operation of switching from the executing node to the service node, so that the automatic switching mechanism of the role of the node server is realized, and therefore, the same node server has two working modes, namely, one node server can be in charge of the work of the executing node and the work of the service node, so that the node server has mobility, can flexibly and automatically switch the working modes, improve the deployment of the scheduling system, and can process the fault of the scheduling system.

When the subsequent node database receives the same node query request scheduling, the self information of the service node, such as the access entry, the task type and the like of the service node upgraded by the execution node, is fed back to the execution node or the agent node to which the node query request belongs, so that the execution node or the agent node executes corresponding operation.

Step S23, after the current node is upgraded to the service node, receiving a task operation request corresponding to the task type, and adding the task operation request to the task pool for the execution node to preempt:

in a typical embodiment, after the executing node finishes upgrading to the service node by sending a role switching instruction to a node database, the proxy node may send a task operation request of a task type corresponding to the service node through an access entry of the service node, so that the service node adds the task operation request to a task pool, and notifies a part of executing nodes reported through the access entry of the task operation request to finish scheduling the task type corresponding to the task operation request.

The task operation request comprises task type information, task operation information and the like so as to represent information of a task to be processed by the proxy node, and after the task operation request is sent to the corresponding service node, the task operation request is added to the task pool by the service node so that the service node can schedule the task operation request to the corresponding execution node.

The task pool refers to a database storing one or more task operation requests received by the service node, the service node obtains and adds an access entry of a proxy node to which the task operation request belongs, calculation resource information required for executing task operation, and the like to the task pool as data thereof by analyzing the task operation requests, generally, the task pool is constructed in a list form so as to orderly schedule the task operation requests, and the service node preferentially processes the task operation request received first.

To more fully describe the details of the specific implementation of performing a node upgrade to the service node, referring to fig. 6, when the node is the service node, the following specific steps may be performed:

step S231, receiving a task operation request meeting a preset task type sent by the proxy node, and adding the task operation request into the task pool:

and the proxy node sends the task operation request to be processed to the corresponding service node through an access entry in the feedback result sent by the node database, so that the service node adds the task operation request to a task pool and schedules the corresponding execution node to process the corresponding task for the task operation request.

The task pool stores all task operation requests received by the service node in a list form, so that the service node processes the task operation request added earliest currently in the task pool preferentially and effectively processes all task operation requests received by the service node.

Step S232, selecting a part of execution nodes for processing the task operation request according to the load balancing information reported by the execution nodes matched with the preset task type:

and the execution node reports the load balancing information to the service node through an access entry of the service node in a feedback result obtained from the node database so that the service node can schedule a task operation request for the service node.

The load balancing information is used for representing the occupation condition of the current computing resource of the execution node to which the load balancing information belongs, the execution node can screen out part of the execution nodes with enough computing resources by accessing an entrance, and the service node can send the task operation requests stored in the task pool to the part of the execution nodes by analyzing the load balancing information.

After receiving the load balancing information reported by the executing node matched with the task type of the service node, the service node stores the load balancing information in a corresponding database so as to analyze all the received load balancing information and select a part of executing nodes suitable for processing the current task operation request.

Step S233, sending a task message notifying processing of the task operation request to the partial execution node:

the service node determines a part of execution nodes suitable for processing the task corresponding to the task operation request to be scheduled at present by analyzing all load balancing information in the execution node database, and sends the task operation request to the part of execution nodes, so that the part of execution nodes seizes the processing right (execution right) of the task corresponding to the task operation request by executing the task operation request.

In step S234, the first responder in the partial execution node performs preemption processing on the task operation request:

and after the service node notifies the task operation request to the screened partial execution nodes, the execution node which responds to the task operation request in the partial execution nodes firstly obtains the processing right of the task corresponding to the task operation request.

And after receiving the task operation request sent by the service node, the partial nodes analyze the task operation request and acquire task operation information so as to determine whether redundant computing resources meet the computing resources required by the task execution corresponding to the task operation request or not, and if so, send a response notice to the service node.

The service node informs the execution node to which the service node belongs of the response received firstly as the execution node for processing the task operation request, and gives the execution node the processing right for processing the task operation request, so that the execution node processes the task to be processed of the proxy node to which the task operation request belongs.

Step S24, when the feedback result includes an access entry of a matched service node, keeping the current node as the execution node for processing the task operation request of the preset task type:

when an executing node sends a node query request to a node database and receives a feedback result sent by the node database, wherein the feedback result contains an access entry, the executing node continuously reports load balancing information to a corresponding service node through the access entry so as to obtain a task operation request of a corresponding task type from the service node and preempt and process the task.

To more fully describe the details of the implementation that the execution node receives the feedback result and includes the access entry of the matched service node, referring to fig. 7, the execution node may execute the following steps:

step S241, continuously uploading load balancing information of the service node matching the task type preset by the service node:

after the execution node receives a feedback result sent by the node database and contains an access entry of a matched service node, the execution node continuously uploads the responsible balance information of the execution node to the service node through the access entry so as to continuously update the load balance information stored on the service node, acquires a task operation request from the service node, preempts the task for processing, and makes the execution node be in a full-load working state as much as possible by using the idle computing resources of the execution node.

Step S242, when receiving the task message which is sent by the service node and informs of processing task operation requests, preempting at least one task operation request from the task pool:

when the responsible balance information of the execution node meets the computing resources required by the task operation request to be processed currently by the service node, the execution node receives the task operation request, and continuously sends the load balance information of the execution node to the service node, so that other task operation requests are continuously received from the service node, meanwhile, the received task operation requests are continuously responded, and the execution right of at least one task is obtained.

In one embodiment, an executing node can process any type of task, and the executing node can obtain a plurality of service nodes to obtain task operation requests, so that when the executing node determines that a currently executable task operation request is responded from the obtained plurality of task operation requests in combination with the current computing resource of the executing node, the executing node notifies the service node sent by the task operation request so as to preempt the task execution right of the task operation request.

Step S243, performing task processing on the successfully preempted task operation request:

when the execution node is the execution node which is sent by the service node and responds to a certain task operation request first, the execution node obtains the execution right of the task corresponding to the task operation request and processes the task.

After the execution node acquires the task execution right of the task operation request, the execution node acquires the task information carried by the task operation request and the proxy node access entry by analyzing the task operation request, executes the corresponding task through the task information, and sends the execution result of the completed task to the corresponding proxy node through the proxy node access entry so that the proxy node can acquire information such as data required to be acquired by the task operation request.

Further, a node switching apparatus of the present application can be constructed by functionalizing the steps in the methods disclosed in the above embodiments, and according to this idea, please refer to fig. 8, wherein in an exemplary embodiment, the apparatus includes:

a request initiating unit 21, configured to initiate a node query request, query a node database for a service node with a task type matching a task type preset by a current node, and obtain a corresponding feedback result;

an instruction initiating unit 22, configured to, when the feedback result indicates an access entry of a service node of a node, keep a current node as the execution node for processing that no data record of a corresponding service node exists in the preset point database, initiate a role switching instruction, so as to store the access entry of the current node and a preset task type of the access entry in the node database in association for query, so as to upgrade the current node to the service node;

a request adding unit 23, configured to receive a task operation request corresponding to the task type after the current node is upgraded to a service node, and add the task operation request to a task pool to perform node preemption processing;

and a feedback determining unit 24, configured to, when the feedback result includes an access entry of a matched service node, maintain the current node as the execution node for processing the task operation request of the preset task type.

Referring to fig. 9, a node coordination method according to the present application is executed by an application program on a service node side, and in an exemplary embodiment, the method includes the following steps:

step S31, adapting to multiple task types, where multiple service nodes of each task type compete for the scheduling right of the task of the type, and determine the service node that obtains the scheduling right is the service node of the type, and is used to schedule a specific execution node for the task of the type:

when a plurality of execution nodes attempt to switch themselves to service nodes of corresponding task types through the role switching instruction, the node database upgrades the execution node of the role switching instruction which responds firstly to the service node, and stores the data record of the execution node into the database, thereby identifying the corresponding service node of the task type.

And after the execution node is upgraded to the service node, continuously receiving the task operation requests sent by the access entrance, and adding the task operation requests into the task pool so as to orderly process the task operation requests.

The determined service node can continuously receive the load balancing information of each execution node uploaded through the access entrance of the service node except for receiving the task operation request, so that the load balancing information of each execution node is updated in real time, and the situation that the load balancing information is not matched with the self load condition of the corresponding execution node when the subsequent determined part of the execution nodes are executed is prevented.

Step S32, each determined service node selects a part of execution nodes for the task requested to be processed by the proxy node according to the load balancing information reported by the execution nodes associated with the service node in the task type:

the service node selects an execution node which meets the computing resource required by executing the task operation request as a part of execution nodes by analyzing the latest load balancing information reported by a plurality of execution nodes associated with the service node on the task type, and sends the task operation request to the part of execution nodes so that the part of execution nodes can respond to the task operation request.

The service node determines the computing resources required by the task corresponding to the task operation request by analyzing the task operation request, and selects at least one execution node meeting the computing resources from the newly received load balancing information as an execution node capable of processing the task corresponding to the task operation request.

In step S33, the determined service node sends a task message about the task requesting processing to the partial execution node:

and after selecting a part of execution nodes, the service node sends the task operation request to the part of execution nodes so that the part of execution nodes respond to the task operation request and sends a response notice to the service node.

After receiving the task operation request sent by the service node, the execution node analyzes whether the current computing resource of the execution node meets the computing resource of the task corresponding to the task operation request, if so, sends a response notification to the service node, and if not, waits for the computing resource of the execution node to meet the computing resource of the request and then sends the response notification to the service node.

The service node records the response time of all the execution nodes receiving the task operation request so as to determine the execution node which responds first in the subsequent steps.

Step S34, the identified service node dispatches the task requesting processing to the winner of the partial execution nodes competing for the execution right of the task requesting processing, and performs task processing:

the contention refers to that the service node selects the execution node with the earliest response time from the response times of the execution nodes receiving all the task operation requests, the execution node with the earliest response time obtains the opportunity of processing the task operation request, and the execution node processes the task corresponding to the task operation request.

The service node gives the execution right of the execution node with the earliest response time to process the task corresponding to the task operation request through a corresponding instruction, after the execution node obtains the execution right of the task, the execution node obtains the task information carried by the execution node and the proxy node access entrance through analyzing the task operation request, completes the processing of the task through the task information, and sends an execution result to the proxy node through the proxy node access entrance, so that the processing of the task is completed.

Further, a node coordination apparatus of the present application can be constructed by functionalizing the steps in the methods disclosed in the above embodiments, and according to this idea, please refer to fig. 10, wherein in an exemplary embodiment, the apparatus includes:

a node determining unit 31, configured to adapt to multiple task types, where multiple service nodes of each task type compete for a scheduling weight of a task serving the task type, and correspondingly determine a service node for scheduling a specific execution node for the task of the corresponding type;

the node screening unit 32 is configured to select, by each determined service node, a part of execution nodes for the task requested to be processed by the agent node according to the load balancing information reported by the plurality of execution nodes associated with the service node in the task type;

a task initiating unit 33, configured to send a task message about the task requesting processing to the partial execution node by the determined service node;

and a task scheduling unit 34, configured to schedule the task requesting processing to a winner of the partial execution nodes that competes for the execution right of the task requesting processing, by the determined service node, and perform task processing.

Further, to facilitate the implementation of the present application, the present application provides an electronic device, including a central processing unit and a memory, where the central processing unit is configured to invoke and run a computer program stored in the memory to perform the steps of the node scheduling method, the node switching method, and the node coordination method in the foregoing embodiments.

It can be seen that the memory is suitable for adopting a nonvolatile storage medium, the aforementioned method is implemented as a computer program and installed in an electronic device such as a mobile phone or a computer, the related program code and data are stored in the nonvolatile storage medium of the electronic device, and further the program is executed by a central processing unit of the electronic device and is called from the nonvolatile storage medium to a memory for execution, so as to achieve the desired purpose of the present application. Therefore, it is understood that in an embodiment of the present application, a non-volatile storage medium may be further provided, in which a computer program implemented according to various embodiments of the node scheduling method, the node switching method, and the node coordination method is stored, and when the computer program is called by a computer, the computer program executes the steps included in the method.

In summary, the application realizes a multi-center distributed task scheduling system, efficiently and orderly schedules various types of tasks preprocessed by a client, and realizes that a node server has two working modes through a role switching module, so that the mobility of the multi-center distributed task scheduling system is embodied to deal with the problems occurring in the task scheduling process.

As will be appreciated by one skilled in the art, the present application includes apparatus that are directed to performing one or more of the operations, methods described herein. These devices may be specially designed and manufactured for the required purposes, or they may comprise known devices in general-purpose computers. These devices have computer programs stored in their memories that are selectively activated or reconfigured. Such a computer program may be stored in a device (e.g., computer) readable medium, including, but not limited to, any type of disk including floppy disks, hard disks, optical disks, CD-ROMs, and magnetic-optical disks, ROMs (Read-Only memories), RAMs (Random Access memories), EPROMs (Erasable Programmable Read-Only memories), EEPROMs (Electrically Erasable Programmable Read-Only memories), flash memories, magnetic cards, or optical cards, or any type of media suitable for storing electronic instructions, and each coupled to a bus. That is, a readable medium includes any medium that stores or transmits information in a form readable by a device (e.g., a computer).

It will be understood by those within the art that each block of the block diagrams and/or flowchart illustrations, and combinations of blocks in the block diagrams and/or flowchart illustrations, can be implemented by computer program instructions. Those skilled in the art will appreciate that the computer program instructions may be implemented by a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, implement the aspects specified in the block or blocks of the block diagrams and/or flowchart illustrations disclosed herein.

Those of skill in the art will appreciate that the various operations, methods, steps in the processes, acts, or solutions discussed in this application can be interchanged, modified, combined, or eliminated. Further, other steps, measures, or schemes in various operations, methods, or flows that have been discussed in this application can be alternated, altered, rearranged, broken down, combined, or deleted. Further, steps, measures, schemes in the prior art having various operations, methods, procedures disclosed in the present application may also be alternated, modified, rearranged, decomposed, combined, or deleted.

The foregoing is only a partial embodiment of the present application, and it should be noted that, for those skilled in the art, several modifications and decorations can be made without departing from the principle of the present application, and these modifications and decorations should also be regarded as the protection scope of the present application.

Claims

1. A node scheduling method is characterized by comprising the following steps:

2. The method of claim 1, wherein: the node query request is originated from the execution node, and when no service node feeds back, the execution node correspondingly sends out the role switching instruction;

alternatively, the first and second electrodes may be,

the node query request comes from a task agent node, and the fed back service node schedules a task operation request of the task for the agent node.

3. The method of claim 1, wherein the node database includes data records for a plurality of different service nodes, each service node corresponding to a task type, and only one service node corresponding to a task type at a time.

4. A node switching method is characterized by comprising the following steps:

5. The method according to claim 4, characterized in that the current node performs the following steps in the role of serving node:

6. The method according to claim 5, characterized in that the current node performs the following steps in the role of serving node:

7. A node coordination method is characterized by comprising the following steps:

8. A node switching apparatus, comprising:

9. An electronic device comprising a central processing unit and a memory, wherein the central processing unit is configured to invoke the execution of a computer program stored in the memory to perform the steps of the node scheduling method according to any one of claims 1 to 3; or performing the steps of the node switching method according to any of claims 4 to 6; or to perform the steps of the node coordination method according to claim 7.

10. A non-volatile storage medium, characterized in that it stores, in the form of computer-readable instructions, a computer program implemented according to the node scheduling method of any one of claims 1 to 3; or a computer program implemented according to the node switching method of any of claims 4 to 6; or a computer program implemented according to the node coordination method of claim 7, which, when invoked by a computer, performs the steps comprised by the respective method.