WO2023077952A1 - 数据处理方法、系统、相关设备、存储介质及产品 - Google Patents

数据处理方法、系统、相关设备、存储介质及产品 Download PDF

Info

Publication number
WO2023077952A1
WO2023077952A1 PCT/CN2022/117512 CN2022117512W WO2023077952A1 WO 2023077952 A1 WO2023077952 A1 WO 2023077952A1 CN 2022117512 W CN2022117512 W CN 2022117512W WO 2023077952 A1 WO2023077952 A1 WO 2023077952A1
Authority
WO
WIPO (PCT)
Prior art keywords
data
edge
type
target
central cluster
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/CN2022/117512
Other languages
English (en)
French (fr)
Inventor
朱海华
李丹
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Tencent Technology Shenzhen Co Ltd
Original Assignee
Tencent Technology Shenzhen Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Tencent Technology Shenzhen Co Ltd filed Critical Tencent Technology Shenzhen Co Ltd
Priority to EP22888983.8A priority Critical patent/EP4379545B1/en
Publication of WO2023077952A1 publication Critical patent/WO2023077952A1/zh
Priority to US18/200,437 priority patent/US12452339B2/en
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/14—Session management
    • H04L67/141—Setup of application sessions
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/46—Multiprogramming arrangements
    • G06F9/48—Program initiating; Program switching, e.g. by interrupt
    • G06F9/4806—Task transfer initiation or dispatching
    • G06F9/4843—Task transfer initiation or dispatching by program, e.g. task dispatcher, supervisor, operating system
    • G06F9/4881—Scheduling strategies for dispatcher, e.g. round robin, multi-level priority queues
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/46—Multiprogramming arrangements
    • G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
    • G06F9/5005—Allocation of resources, e.g. of the central processing unit [CPU] to service a request
    • G06F9/5027—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals
    • G06F9/505—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals considering the load
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/46—Multiprogramming arrangements
    • G06F9/54—Interprogram communication
    • G06F9/546—Message passing systems or structures, e.g. queues
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/46—Multiprogramming arrangements
    • G06F9/54—Interprogram communication
    • G06F9/547—Remote procedure calls [RPC]; Web services
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/01—Protocols
    • H04L67/10—Protocols in which an application is distributed across nodes in the network
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/2866—Architectures; Arrangements
    • H04L67/289—Intermediate processing functionally located close to the data consumer application, e.g. in same machine, in same home or in same sub-network
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/50—Network services
    • H04L67/56—Provisioning of proxy services
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/50—Network services
    • H04L67/56—Provisioning of proxy services
    • H04L67/564—Enhancement of application control based on intercepted application data
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L69/00—Network arrangements, protocols or services independent of the application payload and not provided for in the other groups of this subclass
    • H04L69/14—Multichannel or multilink protocols
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2209/00—Indexing scheme relating to G06F9/00
    • G06F2209/54—Indexing scheme relating to G06F9/54
    • G06F2209/547—Messaging middleware
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2209/00—Indexing scheme relating to G06F9/00
    • G06F2209/54—Indexing scheme relating to G06F9/54
    • G06F2209/548—Queue

Definitions

  • the present application relates to the field of computer technology, and in particular to a data processing method, system, related equipment, storage medium and product.
  • cloud applications have shown a trend of vigorous development.
  • cloud applications are usually deployed in running containers on edge servers.
  • edge servers For business realization, more public service components will be matched and deployed for edge servers to support various business realizations of edge servers in the cloud application process. Since the locations of each client are relatively scattered, in order to make the operation objects of each cloud application have a better operating experience through the corresponding client, the edge servers that support the operation of the cloud application need to match and deploy more public service components. It can be seen that using the existing cloud application running method will result in waste of component resources.
  • Embodiments of the present application provide a data processing method, system, related equipment, storage medium, and product, which can avoid waste of component resources.
  • An embodiment of the present application provides a data processing method, which is applied to a cloud processing system, where the cloud processing system includes a central cluster and a plurality of edge servers, and the plurality of edge servers share various services provided by the central cluster,
  • the method is executed by a target edge server, where the target edge server includes any edge server with the closest physical distance to the client of the target cloud application among the plurality of edge servers, including:
  • the operating data is the first type of data
  • the service matching the first type of data responds to the first type of data
  • the first communication link is a link that matches the real-time requirements of the first type of data
  • the operation data is sent to the central cluster through the second communication link, so that the central cluster adopts the service matching the second type of data for the
  • the second type of data responds, wherein the second communication link is a link that matches the real-time requirements of the second type of data, and the real-time data of the first type of data and the second type of data Sex requirements are different.
  • the embodiment of the present application proposes a cloud processing system, the cloud processing system includes a central cluster and a plurality of edge servers, the plurality of edge servers share various services provided by the central cluster, and the plurality of edge servers Including a target edge server, where the target edge server includes any edge server with the closest physical distance to the client of the target cloud application among the plurality of edge servers;
  • the target edge server is used to obtain the running data of the target cloud application, and when the running data is the first type of data, send the running data to the central cluster; and, when the operational data is the second type of data, sending the operational data to the central cluster through a second communication link with the central cluster;
  • the first The communication link is a link that matches the real-time requirement of the first type of data
  • the second communication link is a link that matches the real-time requirement of the second type of data
  • the first The real-time requirements of the first type of data and the second type of data are different;
  • the central cluster is used to receive the operation data sent by the target edge server, and when the operation data is the first type of data, perform the processing on the first type of data through the service matching the first type of data Response; and, when the operating data is the second type of data, the central cluster responds to the second type of data through a service matching the second type of data.
  • An embodiment of the present application provides a data processing device, which is applied to a cloud processing system, where the cloud processing system includes a central cluster and a plurality of edge servers, and the plurality of edge servers share various services provided by the central cluster,
  • the data processing device is a computer program (including program code) running in the target edge server, and the target edge server includes the closest physical distance to the client of the target cloud application among the plurality of edge servers
  • the device includes:
  • an acquisition unit configured to acquire the running data of the target cloud application
  • a sending unit configured to send the operating data to the central cluster through the first communication link between the target edge server and the central cluster when the operating data is the first type of data, so that The central cluster responds to the first type of data with a service that matches the first type of data;
  • the first communication link is a link that matches the real-time requirements of the first type of data;
  • the sending unit is further configured to send the operating data to the central cluster through a second communication link when the operating data is the second type of data, so that the central cluster adopts the same
  • the service for matching the data of the second type responds to the data of the second type, wherein the second communication link is a link that matches the real-time requirements of the data of the second type, and the data of the first type and The real-time requirements of the second type of data are different.
  • An embodiment of the present application provides an edge server, including a processor, an input device, an output device, and a memory.
  • the computer program of the above method the computer program includes program instructions, the processor is configured to call the program instructions, and perform the following steps:
  • the operating data is the first type of data
  • the service matching the first type of data responds to the first type of data
  • the first communication link is a link that matches the real-time requirements of the first type of data
  • the operation data is sent to the central cluster through the second communication link, so that the central cluster adopts the service matching the second type of data for the
  • the second type of data responds, wherein the second communication link is a link that matches the real-time requirements of the second type of data, and the real-time data of the first type of data and the second type of data Sex requirements are different.
  • An embodiment of the present application provides a computer-readable storage medium, where program instructions are stored in the computer-readable storage medium, and when the program instructions are executed by the processor, when the program instructions are executed by the processor, the The data processing method as described in the first aspect.
  • any edge server in the cloud processing system can share and reuse public services by invoking the corresponding service components in the central cluster , so as to avoid repeated deployment of public components and improve the resource utilization of the central cluster.
  • the communication links between the edge server and the central cluster in the cloud processing system respectively include the first communication link and the second communication link
  • the target edge server obtains the running data of the target cloud application, if it determines The operating data is the first type of data, then the operating data can be sent to the central cluster through the first communication link, and the central cluster responds to the first type of data by invoking the corresponding service, and if the target edge
  • the running data obtained by the server is the second type of data
  • the target edge server can send the running data to the central cluster through the second communication link with the central cluster, and the central cluster calls the corresponding service for the first Response processing of the second type of data
  • the target edge server will implement it through different communication links when invoking the public services provided in the central cluster, that is to say, different services between the target edge server and the central cluster
  • Interaction and data transmission are realized based on different communication links, which means that when the target edge server calls public services in the central cluster based on different services, the data paths between the data of different services are mutually
  • Figure 1a is a schematic diagram of the operation of a cloud application provided by the embodiment of the present application.
  • Fig. 1b is a schematic diagram of a cloud processing system provided by an embodiment of the present application.
  • Fig. 1c is a schematic diagram of another cloud processing system provided by the embodiment of the present application.
  • FIG. 2 is a schematic flow chart of a data processing method provided in an embodiment of the present application.
  • FIG. 3 is a schematic diagram of related data flows in a cloud application provided by an embodiment of the present application.
  • Fig. 4 is a schematic flow chart of another data processing method provided by the embodiment of the present application.
  • FIG. 5a is a schematic diagram of the architecture of a cloud processing system provided by an embodiment of the present application.
  • Fig. 5b is a schematic diagram of the architecture of another cloud processing system provided by the embodiment of the present application.
  • Fig. 5c is a schematic diagram of an automatic discovery process of an edge point provided by an embodiment of the present application.
  • Fig. 5d is a schematic diagram of the processing flow of a target cloud game provided by the embodiment of the present application.
  • FIG. 6 is a schematic block diagram of a data processing device provided by an embodiment of the present application.
  • Fig. 7 is a schematic block diagram of an edge server provided by an embodiment of the present application.
  • the embodiment of the present application proposes a data processing method, which can be applied in a cloud processing system, and in the cloud processing system, multiple edge servers used to run cloud applications are respectively deployed on different physical In the location area (or physical edge area), the edge servers in different location areas are used to support the display of the application screen of the cloud application in the client within the corresponding area, so that the client of the cloud application can access the physical distance
  • An edge server that is relatively close that is, the physical distance is less than or equal to the distance threshold
  • the rendering speed of the client's application screen can be effectively improved, thereby improving the running smoothness of the cloud application on the client.
  • the cloud application is an application program based on cloud computing. In this operating mode, all cloud applications run on the server side, and the rendered application screen is compressed and sent to the client through the network. Then the corresponding , the client can obtain operation events related to the cloud application through the terminal device, such as touch screen events, keyboard and mouse events, joystick events, etc., and transmit them to the server through the network, so as to achieve the purpose of interacting with the cloud application .
  • cloud instances refer to actual A virtual running container running for the cloud application.
  • the cloud application runs in the running container in the edge server, where the running container may be an Android (Android) container.
  • the cloud instance further includes a transmission container and a proxy container, wherein the connection relationship between the running container, the transmission container, the proxy container, and the client can be shown in Figure 1a, and the cloud instance includes the dotted line in Figure 1a
  • the running container is used to install and run the cloud application, and the application screen of the cloud application can be coded and sent to the transmission container.
  • One side of the transmission container establishes a communication connection with the client 104, and the other side communicates with the client 104
  • the running container (or the coding process of the running container) establishes a communication connection.
  • the transmission container obtains the coded data of the application screen from the running container, it sends the coded data to the client, and the client can use the obtained
  • the coded data is decoded, and the application screen is rendered and displayed in the interface.
  • the transmission container can be an instant communication container (WebRTC, Web Real Time Clock) container, and the communication connection established between the transmission container and the client can be Point-to-point (P2P, Peer to Peer) connection.
  • the agent (Instance) container can also become an agent.
  • the agent container is used to maintain the creation/destruction of the application instance and various state transitions during the life cycle, and establish a connection with the central cluster for the running container and the transmission container. Communication channels, such as accessing the database and message middleware of the central cluster.
  • the cloud processing system further includes a central cluster, which is used to deploy public services of cloud applications, such as running analysis services and basic instruction services.
  • the operation analysis service refers to the service used to analyze the operation status of the cloud application, such as the freeze situation and the delay situation
  • the basic instruction service is opposite to the operation service for the cloud application.
  • the operation service refers to the application
  • the service for updating the screen, the basic command service includes services such as login authentication and data storage provided for the cloud application. It can be understood that the update processing of the application screen can be executed by the application container of the edge server, and other operation instructions except the update request for the application screen need to be executed by calling the public components in the central cluster.
  • any other service except the request to update the application screen can be used as the basic instruction service.
  • the connection relationship between the central cluster and edge servers included in the cloud processing system can be shown in Figure 1b, wherein multiple edge servers in the cloud processing system share various services provided by the central cluster , wherein, the central cluster can be composed of one or more computer devices (such as the computer device 10 shown in Figure 1b), and the edge server 11 can also be a server, or a server composed of multiple servers Cluster, in the embodiment of this application, the edge server is mainly used as an example for illustration, and different edge services are used to support the operation of cloud applications in different geographical areas, such as the server 110 shown in Figure 1b It is different from other servers corresponding to the geographical area that supports the running of cloud applications.
  • edge servers in the cloud processing system will share the public services provided by the central cluster, so that the edge servers can communicate with the central cluster to Calling public services also realizes the multiplexing of public services in the central cluster in edge servers in different geographic locations, that is, based on the deployment of public services in the central cluster and the connection of each edge server to the central
  • the sharing and reuse of public services in the cluster can no longer deploy public services in the edge server, so the edge server can be deployed lightly, that is, only relevant components used to support the operation of cloud applications can be deployed in the edge server, That is, only the outflow components related to the application screen of the cloud application can be deployed on the edge server.
  • the deployment of components in the central cluster and edge servers can be as shown in Figure 1c.
  • the central cluster is connected to different application backgrounds through the application service platform 120.
  • the central cluster includes public components 121 that provide public services and basic components 122, and each edge server multiplexes the upper-level central
  • the common components and basic components in the cluster make it unnecessary for each edge server to repeatedly deploy common components and basic components, and only need to deploy cloud application instances in the edge servers.
  • the public service since the public service is implemented by deploying corresponding public components in the central cluster, for example, the public service may include authentication, caching, scheduling, etc., wherein the public components deployed in the central cluster can be Including authentication components, cache components, scheduling components, etc., since each edge server can share the public services provided by each public component in the multiplexing center cluster, it can also avoid repeated deployment of common components in the edge server, thereby improving the center Utilization rate of each public component in the cluster to avoid waste of component resources.
  • the edge server when the edge server is running the cloud application, the edge server is only used to execute the outflow service of the application screen (that is, to output the coded data of the application screen) service), and the implementation of other public services will call the corresponding public components in the central cluster through the communication connection with the central cluster.
  • the communication connection between the edge server and the central cluster respectively includes a first communication link and a second communication link, wherein, when the operation data is the first type of data, the edge server can use the first communication link to The data is sent to the central cluster, so that the central cluster responds to the operational data through a service that matches the operational data.
  • the edge server can send the running data to the central cluster through the second communication link, and make the central cluster call the service that matches the running data. data in response.
  • the first communication link is a link that can meet the real-time requirements of the first type of data
  • the second communication link is a link that can meet the real-time requirements of the second type of data, and the first type of data and the second type of data Different real-time requirements
  • the central cluster is used to receive the operating data sent by the edge server, and when the operating data is the first type of data, respond to the first type of data through the service that matches the first type of data
  • the central cluster responds to the second type of data through services matching the second type of data when the running data is the second type of data.
  • the real-time requirement of the first type of data may be that the transmission delay is less than the first delay threshold
  • the real-time requirement of the second type of data may be that the transmission delay is less than the second delay threshold
  • the first time delay may be less than the second delay threshold.
  • the latency threshold is different from the second latency threshold.
  • the link delay of the first communication link is less than the first delay threshold, that is, the first communication link can meet the real-time requirements of the first type of data
  • the link delay of the second communication link The path delay is less than the second delay threshold, that is, the second communication link can meet the real-time requirement of the second type of data.
  • the target edge server of the application is described in detail as the execution subject, wherein the target cloud application is run in multiple different edge servers, and one edge server is used to support the operation of the cloud application in a client within a corresponding physical location,
  • the target edge server may be any edge server with the closest physical distance to the client of the target cloud application among the multiple edge servers running the target cloud application, or may not be the edge server with the closest physical distance. edge server.
  • the edge server with the closest physical distance to the client of the target cloud application when the edge server with the closest physical distance to the client of the target cloud application is working normally (or, working normally and the occupied processing resources are less than the threshold), priority is given to The edge server with the closest physical distance to the client of the target cloud application is used as the target edge server, and if the edge server with the closest physical distance to the client of the target cloud application is in an abnormal working state or is occupied
  • the processing resource of is greater than or equal to the threshold
  • other non-physically closest edge servers may be used as target edge servers, for example, the edge server corresponding to the second closest physical distance may be used as the target edge server.
  • the target edge server as an edge server with the closest physical distance to the client of the target cloud application as an example.
  • the method includes step S201 to step S202 , and each step will be described below in conjunction with FIG. 2 .
  • the operating data is the first type of data
  • the first communication link is a link that matches the real-time requirements of the first type of data.
  • the target cloud application is any one of the cloud applications supported by the target edge server, and the application screen of the target cloud application is rendered according to the rendering resources obtained by the corresponding client from the target edge server.
  • the target edge server can obtain the running data related to the running process of the target cloud application, wherein the running data of the target cloud application
  • the data mainly includes audio and video data, basic instruction data (or application basic instructions, which can be authentication instructions or data storage instructions, etc.) data, summary data, etc.).
  • the audio and video data is passed through the data path between the game instance 301 in the target edge server and the client 302
  • the transmission is generally carried out by the transmission container in the target edge server to the client.
  • the client since the client renders and displays the application screen through the audio and video data (or the encoded data of the audio and video data) obtained from the target edge server Therefore, the data channel used to transmit the audio and video data has the highest real-time requirements, and the data packet loss rate is required to be lower than the preset packet loss threshold during data transmission, so as to avoid frame loss and stuttering. affect user experience.
  • the basic instruction data can also be referred to as application basic instructions.
  • the application basic instructions refer to some general instructions sent by the client.
  • the basic instruction data can be forwarded to the public component 304 in the central cluster through the gateway 303 to request public service, or be forwarded to the target edge server to request to update and display the application screen.
  • the public service that is forwarded to the public component and requested may include services such as enabling applications, enabling billing, and transfer of instance control rights.
  • the communication link involved in the basic instruction data includes the communication link between the client and the public components in the central cluster, and the communication link between the client and the target edge server.
  • the data path for basic command data has certain real-time requirements, but the basic command data can trigger the client to resend after it is lost.
  • the data path of basic command data has real-time requirements compared with audio and video data.
  • Relatively low that is, the link delay of the data link used to transmit basic quality data may be higher than the link delay of the data link used to transmit audio and video data.
  • the real-time requirement of the data supporting the analysis of the operation status of the target cloud application is The real-time requirements of the above-mentioned data are the lowest.
  • the target edge server since the operation data obtained by the target edge server has different real-time requirements for public service requests, the target edge server may request public services based on the difference in real-time requirements of the operation data.
  • the operating data of the system is divided into the first type of data and the second type of data, wherein, if the first delay threshold corresponding to the first type of data is greater than the second delay threshold corresponding to the second type of data, then the first type of data has a significant impact on real-time The requirements for real-time data are lower than those for the second type of data. If the first delay threshold corresponding to the first type of data is smaller than the second delay threshold corresponding to the second type of data, then the first type of data has higher real-time requirements than the second type of data.
  • the real-time requirements of the first type of data are lower than the real-time requirements of the second type of data, then it can be understood that if the operating data includes basic instruction data, then The basic instruction data is the second type of data, and if the operation data includes data that supports the analysis of the operation of the target cloud application, then the data used to analyze the operation of the target cloud application is the first class data.
  • the data supporting the analysis of the running status of the target cloud application may be summary data of frame data generated by the application instance of the target edge server during the process of sending audio and video data to the client, then, the The summary data can be used to analyze the delay and freeze in the process of sending the audio and video data of the cloud application to the client for rendering and displaying the application screen, and based on the delay and freeze analysis, the cloud application can be finally quantified. operating conditions.
  • the running status of the cloud application may include a quantified degree of program running stalls.
  • the active interval can be divided into multiple sub-intervals according to the preset duration (set to 5 seconds), assuming that every 8 sub-intervals is a sliding window (set to 40 seconds). If the number of times the message response time in each window exceeds the response time threshold (set to 200ms) is equal to or greater than 6 times, the current window is considered to be a stuck window, and a sub-interval is moved to the right to enter the next window until it is judged Get all the windows in the active interval to get the set of stuck windows. Then calculate the ratio of the freeze window to the active interval of the cloud application.
  • the degree of program running lag is characterized by the ratio of the delay of a single operation to the active area of the freeze window stack cloud application.
  • the process based on the analysis of the target cloud application is performed by the operation analysis service provided by the analysis component in the public component, therefore, the target edge server collects and analyzes the data of the operation of the target cloud application (as described above After the summary data), the collected data can be sent to the analysis component for analysis. And because the process of analyzing the running status of the cloud application can be carried out in real time, it can also be executed when there is an analysis requirement. Therefore, after the target edge server collects the data for analyzing the running status of the target cloud application, no The data needs to be responded immediately, but the packet loss rate is required to be less than a certain packet loss threshold. If the packet loss rate of the data is higher than the packet loss threshold, it will affect the analysis result of the operation status of the target cloud application.
  • the data for analyzing the running status of the target cloud application includes not only the above summary data for recording the process of sending audio and video data from the application instance of the target edge server, but also includes The log data generated during the process (such as running in the background), so that the running process of the target cloud application can be comprehensively evaluated, such as the resource occupation of the target cloud application during the running process.
  • the data for analyzing the operation of the target cloud application is the summary data recorded for the transmission process of audio and video data.
  • the target edge server can use the above-mentioned cloud processing system based on the real-time requirements of the above-mentioned relevant data and the data flow direction of the above-mentioned three types of data Different communication links transmit the above three kinds of data, so that the mutual influence between different communication links can be avoided, the reliability of processing related data of cloud applications can be improved, and the stability of the cloud processing system can be improved .
  • the above-mentioned cloud processing system also designs three corresponding different communication links, wherein, in the cloud processing system, the communication link between the central cluster and the edge server (such as the target edge server) includes The first communication link and the second communication link, the communication links included between the edge server and the client are used to transmit the above audio and video data. Since the central cluster in the cloud processing system and the edge server include the first communication link and the second communication link for data interaction, and the basic instruction data related to the target cloud application is used to analyze the operation of the target cloud application After the data is obtained by the target edge server, it needs to be sent to the central cluster, and the relevant public services of the central cluster are called to respond to the basic instruction data, and the running status of the target cloud game is analyzed and processed.
  • the communication link between the central cluster and the edge server such as the target edge server
  • the edge server includes The first communication link and the second communication link
  • the communication links included between the edge server and the client are used to transmit the above audio and video data. Since the central cluster in the cloud processing system and the edge
  • the target edge server can use different communication links to transfer the basic instruction data and the supporting The data for analyzing the operation status of the target cloud application is sent to the central cluster.
  • the target edge server may, after obtaining the operation data, if the obtained operation data is data for analyzing the operation status of the target cloud application, then It can be determined that the obtained operating data is the first type of data, and the operating data can be sent through the first communication link, so that after the central cluster obtains the first type of data, it can use the service pair matching the first type of data
  • the first type of data responds, and if the operating data acquired by the target edge server is basic instruction data, the target edge server can determine that the acquired operating data is the second type of data, and can pass the second communication link Sending the second type of data to realize the real-time requirements based on the operating data uses the first communication link or the second communication link to send data, which can effectively improve the security of the cloud processing system and ensure the data sending process Data Security in .
  • the target edge server can use the first communication link to use the data for the target cloud application
  • the data for analyzing the running status of the target cloud application is sent to the central cluster.
  • the data for analyzing the running status of the target cloud application is sent to the central cluster by the target edge server through the first communication link Analysis component, so as to realize the call to the operation analysis service provided by the analysis component in the central cluster, and if the operation data obtained by the target edge server is the basic instruction data for the target cloud application, it can pass through the second communication link If the data is sent by the channel, then step S203 can be executed.
  • the target edge server can Immediately send the first type of data to the central cluster through the first communication link, so that a real-time analysis of the operation of the target cloud application can be realized, or, after the target edge server obtains the first type of data, First cache the first type of data obtained in the storage space, and when it is determined that there is a need to analyze the running status of the target cloud application, then send the first type of data to the central cluster through the first communication link To analyze the running status of the target cloud application.
  • the target edge server After the target edge server obtains the first type of data, the target edge server can first backup and store the obtained first type of data in the local disk, and then pass the first type of data through the public network through the first communication link sent to the central cluster, so that the target edge server can initiate retransmission based on the first type of data backed up in the local disk when it is determined that the first type of data is lost during transmission, so as to ensure that the central cluster can receive correct And the complete first type of data can improve the accuracy of analyzing the target cloud application.
  • the target edge server after the target edge server acquires the first type of data, it can cache the first type of data in a local storage space, or send the first type of data to the blockchain network for further processing. Cache, and obtain the first type of data from the blockchain network when the first type of data needs to be sent to the central cluster. Based on the cache of the first type of data in the blockchain network, it can effectively prevent the first type of data from being tampered with.
  • the running data of the target cloud application obtained by the target edge server is obtained based on the physical distance range, because any edge server in the cloud processing system is used to support a cloud application corresponding to a customer within a physical range end, and the physical distance between the client of the cloud application supported by any edge server and any edge server must be smaller than the physical distance between the client of the cloud application and other edge servers, then,
  • the running data of the target cloud application obtained by the target edge server is obtained from a client of the target cloud application with the closest physical distance to the target edge server.
  • the target edge server can determine the physical distance between the target edge server and the corresponding client through the Internet Protocol Address (Internet Protocol Address, IP) of the target edge server and the IP address of the terminal device corresponding to the corresponding client.
  • IP Internet Protocol Address
  • the physical location of the target edge server (which may be latitude and longitude information) can be determined based on the IP address of the target edge server, and the physical location of the terminal device can be determined based on the IP address of the terminal device, and then based on the physical location of the target edge server and the physical location of the terminal device to determine the physical distance between the target edge server and the terminal device.
  • the second communication link is a link that can meet the real-time requirements of the second type of data, and the real-time requirements of the first type of data and the second type of data are different.
  • the target edge server When the target cloud application is running in the target edge server, if the running data obtained by the target edge server is basic instruction data, the target edge server can determine that the obtained running data is the second type of data, and then can use the second communication
  • the link sends the basic instruction data to the central cluster for corresponding processing.
  • the basic instruction data is related to the display of the application interface of the target cloud application, and the application screen of the target cloud application is sent to the client for display, then, based on the display of the application screen of the target cloud application by the client,
  • the operation object of the target cloud application can send relevant instruction data for the target edge server to the target edge server through the application screen displayed on the client, wherein the relevant instruction data sent by the application object to the target edge server through the client includes Basic instruction data and operation instruction data.
  • the application screen displayed on the client includes a basic screen and an operation screen
  • the operation screen refers to an interface that enters the actual application environment (such as the main application interface of the target cloud application)
  • the The basic screen refers to the interface displayed to the application object before entering the actual application environment, for example, it may be a login interface, an application resource update interface, and the like.
  • the instruction data generated based on the operations performed by the operating user on the basic screen is the basic instruction data.
  • the instruction data generated based on the operations performed by the operating user on the operation screen becomes the operation instruction data.
  • the operation instruction data is an instruction for controlling the display of the application screen of the target cloud application, such as an instruction for controlling the virtual object in the target cloud game
  • the basic instruction data includes Other common commands other than control can be called basic command data.
  • the operation object refers to an object that triggers the running of the target cloud application and controls the target cloud application.
  • the operation screen refers to the interface that enters the actual game environment. Then, in the operation screen corresponding to the actual game environment of the target cloud game, the corresponding operation object can be used for The avatar displayed on the operation screen performs manipulation.
  • the target edge server Since the target edge server will update the application screen based on the operation instruction data after obtaining the operation instruction data for the target cloud application from the client, therefore, the target edge server can trigger the operation after obtaining the operation instruction data
  • the running container of the target cloud application updates the application screen currently displayed on the client, encodes the updated application screen and sends it to the transmission container, and finally delivers it to the client, so that the The application screen is updated.
  • the target edge server obtains the basic instruction data for the target cloud application from the client, since the target edge server only supports the screen update service, most of the basic instruction data need to request other than the screen update service.
  • the target edge server needs to forward the basic instruction data to the central cluster, so as to request to call the relevant public services in the central cluster to process the basic instruction data. response.
  • the target edge server forwards the basic instruction data to the central cluster, based on the two different communication links existing in the cloud processing system, the target edge server can forward the basic instruction data to the center through the second communication link cluster, so that the central cluster responds to the basic command data through the basic command service.
  • any public service deployed in the central cluster can be realized by one or more service components, and one service component can correspond to one or more servers.
  • any edge server in the cloud processing system can share and reuse public services by invoking the corresponding service components in the central cluster , so as to avoid repeated deployment of public components and improve the resource utilization of the central cluster.
  • the communication links between the edge server and the central cluster in the cloud processing system respectively include the first communication link and the second communication link
  • the target edge server obtains the running data of the target cloud application, if it determines The operating data is the first type of data, then the operating data can be sent to the central cluster through the first communication link, and the central cluster responds to the first type of data by invoking the corresponding service, and if the target edge
  • the running data obtained by the server is the second type of data
  • the target edge server can send the running data to the central cluster through the second communication link with the central cluster, and the central cluster calls the corresponding service for the first Response processing of the second type of data
  • the target edge server will implement it through different communication links when invoking the public services provided in the central cluster, that is to say, different services between the target edge server and the central cluster
  • Interaction and data transmission are realized based on different communication links, which means that when the target edge server calls public services in the central cluster based on different services, the data paths between the data of different services are mutually
  • Figure 4 is a schematic flow chart of another data processing method proposed in the embodiment of the present application.
  • the data processing method proposed in the embodiment of the present application is also applied in the above-mentioned cloud processing system, and the cloud A plurality of edge servers in the processing system share a plurality of public services provided by the central cluster, wherein the data processing method can be executed by a target edge server for running a target cloud application, the target edge server is in the cloud processing system, and Any edge server with the closest physical distance between clients of the target cloud application.
  • the method may include:
  • the running data is acquired by the target edge server, where the running data may be data used to analyze the running status of the target cloud application, and in one embodiment, to analyze the running status of the target cloud application
  • the analyzed data includes the summary data obtained when the target edge server sends the application screen of the target cloud application to the corresponding client through the application instance.
  • the target edge server sends the application screen to the client, the If the target edge server runs the transmission container included in the application instance of the target cloud application for delivery, then the proxy container included in the application instance can obtain the target based on the process of the transmission container delivering the application screen to the client. Summary data for cloud applications.
  • the target edge server including the application instance may also be called an outflow node.
  • the outflow node is the node responsible for running the application instance in the cluster related to the cloud application. It actually includes the server that hosts the running container and the transmission container that is responsible for transmitting the encoded video stream to the user.
  • the server may be, for example, an Advanced Reduced Instruction Machine (ARM, Advanced RISC Machines) server (a server based on supporting mobile application processors), and the server carrying the transport container may be, for example, x86 (a server supporting applications on personal computers). processor) or ARM server.
  • ARM Advanced Reduced Instruction Machine
  • the data used to analyze the operation of the target cloud application may also be log data generated when the target cloud application is running in the target edge server, and the target edge server can be obtained during the running of the target cloud application The log data, or the log data may also be obtained after the target cloud application finishes running.
  • the target edge server hosting and running the target cloud application can be implemented based on the cloud application mid-platform deployment architecture shown in Figure 5a.
  • the cloud application mid-platform deployment architecture is also the above-mentioned cloud processing system.
  • the deployment architecture of the central cluster in the deployment architecture shown in Figure 5a, the node 50 is the outflow node, the service 51 in the dotted box is a public service, and the nodes running the public service are isolated as control nodes, and the control is used to be responsible for A node that manages and controls application instances of cloud applications.
  • This type of node is used to run public components of cloud applications such as database, message middleware, cache component, gateway access component, user authentication component, and outflow node scheduling component.
  • the central cluster can provide gateway services through various gateways, wherein the gateway included in the central cluster can be an edge service gateway (Edge gateway, EdgeGW) for accessing the service platform.
  • the business platform refers to the operating platform of the management background.
  • the edge business gateway is connected to the business platform and provides the same entrance to the management background; the edge business gateway can also be called the background gateway, which is mainly used to process The security checksum is forwarded to the correct internal service according to the unreasonable business.
  • other gateways can also be included in the deployment architecture, such as signaling gateway (signal GW) and streaming gateway (Stream GW).
  • the operation object of the instance is accessed.
  • the access to the operation object corresponding to the outflow instance of the edge cluster can be accessed through the outflow instance of the edge server.
  • the operation object when it accesses the deployment architecture, it can first install the software development and installation package (SDK, Software Development Kit) of the corresponding cloud application in the corresponding terminal device, so that the terminal device of the operation object It has the ability to render the application screen of the corresponding cloud application, and then can access it through a specific input and output interface and the user gateway in the architecture, so as to realize the access of the operation object to the deployment architecture, wherein the input and output interface can be, for example, It is the F5 interface (a concurrent access interface) of the traditional load balancer (CLB, Classic Load Balancer).
  • CLB Classic Load Balancer
  • the business layer included in the public service in the dotted box marked by 51 may include: Edge Scheduler, Recorder, Conf Center, Transcode, etc. .
  • the edge scheduling service is mainly used to manage the life cycle of each outflow instance on the outflow node (creation, initialization, deletion, etc.)
  • the record storage service is also a screen recording service, which is used to transcode, record or connect audio and video streams To the live broadcast platform, so that the operation object can directly broadcast live when running the cloud application, without consuming the local processing capacity of the terminal device corresponding to the cloud application.
  • the configuration center is used to provide configuration capabilities, such as the version of the outflow component, and the resource requirements at runtime (such as how much processing resources and memory to occupy), etc.
  • Encoding is used to encode audio and video streams, etc.
  • any outflow node includes a running container, a transmission container and a proxy container respectively.
  • the outflow node may also include a certificate node (License Node), which is used to create a key and send the created key to the operation object.
  • the public service certificate center (License Center) is used to verify the security of the key sent by the certificate node to the operation object.
  • the key is used to provide security authentication capabilities, and the outflow instance will be generated at the beginning of creation. Several keys, these keys will be assigned to the operation object later, when the operation object requests the gateway to establish a connection based on the key, the gateway will pass the key to the authentication service for authentication.
  • the public services in this deployment architecture also include storage services, which are used to provide basic storage functions.
  • the storage services can be implemented by various storages, such as distributed storage (kafka), relational database (MySQL), based on remote Dictionary service storage (Redis), registry-based storage (Registry) and file storage (Filesever).
  • the public service also includes quota management service (Quota), data preprocessing service (Pre Process), monitoring (Monitor) service, data (Statistics) service.
  • the quota management service is mainly used to unify the current resource usage of the computing cluster, such as when each instance is created, the central processing unit (CPU, Central Processing Unit)/memory/graphics processing unit (GPU, Graphics Processing Unit) that needs to be allocated Unit,) the number of resources, therefore, the quota management service is used to uniformly calculate and allocate the required number of resources; the data preprocessing service is because the data volume of data processing is very large, and the traditional streaming processing is difficult to meet the performance requirements.
  • CPU Central Processing Unit
  • GPU Graphics Processing Unit
  • the data preprocessing service can be used to process the data generated by each instance first, and then put the aggregated data into the stream processing process for analysis and report display (that is, subsequent push to the data center and operation and maintenance monitoring platform for analysis and display); the monitoring service is used for real-time monitoring of various health indicators of the cloud application system; the data service is a service for summarizing the status of cloud application instances, and is used to provide basic data for business monitoring to the monitor.
  • public services are centrally deployed in the central cluster 51, and by deploying edge servers in the edge clusters, the sharing and reuse of public services in the central cluster is realized.
  • the edge server can share and reuse two different public services provided by the central cluster through two different communication links.
  • the central cluster includes a central proxy device and an edge management device, and the target edge server is associated with the edge proxy device; then, if the real-time requirement for data transmission of the first type of data is lower than that of the second type of data Real-time transmission requirements, the communication link between the edge proxy device and the central proxy device included in the central cluster is the first communication link; and the communication link between the target edge server and the edge management device is the second communication link, then the target edge server can call public services in the central cluster based on the first communication link and the second communication link respectively based on business requirements.
  • the target edge server may send the obtained first type of data of the target cloud application to the central cluster through the first communication link, and call corresponding services in the central cluster to respond to the first type of data, That is, turn to step S402, and when the target edge server obtains the second type of data, it will send the second type of data to the central cluster through the second communication link, so as to call the corresponding service of the central cluster to The second type of data responds, that is, go to step S403.
  • the running data is the first type of data
  • the first communication link is a link that can meet the real-time requirements of the first type of data.
  • the operating data is the second type of data
  • the second communication link is a link that can meet the real-time requirements of the second type of data, and the real-time requirements of the first type of data and the second type of data are different.
  • the target edge server includes a running container, a transmission container and a proxy container.
  • the target cloud application runs in the running container of the target edge server.
  • the transmission container is used to establish connections with the client and the running container to send the application screen encoded by the running container to the client for display, and the proxy container is used to transmit
  • the container establishes a communication connection with the running container, and establishes a communication connection with the edge management device in the central cluster. Therefore, the second communication link refers to the communication link after the communication connection between the agent container and the edge management device is established.
  • FIG. 5b a schematic diagram of deploying multiple edge clusters that share and reuse the public services in the central cluster
  • the center agent device 520 that this central cluster includes is kafka proxy (a kind of distributed log proxy)
  • the edge management device 521 is Edge manager (a kind of stateless management node)
  • the edge agent device associated with the target edge server 522 is kafka log proxy (a distributed log proxy, the log proxy and the central proxy device can be the same or different)
  • the edge management device may also be called an edge computing management device (EC Manager, Edge Computing Manager).
  • EC Manager is a stateless edge management device and is mainly responsible for opening up the connection between the edge server and the central cluster.
  • the second communication link is used to deploy services for managing the edge servers.
  • the management services for the edge servers are stateless services.
  • the stateless services refer to the multiple management services deployed in the central cluster. There is no difference between each management service. If one of the management services crashes, another management service can be executed instead of the crashed management service, thereby ensuring the stability of the second communication link.
  • the edge management device can proxy the communication traffic between the public components in the central cluster and the outgoing instance of the target edge server.
  • the edge management device communicates with the proxy container of the outflow instance by establishing a two-way Transport Layer Security (TLS, Transport Layer Security) network socket (WebSocket) connection (long connection).
  • TLS Transport Layer Security
  • WebSocket WebSocket
  • the running data obtained by the target edge server is obtained by the proxy container of the target edge server, then the obtained running data is the second type of data;
  • the second type of data is sent to the edge management in the central cluster through the communication connection between the proxy container and the edge management device in the central cluster device, so as to realize sending the second type of data to the central cluster.
  • the first type of data includes summary data for recording the application instance in the process of outputting the application screen to the client
  • the amount of data carried in the first communication link is relatively large, such as in When the current number of daily active users (DAU, Daily Active User) of the target cloud application is 500,000, the order of magnitude of data sent through the first communication link reaches tens of billions, but the real-time requirements of the first type of data are relatively low, that is, In other words, the target edge server can transmit operating data with low real-time requirement but large data volume through the first communication link.
  • DAU Daily Active User
  • the edge proxy device includes kafka log proxy, and the edge proxy device obtains the first type of data (such as the above-mentioned summary data) sent by the outflow instance (ie, the associated target edge server) ), the first type of data can be landed on the local disk first, and then the first type of data can be sent to the central proxy device (kafka proxy) located in the central cluster through the public network. Therefore, if the network of the edge proxy device is jittered When the network is disconnected, the first communication link will be temporarily interrupted.
  • the first type of data such as the above-mentioned summary data
  • the outflow instance ie, the associated target edge server
  • the first type of data stored in the local disk can be used to transfer the first type of data to Resend, and if the edge proxy device is re-issued due to equipment or service upgrades or other problems, the first type of data can also be reloaded and resumed from the local disk to ensure that the first type of data is not lost.
  • the transmission control protocol (TCP, Transmission Control Protocol) is used to establish a communication connection between the edge proxy device and the central proxy device, and a cloud load balancing (Cloud Load Balancer, CLB) device and The firewall is used to display the source IP of the device that generates the log data, so as to determine the data security of the first communication link.
  • load balancing can provide safe and fast traffic distribution services, access traffic can be automatically distributed to multiple cloud servers in the cloud via CLB, expand the service capability of the system and eliminate single point of failure, and load balancing supports billions of High-level connections and tens of millions of concurrency can easily handle large-traffic access and meet business needs.
  • the central cluster can also provide message sending and receiving services through message middleware (MQ, Message Queue).
  • MQ message middleware
  • Message middleware is a message-oriented system.
  • Software can also be called message queue, which refers to the platform-independent data exchange with efficient and reliable message delivery mechanism, and the integration of distributed systems based on data communication. By providing message delivery and message queue model, it can be used in distributed environment communication under the extended process.
  • Psrxy Server (a network proxy deployed in the central cluster) and Psrxy Client (a network proxy deployed in the edge server) serve as the proxy network path connecting the central cluster and edge points of k8s (a resource platform), so that When k8s manages the edge point, it does not need to pay attention to the network situation of the edge point (regardless of the public network or the private network).
  • k8s also known as Kubernetes, is a portable and scalable open source platform for managing containerized workloads and services that facilitates declarative configuration and automation.
  • Kubernetes has a large and rapidly growing ecosystem , Kubernetes services, support, and tools are widely available.
  • a target resource (DeamonSet) is deployed in the edge proxy device, and the target resource is used to associate a preset tag with the edge proxy device.
  • the preset tag may be tencent.cr/ec- role: ec-kafka-proxy, then, when other devices in the physical edge area are added with preset tags, other devices with preset tags are used as edge proxy devices, then based on the target resources deployed in the edge proxy devices , so that the kafka log proxy component in the edge cluster will be automatically deployed to the edge proxy device where the target resource is deployed, and the edge server associated with the edge proxy device will also be automatically ready to accept the scheduling service located in the central cluster, thereby realizing
  • the simplification of the construction and maintenance of edge points (including edge servers and associated edge proxy devices) enables edge points to be automatically discovered and connected to the central cluster.
  • step S501 the edge management device monitors the edge proxy device.
  • the first communication link and the second communication link mentioned above are both established after the target edge point is marked as available, and the target edge point is marked as available by the edge management device , then, when the edge management device marks the target edge point as an available state, the edge management device may first obtain the resource object creation information of the edge proxy device of the target edge point.
  • Step S502 the edge proxy device creates a resource object.
  • the resource object pod created by the edge proxy device is a collection of one or more containers.
  • step S503 the edge management device obtains the object identifier.
  • the edge management device determines that the edge proxy device has created the resource object (pod) according to the resource object creation information, it obtains the device identifier of the edge proxy device and the zone (zone) identifier of the physical edge area where the edge proxy device is located .
  • Step S504 the edge management device sends the device identifier of the created object resource to the container operating platform k8s.
  • Step S505 the edge management device sends the area identifier to k8s.
  • step S506 the edge management device associates the storage device ID with the area ID in the central cluster.
  • step S507 the edge management device marks the availability of the target edge point.
  • the edge management device can mark the target edge point as available after the central cluster associates the storage device ID with the area ID .
  • the network type of the edge proxy device is the host network type (i.e., Host Network), and the device identification (i.e. node IP, Node IP) of the edge proxy device configured as the host network type is the same as that configured as the host network type.
  • the object identifiers of the resource objects created in network-type edge proxy devices are consistent; then, when the edge management device obtains the device identifiers of the edge proxy devices, it can obtain the object identifiers of the resource objects created in the edge proxy devices (that is, pod IP), and the obtained object identifier is used as the device identifier of the edge proxy device, then the region identifier obtained by the edge management device is also the physical region (or physical region) of the terminal device corresponding to the client supported by the corresponding edge server edge area), then, after the edge management device associates and stores the device identifier and the area identifier, the discovery process of the newly added edge point (such as the above-mentioned target edge point) is completed.
  • the edge management device obtains the device identifiers of the edge proxy devices, it can obtain the object identifiers of the resource objects created in the edge proxy devices (that is, pod IP), and the obtained object identifier is used as the device identifier of the edge proxy device, then the region identifier obtained by
  • the target edge point is also deleted by the edge management device. Then, the edge management device can stop running the edge proxy device when the edge management device obtains the deletion trigger instruction for the target edge point, and delete the The device ID and area ID associated in the storage, and after the device ID and area ID are deleted, the target edge point is unavailable.
  • the target cloud application is run in the edge servers included in different edge points, and one edge point corresponds to a physical edge area; the edge server of any edge point uses the area identifier of the corresponding physical edge area to The output of the application screen is displayed to the client in the physical edge area.
  • the various services provided by the central cluster include at least operation analysis services and basic instruction services.
  • the first type of data is the data that supports the analysis of the operation status of the target cloud application
  • the second type of data is for The basic instruction data of the target cloud application. Then when the first type of data is the data that supports the analysis of the running status of the target cloud application, the service that matches the first type of data is the operation analysis service; when the second type of data is basic instruction data, the service that matches the second type of data
  • the matching service is the base instruction service.
  • the target edge server After the target edge server sends the first type of data to the central cluster through the first communication link, and after the central cluster analyzes the operation of the target cloud application by running the analysis service, the target edge server can also use the first communication
  • the link obtains the analysis results for the target cloud application, and outputs the obtained analysis results to the quality monitoring device, or, the quality monitoring device for the target cloud application can also be directly connected to the central cluster, and from The analysis result for the target cloud application is acquired in the central cluster.
  • the second type of data with higher real-time requirements sent by the target edge server to the central cluster through the second communication link is obtained by the target edge server after displaying the application screen of the target cloud application on the client. If the target edge server sends the second type of data to the central cluster through the second communication link, and after the central cluster responds to the second type of data through the basic command service, the target edge server will pass the second communication link The link obtains the feedback data generated after responding to the second type of data from the central cluster.
  • the target edge server can obtain the feedback data from the edge management device of the central cluster through the proxy container in the target edge server.
  • the target edge server After the target edge server obtains the feedback data for the second type of data, it can use the feedback data to update the application screen, and display the updated application screen on the client.
  • the target edge server can The feedback data is sent to the running container in the target edge server to update the application screen with the feedback data in the running container, then, after the updated application screen is encoded in the running container, it is sent to the client through the transmission container in the target edge server terminal, and display the updated application screen on the client.
  • the public services in the central cluster can be centrally managed, such as centrally deployed in the Internet Data Center (IDC, Internet Data Center), the edge points can be deployed in computer rooms that are physically close to the operating objects.
  • IDC Internet Data Center
  • the edge points can be deployed in Huizhou and Dongguan respectively. Deploy to provide a more high-quality and streamlined cloud application access experience for users in Huizhou and Dongguan.
  • the deployment of edge points should be as lightweight as possible.
  • the edge points only include outflow nodes that support the operation of outflow instances. Other services will exist attached to the central cluster to reduce the construction and maintenance costs of edge points. Outgoing nodes will also be deployed in the cluster, so that the operation objects covered by the central cluster can access cloud applications through the central cluster.
  • the target cloud application includes the target cloud game
  • the client of the target cloud application includes the game client corresponding to the target cloud game
  • the first type of data that supports the analysis of the operation of the target cloud application includes the target cloud game
  • the basic instruction data for the target cloud application includes game basic instructions
  • the game basic instructions include any one or more of the following: storage instructions and authentication instructions.
  • the cloud game runs based on the cloud game processing system (that is, the above-mentioned cloud processing system), and the cloud game processing system includes a central cluster and a plurality of edge servers, and the plurality of edge servers share the quality analysis service 531 provided by the central cluster And the game basic service 532, the target cloud game will access the corresponding edge server with the closest distance as the target edge server based on the distance from each edge server, and the target edge server can pass the transmission included in the game instance of the target cloud game
  • the container establishes a connection with the game client, and triggers the running of the target cloud game in the game container.
  • the game container will continuously send audio and video streams to the game client through the transmission container, and display the corresponding game screen in the game client.
  • the summary data generated by sending audio and video streams (or the log data generated by running the cloud game) is collected by the target edge server, the edge proxy device associated with the target edge server and the central proxy device in the central cluster are sent to In the central cluster, the quality analysis of the running process of the target cloud game is performed.
  • the game operation user can send the basic game instructions through the terminal device where the game client is located, and the basic game instructions can be passed between the target edge server and the edge management device in the central cluster.
  • the long connection between send the game basic command to the central cluster, and obtain the feedback data for the game basic command from the edge management device, the feedback data can update the currently displayed game screen in the game container Afterwards, the game container sends the updated game screen to the game client through the transmission container for display.
  • the target cloud game processing system also includes a virtual game controller access gateway (GamePad GW).
  • GamePad GW virtual game controller access gateway
  • the game virtual handle is to solve the problem that when there is no physical handle, the game operation object can use the terminal device as the physical carrier for accessing the handle, and generate a touch-screen version of the game handle on the terminal screen.
  • the central cluster An interface for live broadcast access can also be provided, so that in the process of cloud games, the game can be broadcast directly based on the central cluster and then to the live broadcast platform.
  • the target cloud application when the target cloud application is running in the cloud processing system, the running container of the target edge server closest to the client corresponding to the target cloud application executes the operation, and the target edge server is connected to multiple other edge servers.
  • Servers share multiple services provided by the central cluster, thereby ensuring the lightweight deployment of edge servers, and based on the sharing and multiplexing of services in the central cluster, the utilization of resources in the central cluster can be improved.
  • the target edge server After the target edge server obtains the running data for the target cloud application, when the running data is the first type of data, it sends the first type of data to the central cluster through the first communication link for response, and the running data When it is the second type of data, the second type of data is sent to the central cluster through the second communication link for response, and then the feedback data for the operation data can be obtained through the corresponding communication link, because in the cloud processing system
  • the edge point is deployed, the automatic discovery of the edge point can be realized. Therefore, the deployment progress of the edge point and the discovery efficiency of the edge point can be improved. Based on the separation of the data transmission channel, the data security of the cloud application can be effectively improved. performance, and improve the operation fluency of cloud applications.
  • the embodiment of the present application also proposes a data processing device, which is applied to a cloud processing system.
  • the cloud processing system includes a central cluster and a plurality of edge servers, and the plurality of edge servers share
  • the data processing device may be a computer program (including program code) running in the above-mentioned target edge server, the target edge server includes the plurality of edge servers, and Any edge server with the closest physical distance between clients of the target cloud application.
  • the data processing device may be used to execute the data processing methods described in FIGS. 2 and 4 . Please refer to FIG. 6 , and the data processing device includes: an acquiring unit 601 and a sending unit 602 .
  • the obtaining unit 601 is configured to obtain the operation data of the target cloud application
  • the sending unit 602 is configured to, when the operation data is the first type of data, send the operation data to the central cluster through the first communication link between the target edge server and the central cluster, so as to making the central cluster respond to the first type of data with a service that matches the first type of data;
  • the first communication link is a link that matches the real-time requirements of the first type of data ;
  • the sending unit 602 is further configured to send the operating data to the central cluster through a second communication link when the operating data is the second type of data, so that the central cluster uses the same
  • the second type of data matching service responds to the second type of data, wherein the second communication link is a link that matches the real-time requirements of the second type of data, and the first type of data It is different from the real-time requirements of the second type of data.
  • the central cluster includes a central proxy device and an edge management device, and the target edge server is associated with the edge proxy device; if the real-time requirement of the first type of data is lower than that of the second type of data Real-time requirements, the communication link between the edge proxy device and the central proxy device is the first communication link; the communication link between the target edge server and the edge management device is the first communication link the second communication link.
  • the target edge server includes a running container, a transport container, and a proxy container; the target cloud application runs in the running container of the target edge server, and the transport container is used to communicate with the client respectively Establishing a connection with the running container, so as to encode the application screen of the target cloud application generated by the running container, and send the coded application screen to the client for display; the proxy container is used for The transmission container establishes a communication connection with the running container, and establishes a communication connection with the edge management device in the central cluster, wherein the second communication link refers to establishing a communication connection between the proxy container and the edge management device subsequent communication link.
  • the sending unit 602 is configured to: when the running When the data is the second type of data, the running data is sent to the edge management device in the central cluster through the communication connection between the proxy container and the edge management device in the central cluster.
  • the multiple services provided by the central cluster include at least operation analysis services and basic instruction services
  • the first type of data is data that supports the analysis of the operation status of the target cloud application
  • the second type The second type of data is the basic instruction data for the target cloud application; wherein, when the first type of data supports the analysis of the operation of the target cloud application, it matches the first type of data
  • the service provided is the operation analysis service; when the second type of data is the basic instruction data, the service matching the second type of data is the basic instruction service.
  • the basic instruction data is acquired after the client of the target cloud application displays the application screen of the target cloud application; the device further includes a processing unit 603 .
  • the obtaining unit 601 is further configured to obtain feedback data generated after responding to the second type of data from the central cluster through the second communication link;
  • the processing unit 603 is configured to update the application screen by using the feedback data, and send the updated application screen to the client of the target cloud application, so as to display the updated application on the client of the target cloud application screen.
  • the obtaining unit 601 is further configured to: obtain the feedback data from the edge management device of the central cluster through the proxy container in the target edge server;
  • the processing unit 603 is further configured to: send the feedback data to a running container in the target edge server, so as to use the feedback data in the running container to update the application screen; the updated application screen After being encoded in the running container, it is sent to the client of the target cloud application through the transmission container in the target edge server, and the updated application screen is displayed on the client of the target cloud application.
  • the target cloud application includes a target cloud game
  • the client of the target cloud application includes a game client corresponding to the target cloud game
  • One type of data includes log data of the target cloud game
  • the basic instruction data for the target cloud application includes game basic instructions
  • the game basic instructions include at least one of the following: storage instructions and authentication instructions.
  • the target edge server and the edge proxy device associated with the target edge server constitute a target edge point, and both the first communication link and the second communication link are at the target edge point It is established after being marked as an available state; the target edge point is marked as an available state by the edge management device, and when the edge management device marks the target edge point as an available state, the edge management device Acquiring the resource object creation information of the edge proxy device associated with the target edge point, and after the edge management device determines that the edge proxy device has created a resource object according to the resource object creation information, acquiring the resource object creation information of the edge proxy device The device ID and the area ID of the physical edge area where the edge proxy device is located; after the central cluster associates and stores the device ID and the area ID, marks the target edge point as an available state.
  • the network type of the edge proxy device is a host network type
  • the device identifier configured as the edge proxy device of the host network type is the same as that of the edge proxy device configured as the host network type.
  • the object identifiers of the resource objects created in the proxy device are consistent; when acquiring the device identifier of the edge proxy device, the object identifier of the resource object created in the edge proxy device is obtained, and the obtained object identifier is used as the Device ID of the edge proxy device.
  • target resources are deployed in the edge proxy device, and the target resource is used to associate a preset tag with the edge proxy device; when other devices in the physical edge area are added the preset After the tag is added, other devices to which the preset tag is added are used as the edge proxy device.
  • the target edge point is also deleted by the edge management device
  • the method for deleting the target edge point by the edge management device is that when the edge management device obtains the When the deletion of the point triggers the instruction, stop the operation of the edge proxy device, and delete the device identification and the area identification associated in the central cluster; after the equipment identification and the area identification are deleted, The target edge point is in an unavailable state.
  • the target cloud application is run in edge servers included in different edge points, one edge point corresponds to a physical edge area; the edge server of any edge point will The application screen output of the target cloud application is displayed on the client in the physical edge area.
  • any edge server in the cloud processing system can share and reuse public services by invoking the corresponding service components in the central cluster , so as to avoid repeated deployment of public components and improve the resource utilization of the central cluster.
  • the sending unit 602 may send the operating data to the central cluster through the first communication link when it is determined that the operating data is the first type of data, and respond to the first type of data by invoking the corresponding service in the central cluster, and If the acquiring unit 601 acquires that the running data for the target cloud application is the second type of data, the sending unit 602 may send the second type of data to the central cluster through the second communication link with the central cluster, and the The central cluster invokes the corresponding services to respond, because when invoking the public services provided in the central cluster, it will be realized through different communication links, that is, different business interactions and data transmissions with the central cluster are It is implemented based on different communication links, that is, when the public services in the central cluster are called based on different businesses, the data paths between the data of different businesses are independent of each other and do not affect
  • FIG. 7 is a schematic structural block diagram of an edge server (such as the above-mentioned target edge server) provided by an embodiment of the present application.
  • the target edge server shares various services provided by the central cluster included in the data processing system with other multiple edge servers, and the target edge server is any edge server with the closest physical distance to the client of the target cloud application
  • the edge server in this embodiment as shown in FIG. 7 may include: one or more processors 701 ; one or more input devices 702 , one or more output devices 703 and a memory 704 .
  • the aforementioned processor 701 , input device 702 , output device 703 and memory 704 are connected through a bus 705 .
  • the memory 704 is used to store computer programs, and the computer program includes program instructions, and the processor 701 is used to execute the program instructions stored in the memory 704 .
  • Described memory 704 can comprise volatile memory (Volatile Memory), such as random access memory (RAM, Random-Access Memory); Memory 704 can also comprise nonvolatile memory (non-volatile memory), such as flash memory (Flash Memory), solid-state hard drive (SSD, Solid-State Drive) etc.; Memory 704 can also comprise the combination of above-mentioned type memory.
  • volatile memory volatile Memory
  • RAM random access memory
  • Random-Access Memory Random-Access Memory
  • Memory 704 can also comprise nonvolatile memory (non-volatile memory), such as flash memory (Flash Memory), solid-state hard drive (SSD, Solid-State Drive) etc.
  • Flash Memory flash Memory
  • SSD Solid-State Drive
  • the processor 701 may be a CPU.
  • the processor 701 may further include a hardware chip.
  • the above-mentioned hardware chip can be an application-specific integrated circuit (ASIC, Application-Specific Integrated Circuit), a programmable logic device (PLD, Programmable Logic Device) and the like.
  • the PLD may be a Field Programmable Logic Gate Array (FPGA, Field-Programmable Gate array), a general array logic (GAL, Generic Array Logic) and the like.
  • the processor 701 may also be a combination of the above structures.
  • the memory 704 is used to store a computer program
  • the computer program includes program instructions
  • the processor 701 is used to execute the program instructions stored in the memory 704, so as to implement the above-mentioned corresponding methods in Figure 2 and Figure 4 A step of.
  • the processor 701 is configured to call the program instructions to perform: acquiring the running data of the target cloud application; when the running data is the first type of data, passing the running data through the sending the first communication link between the target edge server and the central cluster to the central cluster, so that the central cluster responds to the first type of data with a service matching the first type of data ;
  • the first communication link is a link that matches the real-time requirements of the first type of data; when the operating data is the second type of data, the operating data is sent through the second communication link to the central cluster, so that the central cluster responds to the data of the second type with a service matching the data of the second type, wherein the second communication link is connected to the data of the second type
  • the real-time performance of the data requires matching links, and the real-time performance requirements of the first type of data and the second type of data are different.
  • the central cluster includes a central proxy device and an edge management device, and the target edge server is associated with the edge proxy device; if the real-time requirement of the first type of data is lower than that of the second type of data Real-time requirements, the communication link between the edge proxy device and the central proxy device is the first communication link; the communication link between the target edge server and the edge management device is the first communication link the second communication link.
  • the target edge server includes a running container, a transport container, and a proxy container; the target cloud application runs in the running container of the target edge server, and the transport container is used to communicate with the client respectively Establishing a connection with the running container, so as to encode the application screen of the target cloud application generated by the running container, and send the coded application screen to the client for display; the proxy container is used for The transmission container establishes a communication connection with the running container, and establishes a communication connection with the edge management device in the central cluster, wherein the second communication link refers to establishing a communication connection between the proxy container and the edge management device subsequent communication link.
  • the processor 701 is configured to call the program instruction, For execution: when the operation data is the second type of data, send the operation data to the edge management device in the center cluster through the communication connection between the agent container and the edge management device in the center cluster Edge management device.
  • the multiple services provided by the central cluster include at least operation analysis services and basic instruction services
  • the first type of data is data that supports the analysis of the operation status of the target cloud application
  • the second type The second type of data is the basic instruction data for the target cloud application; wherein, when the first type of data supports the analysis of the operation of the target cloud application, it matches the first type of data
  • the service provided is the operation analysis service; when the second type of data is the basic instruction data, the service matching the second type of data is the basic instruction service.
  • the basic instruction data is acquired after the client of the target cloud application displays the application screen of the target cloud application; the processor 701 is configured to call the program instruction to use In execution: obtain feedback data generated after responding to the second type of data from the central cluster through the second communication link; use the feedback data to update the application screen, and send the updated application screen to the The client of the target cloud application is connected to display the updated application screen on the client of the target cloud application.
  • the processor 701 is configured to invoke the program instructions, for executing: obtaining the feedback data from the edge management device of the central cluster through the proxy container in the target edge server;
  • the processor 701 is configured to invoke the program instructions to execute: sending the feedback data to the running container in the target edge server, so as to update the application in the running container with the feedback data screen; after the updated application screen is coded in the running container, it is sent to the client of the target cloud application through the transmission container in the target edge server, and the client of the target cloud application displays the The updated application screen is described.
  • the target cloud application includes a target cloud game
  • the client of the target cloud application includes a game client corresponding to the target cloud game
  • the first type of data supporting the analysis of the operation of the target cloud application includes log data of the target cloud game
  • the basic instruction data for the target cloud application includes game basic instructions
  • the game basic instructions include at least the following: One: storage instruction and authentication instruction.
  • the target edge server and the edge proxy device associated with the target edge server constitute a target edge point, and both the first communication link and the second communication link are at the target edge point
  • the target edge point is marked as available; the target edge point is marked as available by the edge management device, and the processor 701 is configured to call the program instructions to execute: the edge management device obtains the The resource object creation information of the edge proxy device associated with the target edge point, and after the edge management device determines that the edge proxy device has created a resource object according to the resource object creation information, obtain the device identification of the edge proxy device and the area ID of the physical edge area where the edge proxy device is located; after the central cluster associates and stores the device ID and the area ID, marks the target edge point as an available state.
  • the network type of the edge proxy device is a host network type
  • the device identifier configured as the edge proxy device of the host network type is the same as that of the edge proxy device configured as the host network type.
  • the object identifiers of the resource objects created in the proxy device are consistent; the processor 701 is configured to call the program instructions to execute: obtain the object identifiers of the resource objects created in the edge proxy device, and convert the obtained object The ID is used as the device ID of the edge proxy device.
  • target resources are deployed in the edge proxy device, and the target resource is used to associate a preset tag with the edge proxy device; when other devices in the physical edge area are added the preset After the tag is added, other devices to which the preset tag is added are used as the edge proxy device.
  • the target edge point is also deleted by the edge management device, and the processor 701 is configured to invoke the program instructions to execute: when the edge management device obtains the When the deletion of the edge point triggers the instruction, stop the operation of the edge proxy device, and delete the device identification and the area identification stored in the central cluster; after the device identification and the area identification are deleted , the target edge point is unavailable.
  • the target cloud application is run in edge servers included in different edge points, one edge point corresponds to a physical edge area; the edge server of any edge point will The application screen output of the target cloud application is displayed on the client in the physical edge area.
  • An embodiment of the present application provides a computer program product or computer program, where the computer program product or computer program includes computer instructions, and the computer instructions are stored in a computer-readable storage medium.
  • the processor of the computer device reads the computer instruction from the computer-readable storage medium, and the processor executes the computer instruction, so that the computer device executes the above-mentioned method embodiment as shown in FIG. 2 or FIG. 4 .
  • the computer-readable storage medium can be a magnetic disk, an optical disk, a read-only memory (Read-Only Memory, ROM) or a random access memory (Random Access Memory, RAM), etc.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Software Systems (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Security & Cryptography (AREA)
  • Information Transfer Between Computers (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

本申请实施例公开了一种数据处理方法、系统、相关设备、存储介质及产品，其中方法包括：获取目标云应用的运行数据；当运行数据为第一类数据时，将运行数据通过目标边缘服务器与中心集群之间的第一通信链路发送给中心集群，以使中心集群采用与第一类数据匹配的服务对第一类数据进行响应；第一通信链路是与第一类数据的实时性要求相匹配的链路；当运行数据为第二类数据时，将运行数据通过第二通信链路发送给中心集群，以使中心集群采用与第二类数据匹配的服务对第二类数据进行响应。

Description

数据处理方法、系统、相关设备、存储介质及产品
相关申请的交叉引用
本申请基于申请号为202111286046.0、申请日为2021年11月02日的中国专利申请提出，并要求以上中国专利申请的优先权，以上中国专利申请的全部内容在此引入本申请作为参考。
技术领域
本申请涉及计算机技术领域，尤其涉及一种数据处理方法、系统、相关设备、存储介质及产品。
背景技术
随着云计算的不断深入发展，云应用呈现了蓬勃发展的趋势，当前为了支持云应用的运行，通常是将云应用部署在边缘服务器的运行容器中运行，此外，为了支持云应用各种复杂业务的实现，还会为边缘服务器匹配部署较多的公共服务组件，以支持边缘服务器在云应用过程中的各种业务实现。而由于各客户端的位置较为分散，因此，为了使各云应用的操作对象通过相应的客户端具备较好的操作体验，支持该云应用运行的边缘服务器均需要匹配部署较多的公共服务组件。由此可见，采用现有的云应用运行的方法，将会造成组件资源的浪费。
发明内容
本申请实施例提供了一种数据处理方法、系统、相关设备、存储介质及产品，可避免对组件资源的浪费。
本申请实施例提供了一种数据处理方法，应用于云处理系统，所述云处理系统包含中心集群和多个边缘服务器，所述多个边缘服务器共享由所述中心集群提供的多种服务，所述方法由目标边缘服务器执行，所述目标边缘服务器包括所述多个边缘服务器中，与目标云应用的客户端之间的物理距离最近的任一个边缘服务器，包括：
获取目标云应用的运行数据；
当所述运行数据为第一类数据时，将所述运行数据通过所述目标边缘服务器与所述中心集群之间的第一通信链路发送给所述中心集群，以使所述中心集群采用与所述第一类数据匹配的服务对所述第一类数据进行响应；所述第一通信链路是与所述第一类数据的实时性要求相匹配的链路；
当所述运行数据为第二类数据时，将所述运行数据通过第二通信链路发送给所述中心集群，以使所述中心集群采用与所述第二类数据匹配的服务对所述第二类数据进行响应，其中，所述第二通信链路是与所述第二类数据的实时性要求相匹配的链路，且所述第一类数据和所述第二类数据的实时性要求不同。
本申请实施例提出了一种云处理系统，所述云处理系统包含中心集群和多个边缘服务器，所述多个边缘服务器共享由所述中心集群提供的多种服务，所述多个边缘服务器包括目标边缘服务器，所述目标边缘服务器包括所述多个边缘服务器中，与目标云应 用的客户端之间的物理距离最近的任一个边缘服务器；
所述目标边缘服务器用于获取目标云应用的运行数据，并在所述运行数据为第一类数据时，通过与所述中心集群之间的第一通信链路，将所述运行数据发送给所述中心集群；以及，在所述运行数据为第二类数据时，通过与所述中心集群之间的第二通信链路，将所述运行数据发送给所述中心集群；所述第一通信链路是与所述第一类数据的实时性要求相匹配的链路，所述第二通信链路是与所述第二类数据的实时性要求相匹配的链路，且所述第一类数据和所述第二类数据的实时性要求不同；
所述中心集群用于接收由所述目标边缘服务器发送的运行数据，并在所述运行数据为第一类数据时，通过与所述第一类数据匹配的服务对所述第一类数据进行响应；以及，所述中心集群还在所述运行数据为第二类数据时，通过与所述第二类数据匹配的服务对所述第二类数据进行响应。
本申请实施例提供了一种数据处理装置，应用于云处理系统，所述云处理系统包含中心集群和多个边缘服务器，所述多个边缘服务器共享由所述中心集群提供的多种服务，所述数据处理装置是运行于目标边缘服务器中的一个计算机程序(包括程序代码)，所述目标边缘服务器包括所述多个边缘服务器中，与目标云应用的客户端之间的物理距离最近的任一个边缘服务器，所述装置包括：
获取单元，配置为获取目标云应用的运行数据；
发送单元，配置为当所述运行数据为第一类数据时，将所述运行数据通过所述目标边缘服务器与所述中心集群之间的第一通信链路发送给所述中心集群，以使所述中心集群采用与所述第一类数据匹配的服务对所述第一类数据进行响应；所述第一通信链路是与所述第一类数据的实时性要求相匹配的链路；
所述发送单元，还配置为当所述运行数据为第二类数据时，将所述运行数据通过第二通信链路发送给所述中心集群，以使所述中心集群采用与所述第二类数据匹配的服务对所述第二类数据进行响应，其中，所述第二通信链路是与所述第二类数据的实时性要求相匹配的链路，且所述第一类数据和所述第二类数据的实时性要求不同。
本申请实施例提供了一种边缘服务器，包括处理器、输入设备、输出设备和存储器，所述处理器、输入设备、输出设备和存储器相互连接，其中，所述存储器用于存储支持边缘服务器执行上述方法的计算机程序，所述计算机程序包括程序指令，所述处理器被配置用于调用所述程序指令，执行如下步骤：
获取目标云应用的运行数据；
当所述运行数据为第一类数据时，将所述运行数据通过所述目标边缘服务器与所述中心集群之间的第一通信链路发送给所述中心集群，以使所述中心集群采用与所述第一类数据匹配的服务对所述第一类数据进行响应；所述第一通信链路是与所述第一类数据的实时性要求相匹配的链路；
当所述运行数据为第二类数据时，将所述运行数据通过第二通信链路发送给所述中心集群，以使所述中心集群采用与所述第二类数据匹配的服务对所述第二类数据进行响应，其中，所述第二通信链路是与所述第二类数据的实时性要求相匹配的链路，且所述第一类数据和所述第二类数据的实时性要求不同。
本申请实施例提供了一种计算机可读存储介质，所述计算机可读存储介质中存储有程序指令，所述程序指令被处理器执行时，所述程序指令被处理器执行时，用于执行如第一方面所述的数据处理方法。
在本申请实施例中，基于云处理系统对提供公共服务的中心集群的集中管理，该云处理系统中的任一边缘服务器可通过对中心集群中相应服务组件的调用进行公共服务的共享复用，从而可避免对公共组件的重复部署，提升中心集群的资源利用率。由于 该云处理系统中的边缘服务器与中心集群之间的通信链路分别包括第一通信链路和第二通信链路，那么，如果目标边缘服务器获取到目标云应用的运行数据后，若确定该运行数据为第一类数据，则可通过该第一通信链路将该运行数据发送给中心集群，并在中心集群中通过调用相应的服务对该第一类数据进行响应，而如果目标边缘服务器获取到的运行数据为第二类数据，则该目标边缘服务器可通过与中心集群之间的第二通信链路将该运行数据发送给中心集群，并由中心集群调用相应的服务对该第二类数据进行响应处理，由于目标边缘服务器在对中心集群中提供的公共服务进行调用时，将通过不同的通信链路来实现，也就是说，该目标边缘服务器与中心集群之间的不同业务交互和数据传输是基于不同的通信链路来实现的，也就使得该目标边缘服务器在基于不同业务对中心集群中的公共服务进行调用时，不同业务的数据之间的数据通路是相互独立且互不影响的，从而可实现对不同业务的并发执行，并可提升各业务数据在目标边缘服务器和中心集群之间传输的安全性。
附图说明
为了更清楚地说明本申请实施例技术方案，下面将对实施例描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图是本申请的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。
图1a是本申请实施例提供的一种云应用的运行示意图；
图1b是本申请实施例提供的一种云处理系统的示意图；
图1c是本申请实施例提供的另一种云处理系统的示意图；
图2是本申请实施例提供的一种数据处理方法的示意流程图；
图3是本申请实施例提供的一种云应用中相关的数据流的示意图；
图4是本申请实施例提供的另一种数据处理方法的示意流程图；
图5a是本申请实施例提供的一种云处理系统的架构的示意图；
图5b是本申请实施例提供的另一种云处理系统的架构的示意图；
图5c是本申请实施例提供的一种边缘点的自动发现流程的示意图；
图5d是本申请实施例提供的一种目标云游戏的处理流程的示意图；
图6是本申请实施例提供的一种数据处理装置的示意性框图；
图7是本申请实施例提供的一种边缘服务器的示意性框图。
具体实施方式
下本申请实施例提出了一种数据处理方法，该数据处理方法可应用在云处理系统中，而在该云处理系统中，用于运行云应用的多个边缘服务器被分别部署到不同的物理位置区域(或物理边缘区域)中，而处于不同位置区域的边缘服务器用于支持相应区域范围内的客户端中云应用的应用画面的显示，从而也就使得云应用的客户端可访问物理距离相对较近(即物理距离小于等于距离阈值)的边缘服务器，并从访问的边缘服务器中获取到针对该云应用画面的渲染资源，从而可进一步基于该渲染资源对该云应用的应用画面进行渲染显示。通过对物理距离相对较近的边缘服务器的访问并进行渲染资源的获取，可有效提升客户端的应用画面的渲染速度，从而提升云应用在客户端中的运行流畅度。其中，云应用是以云计算为基础的应用程序，在该运行模式下，所有的云应用都在服务器端运行，并将渲染完毕后的应用画面压缩后通过网络传送给客户端，那么对应 地，客户端则可通过终端设备获取与该云应用相关的操作事件，如触屏事件、键盘鼠标事件、摇杆事件等，并通过网络传输到服务器端，从而达到与该云应用进行交互的目的。可以理解，用于支持云应用运行的服务器端即是上述的边缘服务器，该边缘服务器通过创建云端实例以支持云应用的运行，在一个实施例中，云端实例是指在云应用过程中，实际为该云应用运行的一个虚拟的运行容器，云应用在边缘服务器中即是运行在该运行容器中的，其中，该运行容器可以是安卓(Android)容器。
在一个实施例中，云端实例还包括传输容器和代理容器，其中，运行容器、传输容器、代理容器及客户端之间的连接关系可如图1a所示，该云端实例包括由图1a的虚线框标识的运行容器101、传输容器102及代理容器103。其中，该运行容器用于安装云应用，并运行云应用，且可将该云应用的应用画面编码后发送给传输容器，传输容器的一侧与客户端104建立通信连接，另一侧则和运行容器(或运行容器的编码进程)建立通信连接，因此，该传输容器可在从运行容器获取到应用画面的编码数据后，将该编码数据发送给客户端，客户端从而可对获取到的编码数据进行解码，并在界面中对应用画面进行渲染显示，其中，该传输容器可以是即时通信容器(WebRTC，Web Real Time Clock)容器，该传输容器和客户端之间建立的通信连接可以是点对点(P2P，Peer to Peer)连接。另外，该代理(Instance)容器也可成为代理器，该代理容器用于维护应用实例的创建/销毁以及生命周期过程中的各种状态变迁，并为运行容器和传输容器建立与中心集群内部的通信通道，比如访问中心集群的数据库和消息中间件等。
在一个实施例中，该云处理系统还包括中心集群，该中心集群用于部署云应用的公共服务，如运行分析服务和基础指令服务等。其中，运行分析服务指用于对云应用的运行情况，如卡顿情况和延迟情况进行分析的服务，而基础指令服务是与针对该云应用的操作服务相对的，该操作服务是指对应用画面进行更新的服务，基础指令服务包括为该云应用提供的登陆鉴权以及数据存储等服务。可以理解，对应用画面的更新处理可由边缘服务器的应用容器来执行，而除针对该应用画面的更新请求之外的其他操作指令均需要通过调用中心集群中的公共组件来执行。所以，在本申请实施例中，可将除请求对应用画面进行更新的任一其他服务作为基础指令服务。在一个实施例中，该云处理系统包括的中心集群及边缘服务器之间的连接关系可如图1b所示，其中，该云处理系统中的多个边缘服务器共享由中心集群提供的多种服务，其中，该中心集群可以是由一个或多个计算机设备(如图1b中所示的计算机设备10)构成，该边缘服务器11也可以是一个服务器，同样也可以是由多个服务器构成的服务器集群，在本申请实施例中，主要以该边缘服务器为一个服务器为例进行说明，且不同的边缘服务用于支持不同地理位置区域中的云应用的运行，如图1b中所示的服务器110与其他服务器对应支持云应用运行的地理位置区域不同。
基于该云处理系统的部署，该云处理系统中的多个边缘服务器将共享由中心集群提供的公共服务，从而也就可使边缘服务器可通过与中心集群之间的通信连接，对中心集群的公共服务进行调用，也就实现了将中心集群中的公共服务在不同地理位置区域的边缘服务器中进行复用，也就是说，基于在中心集群中进行公共服务的部署，及各边缘服务器对中心集群中的公共服务的共享复用，可不再在边缘服务器中部署公共服务，因此，该边缘服务器则可轻量部署，即在该边缘服务器中可仅部署用于支持云应用运行的相关组件，即是可仅在该边缘服务器中部署与云应用的应用画面相关的出流组件。在一个实施例中，中心集群和边缘服务器中组件部署可如图1c所示。如图1c所示，该中心集群通过应用业务平台120接入到不同的应用后台，该中心集群中包括提供各公共服务的公共组件121，以及基础组件122，而各边缘服务器通过复用上层中心集群中的公共组件和基础组件，使得各边缘服务器可不再进行公共组件和基础组件的重复部署，仅需 要在边缘服务器中部署云应用实例即可。
在一个实施例中，由于该公共服务是通过在中心集群中部署相应的公共组件实现的，如该公共服务可包括鉴权、缓存、调度等，其中，在中心集群中部署的公共组件则可包括鉴权组件、缓存组件、调度组件等，那么由于各边缘服务器可通过共享复用中心集群中各公共组件所提供的公共服务，也就可避免在边缘服务器中重复部署公共组件，进而提升中心集群中各公共组件的使用率，避免组件资源的浪费。
在一个实施例中，基于该云处理系统中的各边缘服务的部署方式，使得边缘服务器在运行云应用时，该边缘服务器仅用于执行应用画面的出流服务(即输出应用画面的编码数据的服务)，而其他的公共服务的实现将通过与中心集群之间的通信连接对中心集群中相应的公共组件进行调用。而边缘服务器与中心集群之间的通信连接分别包括第一通信链路和第二通信链路，其中，当运行数据为第一类数据时，该边缘服务器可通过第一通信链路将该运行数据发送给中心集群，以使中心集群通过与该运行数据据匹配的服务对该运行数据进行响应。而在运行数据为第二类数据时，则该边缘服务器可将该运行数据通过第二通信链路发送到中心集群中，并使该中心集群可通过调用与该运行数据匹配的服务对该运行数据进行响应。第一通信链路是能够满足第一类数据的实时性要求的链路，第二通信链路是能够满足第二类数据的实时性要求的链路，且第一类数据和第二类数据的实时性要求不同；该中心集群则用于接收由边缘服务器发送的运行数据，并在运行数据为第一类数据时，通过与第一类数据匹配的服务对第一类数据进行响应；以及，中心集群还在运行数据为第二类数据时，通过与第二类数据匹配的服务对第二类数据进行响应。可以理解，通过两条不同的通信链路分别对该中心集群中的不同服务的调用，可有效避免不同通路之间的相互影响，进而保证边缘服务器在调用中心集群中各公共服务时的业务安全性，也就可提升数据的安全性，以及该云处理系统的稳定性。
在本申请实施例中，第一类数据的实时性要求可以是传输时延小于第一时延阈值，第二类数据的实时性要求可以是传输时延小于第二时延阈值，第一时延阈值与第二时延阈值不同。在实际应用时，第一通信链路的链路时延是小于该第一时延阈值的，也即第一通信链路能够满足第一类数据的实时性要求；第二通信链路的链路时延是小于第二时延阈值的，也即第二通信链路能够满足第二类数据的实时性要求。
基于上述的云处理系统，以云应用为目标云应用为例，对基于该云处理系统对该目标云应用相关数据指令的处理过程进行详细描述，其中，本申请实施例主要以执行该目标云应用的目标边缘服务器为执行主体进行详细说明，其中，该目标云应用被运行在多个不同的边缘服务器中，一个边缘服务器用于支持一个相应物理位置范围内的客户端中云应用的运行，而该目标边缘服务器则可以是运行该目标云应用的多个边缘服务器中，与该目标云应用的客户端之间的物理距离最近的任一个边缘服务器，或者，也可以不是对应物理距离最近的边缘服务器。其中，在多个边缘服务器中，与目标云应用的客户端之间的物理距离最近的边缘服务器处于正常工作的情况下(或者，正常工作且被占用的处理资源小于阈值的情况下)，优先将与该目标云应用的客户端之间的物理距离最近的边缘服务器作为目标边缘服务器，而如果与该目标云应用的客户端之间的物理距离最近的边缘服务器处于异常工作状态，或者被占用的处理资源大于等于阈值的情况下，可将其他非物理距离最近的边缘服务器作为目标边缘服务器，如可将对应物理距离次近的边缘服务器作为目标边缘服务器。在本申请实施例中，主要以该目标边缘服务器是与该目标云应用的客户端之间的物理距离最近的边缘服务器为例进行详细说明的。其中，可参见如图2所示的数据处理方法的示意流程图，如图2所示，该方法包括步骤S201至步骤S202，以下结合图2对各个步骤进行说明。
S201，获取目标云应用的运行数据。
S202，当运行数据为第一类数据时，将运行数据通过目标边缘服务器与中心集群之间的第一通信链路发送给中心集群，以使中心集群采用与第一类数据匹配的服务对第一类数据进行响应；第一通信链路是与第一类数据的实时性要求相匹配的链路。
在步骤S201和步骤S202中，目标云应用是目标边缘服务器支持运行的云应用中的任意一个，而该目标云应用的应用画面是根据相应的客户端从该目标边缘服务器中获取的渲染资源进行渲染得到的，在一个实施例中，当目标云应用在目标边缘服务器中运行时，该目标边缘服务器可获取得到与该目标云应用的运行过程相关的运行数据，其中，该目标云应用的运行数据主要包括音视频数据，针对该目标云应用的基础指令数据(或应用基础指令，可以是鉴权指令或者数据存储指令等)和支持对该目标云应用的运行情况进行分析的数据(如日志数据、摘要数据等)。另外，与该目标云应用相关的数据通路有如图3所示的三种类型，如图3所示，该音视频数据是通过目标边缘服务器中的游戏实例301与客户端302之间的数据通路进行传输的，一般是由目标边缘服务器中的传输容器流向客户端，可以理解，由于客户端是通过从目标边缘服务器获取的音视频数据(或音视频数据的编码数据)进行应用画面的渲染显示的，所以，用于传输该音视频数据的数据通道的实时性要求最高，且要求在数据传输过程中，数据的丢包率低于预设的丢包阈值，以避免丢帧卡顿现象而影响用户体验。
此外，该基础指令数据也可称为应用基础指令，该应用基础指令是指客户端所发送的一些通用指令，该基础指令数据可通过网关303被转发到中心集群中的公共组件304，以请求公共服务，或者被转发到目标边缘服务器中，以请求对应用画面进行更新显示。其中，该被转发到公共组件并进行请求的公共服务可以包括开启应用的服务、开启计费的服务以及实例控制权转移等服务。由此可见，基础指令数据涉及的通信链路包括客户端与中心集群中公共组件之间的通信链路，以及客户端与目标边缘服务器之间的通信链路。而针对基础指令数据的数据通路有一定的实时性要求，但基础指令数据在丢失后可触发客户端进行重发，所以，基础指令数据的数据通路相比于音视频数据而言的实时性要求相对较低，也即用于传输基础质量数据的数据链路的链路时延可以高于用于传输音视频数据的数据链路的链路时延。另外，由于对目标云应用的运行情况进行分析的过程可以是实时进行的过程，也可以是按照分析需求执行的，所以，支持对该目标云应用的运行情况进行分析的数据的实时性要求是上述提及数据的实时性要求最低的。在一个实施例中，由于目标边缘服务器获取到的运行数据进行公共服务请求的实时性要求存在差异，因此，该目标边缘服务器可基于运行数据的实时性要求差异，将获取到的需要请求公共服务的运行数据划分为第一类数据和第二类数据，其中，如果该第一类数据对应第一时延阈值大于第二类数据对应的第二时延阈值，那么第一类数据对实时性的要求低于第二类数据对实时性的要求。如果第一类数据对应的第一时延阈值小于第二类数据对应的第二时延阈值，那么第一类数据对实时性的要求高于第二类数据对实时性的要求。在本申请实施例中，以该第一类数据的实时性要求低于第二类数据的实时性要求的情况进行说明，那么也就可以理解，如果该运行数据包括的是基础指令数据，那么该基础指令数据为第二类数据，而如果该运行数据包括的是支持对该目标云应用的运行情况进行分析的数据，那么用于对该目标云应用的运行情况进行分析的数据为第一类数据。
在一个实施例中，支持对该目标云应用的运行情况进行分析的数据可以是由目标边缘服务器的应用实例在将音视频数据发送到客户端的过程中产生的帧数据的摘要数据，那么，该摘要数据则可用于对将云应用的音视频数据发送到客户端进行应用画面的渲染显示的过程进行延迟和卡顿分析等，并可基于该延迟和卡顿分析，最终量化出该云应用的运行情况。
例如，云应用的运行情况可以包括量化出的程序运行卡顿程度。在确定程序运行 卡顿程度时，可以对摘要数据中的所有的操作事件和用户界面更新时间进行统计，剔除没有操作事件和用户界面更新事件的时间段，得到云应用的活跃区间，然后确定各个事件的响应时间，统计超过响应时间阈值的次数，记为n，并确定云应用程序运行期间出现一次长时间响应延迟情况的平均间隔时间，从而得到云应用单次操作的长时间响应延迟的情况，然后再利用滑动窗口，统计摘要数据中在连续时间内出现多次(例如预设次数)响应延迟的情况，进而确定出卡顿窗口。在实现时，可以将活跃区间按预设时长(设定为5秒)划分成多个子区间，假定每8个子区间为一个滑动窗口(设定为40秒)。若每个窗口中消息响应时间超过响应时间阈值(设定为200ms)的次数大等于6次，则认为当前窗口是一个卡顿窗口，并右移一个子区间，进入下一个窗口，直至判断出活跃区间的所有窗口，得到卡顿窗口集合。再计算出卡顿窗口占云应用活跃区间的比值。通过单次操作的延迟情况和卡顿窗口栈云应用活跃区域的比值来表征程序运行卡顿程度。
其中，基于对目标云应用进行分析的过程是由公共组件中的分析组件提供的运行分析服务来执行的，因此，目标边缘服务器在采集对该目标云应用的运行情况进行分析的数据(如上述的摘要数据)后，可将采集到的数据发送到该分析组件来进行分析。而由于对云应用的运行情况进行分析的过程可以是实时进行的，也可以在存在分析需求时才执行，所以，在目标边缘服务器采集到对目标云应用的运行情况进行分析的数据后，不需要在立即对该数据进行响应，但要求丢包率要小于一定的丢包阈值。如果该数据的丢包率高于该丢包阈值将影响针对该目标云应用的运行情况的分析结果。此外，对该目标云应用的运行情况进行分析的数据除了包括上述的用于对该目标边缘服务器的应用实例发送音视频数据的过程进行记录的摘要数据，还可包括该目标云应用在其他运行过程(如后台运行时)时产生的日志数据，从而可对该目标云应用的运行过程进行全面评估，如对该目标云应用在运行过程中的资源占用情况等进行评估。在本申请实施例中，是以对该目标云应用的运行情况进行分析的数据为对音视频数据传输过程进行记录的摘要数据为例进行说明的。
基于上述的目标云应用在运行过程中产生的相关数据所对应的数据通路，目标边缘服务器可基于上述相关数据的实时性要求，以及上述三种数据的数据流向，采用上述云处理系统中涉及的不同通信链路，对上述三种数据进行传输，从而可以避免不同通信链路之间的相互影响，提升对云应用的相关数据进行处理时的可靠性，并且能够提升该云处理系统的稳定性。
在一个实施例中，上述的云处理系统同样设计了三种对应的不同通信链路，其中，在该云处理系统中，中心集群与边缘服务器(如目标边缘服务器)之间的通信链路包括第一通信链路和第二通信链路，边缘服务器与客户端之间包括的通信链路用于传输上述的音视频数据。由于云处理系统中的中心集群与边缘服务器之间包括第一通信链路和第二通信链路进行数据交互，且目标云应用相关的基础指令数据和用于对目标云应用的运行情况进行分析的数据在被目标边缘服务器获取到后，均需要发送到中心集群，并调用中心集群的相关公共服务对该基础指令数据进行响应，及对该目标云游戏的运行情况进行分析处理。因此，为了避免在发送基础指令数据和用于对目标云应用的运行情况进行分析的数据过程中所产生的业务冲突问题，目标边缘服务器可分别采用不同的通信链路将基础指令数据和支持对该目标云应用的运行情况进行分析的数据发送到中心集群中。
基于目标边缘服务器在获取到运行数据后的实时性要求，该目标边缘服务器可在获取到该运行数据后，如果获取到的运行数据为用于对目标云应用的运行情况进行分析的数据，则可确定获取到的运行数据为第一类数据，并通过第一通信链路发送该运行数据，从而可在中心集群获取到该第一类数据后，采用与该第一类数据匹配的服务对该第一类数据进行响应，而如果该目标边缘服务器获取到的运行数据为基础指令数据，则该 目标边缘服务器可确定获取到的运行数据为第二类数据，并可通过第二通信链路发送该第二类数据，以实现基于运行数据的实时性要求分别采用第一通信链路或第二通信链路进行数据的发送，可有效提升云处理系统的安全性，并且可保证数据发送过程中的数据安全性。也就是说，如果该目标边缘服务器获取到的运行数据是用于对该目标云应用的运行情况进行分析的数据，则该目标边缘服务器可通过第一通信链路将用于对该目标云应用的运行情况进行分析的数据发送给中心集群，在一些实施例中，用于对该目标云应用的运行情况进行分析的数据是由该目标边缘服务器通过该第一通信链路发送给中心集群的分析组件的，从而实现对中心集群中由该分析组件提供的运行分析服务的调用，而如果目标边缘服务器获取到的运行数据是针对目标云应用的基础指令数据后，则可通过第二通信链路进行数据的发送，即可转而执行步骤S203。
在一个实施例中，如果该第一类数据是实时性要求较低的用于对目标云应用的运行情况进行分析的数据，那么，该目标边缘服务器在获取到该第一类数据后，可立即通过第一通信链路将该第一类数据发送到中心集群，从而可实现对目标云应用的运行情况进行实时分析，或者，该目标边缘服务器也可在获取到该第一类数据后，先将获取到的第一类数据缓存到存储空间中，并在确定存在对目标云应用的运行情况进行分析的需求时，再通过第一通信链路将该第一类数据发送到中心集群中以对该目标云应用的运行情况进行分析。而在目标边缘服务器获取到该第一类数据后，该目标边缘服务器可先将获取到该第一类数据备份存储到本地磁盘中，然后再通过公网将第一类数据通过第一通信链路发送到中心集群中，从而使得目标边缘服务器可在确定第一类数据在传输过程中被丢失时，基于在本地磁盘中备份的第一类数据发起重传，以保证中心集群可接收到正确且完整的第一类数据，进而可提升对目标云应用进行分析的准确度。
在一个实施例中，该目标边缘服务器在获取到第一类数据后，可将该第一类数据缓存到本地的存储空间中，也可将该第一类数据发送到区块链网络中进行缓存，并在后续需要将该第一类数据发送到中心集群时，从该区块链网络中获取该第一类数据。基于在区块链网络中对第一类数据的缓存，可有效避免第一类数据被篡改。在一个实施例中，该目标边缘服务器获取的目标云应用的运行数据，是基于物理距离范围进行获取的，由于云处理系统中的任一边缘服务器用于支持一个物理范围内的云应用对应客户端的运行，且该任一边缘服务器所支持运行的云应用的客户端与该任一边缘服务器之间的物理距离，必然小于该云应用的客户端与其他边缘服务器之间的物理距离，那么，目标边缘服务器获取的目标云应用的运行数据则是从与该目标边缘服务器之间的物理距离最近的目标云应用的客户端中获取得到的。其中，目标边缘服务器可通过该目标边缘服务器的互联网协议地址(Internet Protocol Address，IP)与相应客户端对应终端设备的IP地址，确定该目标边缘服务器与相应客户端之间的物理距离。在实现时，可以基于目标边缘服务器的IP地址确定出目标边缘服务器的物理位置(可以是经纬度信息)，并基于终端设备的IP地址确定出终端设备的物理位置，然后基于目标边缘服务器的物理位置和终端设备的物理位置，确定出目标边缘服务器与终端设备之间的物理距离。
S203，当运行数据为第二类数据时，将运行数据通过第二通信链路发送给中心集群，以使中心集群采用与第二类数据匹配的服务对第二类数据进行响应。
其中，第二通信链路是能够满足第二类数据的实时性要求的链路，且第一类数据和第二类数据的实时性要求不同。
目标云应用在目标边缘服务器中运行时，如果该目标边缘服务器获取到的运行数据为基础指令数据，则该目标边缘服务器可确定获取到的运行数据为第二类数据，进而可通过第二通信链路将该基础指令数据发送到中心集群进行相应处理。其中，基础指令数据与目标云应用的应用界面的显示相关，而该目标云应用的应用画面是被下发到客户 端进行显示的，那么，基于客户端对目标云应用的应用画面的显示，该目标云应用的操作对象可通过在客户端中显示的应用画面，向目标边缘服务器发送针对该目标云应用的相关指令数据，其中，应用对象通过客户端向目标边缘服务器发送的相关指令数据包括基础指令数据和操作指令数据。在一个实施例中，在该客户端中显示的应用画面包括基础画面和操作画面，该操作画面是指进入到实际的应用环境中的界面(如该目标云应用的应用主界面)，而该基础画面则是指未进入到实际的应用环境之前，显示给应用对象的界面，例如可以是登陆界面、应用资源更新界面等。可以理解，当在客户端中显示的应用画面为基础画面时，基于操作用户在该基础画面中执行的操作所生成的指令数据为基础指令数据。而在客户端显示的操作画面时，基于操作用户在操作画面中执行的操作生成的指令数据则成为操作指令数据。其中，由于该操作指令数据是用于对目标云应用的应用画面的显示进行控制的指令，如针对目标云游戏中的虚拟对象进行控制的指令，那么，该基础指令数据则包括除对虚拟对象进行控制外的其他通用指令均可称为基础指令数据。其中，该操作对象是指触发运行目标云应用，并对该目标云应用进行操控的对象。在一个实施例中，如果该目标云应用是目标云游戏，该操作画面则是指进入到实际游戏环境的界面，那么，目标云游戏实际的游戏环境对应的操作画面中，对应操作对象可对显示在该操作画面中的虚拟人物进行操控。
由于目标边缘服务器在从客户端获取到针对目标云应用的操作指令数据后，将基于该操作指令数据对应用画面进行更新，因此，该目标边缘服务器可在获取到该操作指令数据后，触发运行该目标云应用的运行容器对当前显示在客户端中的应用画面进行更新，并将更新后的应用画面编码后发送给传输容器，并最终下发到客户端中，从而使客户端中显示的应用画面被更新。在另一种实现方式中，如果目标边缘服务器从客户端获取到针对目标云应用的基础指令数据，由于该目标边缘服务器仅支持针对画面更新服务，而基础指令数据大都需要请求除画面更新服务之外的其他公共服务，因此，该目标边缘服务器在获取到该基础指令数据后，需要将该基础指令数据转发到中心集群中，以请求调用该中心集群中的相关公共服务对该基础指令数据进行响应。在目标边缘服务器将该基础指令数据转发到中心集群时，基于该云处理系统中存在的两条不同的通信链路，该目标边缘服务器可将该基础指令数据通过第二通信链路转发到中心集群，以使中心集群通过基础指令服务对该基础指令数据进行响应。其中，在中心集群中部署的任一公共服务可由一个或多个服务组件实现，一个服务组件可以对应一个或多个服务器。
在本申请实施例中，基于云处理系统对提供公共服务的中心集群的集中管理，该云处理系统中的任一边缘服务器可通过对中心集群中相应服务组件的调用进行公共服务的共享复用，从而可避免对公共组件的重复部署，提升中心集群的资源利用率。由于该云处理系统中的边缘服务器与中心集群之间的通信链路分别包括第一通信链路和第二通信链路，那么，如果目标边缘服务器获取到目标云应用的运行数据后，若确定该运行数据为第一类数据，则可通过该第一通信链路将该运行数据发送给中心集群，并在中心集群中通过调用相应的服务对该第一类数据进行响应，而如果目标边缘服务器获取到的运行数据为第二类数据，则该目标边缘服务器可通过与中心集群之间的第二通信链路将该运行数据发送给中心集群，并由中心集群调用相应的服务对该第二类数据进行响应处理，由于目标边缘服务器在对中心集群中提供的公共服务进行调用时，将通过不同的通信链路来实现，也就是说，该目标边缘服务器与中心集群之间的不同业务交互和数据传输是基于不同的通信链路来实现的，也就使得该目标边缘服务器在基于不同业务对中心集群中的公共服务进行调用时，不同业务的数据之间的数据通路是相互独立且互不影响的，从而可实现对不同业务的并发执行，并可提升各业务数据在目标边缘服务器和中心集群之间传输的安全性。
请参见图4，是本申请实施例提出的另一种数据处理方法的示意流程图，同样的，在本申请实施例中提出的数据处理方法也应用在上述的云处理系统中，且该云处理系统中的多个边缘服务器共享由中心集群提供的多个公共服务，其中，该数据处理方法可由用于运行目标云应用的目标边缘服务器执行，该目标边缘服务器是该云处理系统中，与目标云应用的客户端之间的物理距离最近的任一个边缘服务器。如图4所示，该方法可包括：
S401，获取目标云应用的运行数据。
该运行数据是由目标边缘服务器获取到的，其中，该运行数据可以是用于对该目标云应用的运行情况进行分析的数据，在一个实施例中，用于对目标云应用的运行情况进行分析的数据包括目标边缘服务器通过应用实例将目标云应用的应用画面下发到相应客户端的过程中获取到的摘要数据，其中，在目标边缘服务器在将应用画面下发到客户端时，是由目标边缘服务器运行该目标云应用的应用实例中包括的传输容器进行下发的，那么，该应用实例中包括的代理容器则可基于该传输容器下发应用画面到客户端的过程，获取得到该目标云应用的摘要数据。在一个实施例中，由于该目标边缘服务器中的应用实例将负责目标云应用的应用画面的传输过程，因此，包括了应用实例的目标边缘服务器也可称为出流节点。那么可以理解，出流节点是与云应用相关的集群中负责运行应用实例的节点，实际包括承载运行容器的服务器，以及负责将编码后的视频流传输给用户的传输容器，其中承载运行容器的服务器例如可以是高级精简指令机器(ARM，Advanced RISC Machines)服务器(一种基于支持移动端应用处理器的服务器)，而承载该传输容器的服务器例如可以是x86(一种支持个人电脑端应用的处理器)或ARM服务器。此外，用于对该目标云应用的运行情况进行分析的数据还可以是该目标云应用在目标边缘服务器中运行时产生的日志数据，该目标边缘服务器可在目标云应用的运行过程中获取到该日志数据，或者也可在该目标云应用运行结束后获取到该日志数据。
在一个实施例中，目标边缘服务器承载运行目标云应用可基于如图5a所示的云应用中台部署架构实现的，可以理解，该云应用中台部署架构也即是上述的云处理系统的中心集群的部署架构，在如图5a所示的部署架构中，节点50为出流节点，虚线框中的服务51为公共服务，且运行公共服务的节点被隔离作为管控节点，管控用于负责管理和控制云应用的应用实例的节点。该类节点用于运行数据库、消息中间件、缓存组件、网关接入组件、用户鉴权组件、出流节点调度组件等云应用的公共组件。基于该部署架构，中心集群可通过各种不同的网关提供网关服务，其中，该中心集群包括的网关可以是用于与业务平台接入的边缘业务网关(Edge gateway，EdgeGW)。该业务平台是指管理后台的运行平台，边缘业务网关通过和业务平台连接，并对管理后台提供同一的入口；该边缘业务网关也可称为后台网关，主要用于针对管理后台过来的请求进行安全校验和根据不通的业务进行转发到正确的内部服务上。另外，在该部署架构中还可包括其他网关，如信令网关(signal GW)和流网关(Stream GW)，该信令网关和流网关可统称为用户网关，用于和用户集群中出流实例的操作对象进行接入。其中，边缘集群的出流实例对应操作对象的接入可通过边缘服务器的出流实例内部进行接入。在一个实施例中，操作对象在接入到该部署架构时，可先在相应的终端设备中安装相应云应用的软件开发安装包(SDK，Software Development Kit)，以使该操作对象的终端设备具备对相应云应用的应用画面的渲染能力，进而可通过特定的输入输出接口和架构中的用户网关进行接入，从而实现将操作对象接入到部署架构中，其中，该输入输出接口例如可以是传统负载均衡(CLB，Classic Load Balancer)的F5接口(一种并发访问接口)。需要说明的是，上述提及的边缘业务网关、信令网关和流网关均分别是独立的网关，也可以是集成在一个网关设备上的不同服务单元。
在一个实施例中，由51标记的虚线框中的公共服务包括的业务层可包括：边缘调度服务(Edge Scheduler)、记录存储服务(Recorder)、配置中心(Conf Center)和编码(Transcode)等。其中，边缘调度服务主要用于管理每个出流实例在出流节点上的生命周期(创建、初始化、删除等等)，记录存储服务也是录屏服务，用于将音视频流转码录制或对接到直播平台，使得操作对象可在进行云应用时进行直接直播，无需消耗云应用对应终端设备的本地处理能力。配置中心用于提供配置能力，比如出流组件的版本，还有运行时对资源的需求(比如需要占用多少处理资源和内存)等。编码则用于对音视频流进行编码处理等。而在出流节点50中，任一出流节点均分别包括运行容器、传输容器和代理容器。另外，该出流节点还可包括证书节点(License Node)，该证书节点用于创建密钥，并将创建的密钥发送给操作对象。而公共服务的证书中心(License Center)则用于对证书节点发送给操作对象的密钥进行安全性校验，其中，该密钥用于提供安全认证能力，出流实例在创建之初会生成若干密钥，这些密钥后续会分配给操作对象，操作对象基于密钥向网关请求建立连接的时候，网关会将该密钥传给认证服务进行认证。
此外，在该部署架构中的公共服务还包括存储服务，用于提供基础的存储功能，该存储服务可由各种存储器来实现，如分布式存储器(kafka)、关系型数据库(MySQL)、基于远程字典服务的存储器(Redis)、基于注册表的存储器(Registry)及文件存储器(Filesever)。另外，该公共服务还包括配额管理服务(Quota)、数据预处理服务(Pre Process)、监测(Monitor)服务、数据(Statistics)服务。其中，配额管理服务主要用于统一计算集群当前对资源的使用情况，比如每个实例创建的时候，需要分配的中央处理器(CPU，Central Processing Unit)/内存/图形处理器(GPU，Graphics Processing Unit，)资源数，因此，该配额管理服务用于对所需的资源数进行统一的计算和分配；数据预处理服务是由于数据处理数据量非常大，传统的流式处理难以满足性能需求，可采用数据预处理服务将每个实例产生的数据先经过预处理服务处理，再将汇总数据放到流式处理流程中进行分析和报表展示(即后续推送到数据中台和运维监测平台进行分析和展示)；监测服务用于对云应用系统各项健康指标的实时监测；数据服务是云应用实例状态汇总的服务，用于给监测器(monitor)提供业务监测的基础数据。
基于上述的针对云应用的中台部署架构，公共服务集中部署在中心集群51，并通过在边缘集群中部署边缘服务器，以实现对中心集群中公共服务的共享复用，其中，边缘集群中的边缘服务器可分别通过两条不同的通信链路实现对中心集群提供的两种不同的公共服务进行共享复用。在一个实施例中，中心集群包括中心代理设备和边缘管理设备，该目标边缘服务器与边缘代理设备相关联；那么，如果对第一类数据进行数据传输的实时性要求低于对第二类数据进行传输的实时性要求，则该边缘代理设备与中心集群包括的中心代理设备之间的通信链路为第一通信链路；而目标边缘服务器与边缘管理设备之间的通信链路为第二通信链路，那么，目标边缘服务器则可基于业务需求分别基于第一通信链路和第二通信链路对中心集群中的公共服务进行调用。在一个实施例中，目标边缘服务器可将获取到的目标云应用的第一类数据通过第一通信链路发送给中心集群，并调用中心集群中的相应服务对该第一类数据进行响应，即转而执行步骤S402，而在目标边缘服务器获取到的是第二类数据时，则将通过第二通信链路将该第二类数据发送给中心集群，以调用中心集群的相应服务对该第二类数据进行响应，即转而执行步骤S403。
S402，当运行数据为第一类数据时，将运行数据通过目标边缘服务器与中心集群之间的第一通信链路发送给中心集群，以使中心集群采用与第一类数据匹配的服务对第一类数据进行响应；第一通信链路是能够满足第一类数据的实时性要求的链路。
S403，当运行数据为第二类数据时，将运行数据通过第二通信链路发送给中心集群，以使中心集群采用与第二类数据匹配的服务对第二类数据进行响应，其中，第二通信链路是能够满足第二类数据的实时性要求的链路，且第一类数据和第二类数据的实时性要求不同。
在步骤S402和步骤S403中，由于目标边缘服务器包括运行容器、传输容器及代理容器。目标云应用运行在目标边缘服务器的运行容器中，传输容器用于分别与客户端及运行容器建立连接，以将运行容器编码后的应用画面发送给客户端进行显示，而代理容器用于为传输容器和运行容器建立通信连接，并与中心集群中的边缘管理设备建立通信连接，因此，第二通信链路是指代理容器和边缘管理设备建立通信连接后的通信链路。在一个实施例中，基于上述的部署架构，将公共服务集中在中心集群进行部署，并部署多个对中心集群中的公共服务进行共享复用的边缘集群的示意图可如图5b所示，其中，该中心集群包括的中心代理设备520为kafka proxy(一种分布式的日志代理)，边缘管理设备521为Edge manager(一种无状态的管理节点)，而在目标边缘服务器关联的边缘代理设备522则为kafka log proxy(一种分布式的日志代理，该日志代理和中心代理设备可以相同或不同)，那么可以理解，该目标边缘服务器与中心集群之间的第一通信链路(或称为数据链路(Data channel))即是由中心代理设备和边缘代理设备之间的数据通信构成，而目标边缘服务器与中心集群之间的第二通信链路则是由边缘管理设备和该目标边缘服务器之间的数据通信构成的。
在一个实施例中，该边缘管理设备也可称为边缘计算管理设备(EC Manager，Edge Computing Manager)，EC Manager是一种无状态的边缘管理设备，主要负责打通边缘服务器和中心集群之间的第二通信链路，并用于部署对边缘服务器进行管理的服务，针对边缘服务器的管理服务是无状态的服务，其中，无状态的服务是指在中心集群中部署的多个管理服务之间，每个管理服务没有区别，如果其中一个管理服务崩溃，另一管理服务可替代崩溃的管理服务执行，从而可保证第二通信链路的稳定性。也就是说，本申请实施例通过在中心集群中部署边缘管理设备，从而可通过该边缘管理设备代理中心集群内的公共组件和目标边缘服务器的出流实例之间的通信流量。其中，该边缘管理设备是通过与出流实例的代理容器建立双向安全传输层协议(TLS，Transport Layer Security)的网络套接字(WebSocket)连接(长连接)来实现通信的。那么可以理解，如果目标边缘服务器获取的运行数据是由目标边缘服务器的代理容器获取到的，那么获取的运行数据则为第二类数据；而在目标边缘服务器将第二类数据通过目标边缘服务器与中心集群之间的第二通信链路发送给中心集群时，即是通过代理容器与中心集群中的边缘管理设备之间的通信连接，将该第二类数据发送给中心集群中的边缘管理设备，从而实现将该第二类数据发送到中心集群中。
在一个实施例中，由于第一类数据包括用于记录应用实例在将应用画面输出到客户端过程中的摘要数据，由此可见，第一通信链路中承载的数据量较大，如在目标云应用当前的日活跃用户数(DAU，Daily Active User)为50万时，通过第一通信链路发送的数据数量级达百亿级别，但第一类数据的实时性要求较低，也就是说，目标边缘服务器可通过第一通信链路进行实时性要求低，但数据量较大的运行数据的传输。如图5b所示的第一通信链路，边缘代理设备包括kafka log proxy，该边缘代理设备在获取到出流实例(即关联的目标边缘服务器)发送的第一类数据(如上述的摘要数据)后，可先将第一类数据落地到本地磁盘中，再通过公网将第一类数据发送到位于中心集群的中心代理设备(kafka proxy)中，因此，如果边缘代理设备的网络出现抖动而断网，第一通信链路将会出现短暂的中断，在该边缘代理设备会持续重连并等待网络重新连接之后，则可通过本地磁盘中存储的第一类数据，将第一类数据进行重新发送，且如果由于设备 或服务升级或其他问题导致的边缘代理设备重新的问题，第一类数据也可从本地磁盘中进行重新加载和续传，以确保第一类数据不被丢失。
在一个实施例中，边缘代理设备和中心代理设备之间采用传输控制协议(TCP，Transmission Control Protocol)建立通信连接，并在中心代理设备前部署有云负载均衡(Cloud Load Balancer，CLB)设备和防火墙，以对产生该日志数据的设备的源IP进行显示，确定第一通信链路的数据安全性。在一个实施例中，负载均衡能提供安全快捷的流量分发服务，访问流量经由CLB可以自动分配到云中的多台云服务器上，扩展系统的服务能力并消除单点故障，且负载均衡支持亿级连接和千万级并发，可轻松应对大流量访问，满足业务需求。另外，中心集群中还可通过消息中间件(MQ，Message Queue)提供消息发送和接收服务，消息中间件是面向消息的系统，在分布式系统中完成消息的发送和接收的基础软件，消息中间件也可以称消息队列，是指用高效可靠的消息传递机制进行与平台无关的数据交流，并基于数据通信来进行分布式系统的集成，通过提供消息传递和消息队列模型，可以在分布式环境下扩展进程的通信。而Psrxy Server(一种部署在中心集群的网络代理)和Psrxy Client(一种部署在边缘服务器的网络代理)作为在打通k8s(一种资源平台)的中心集群和边缘点的代理网络通路，让k8s在管理边缘点的时候可以无需关注该边缘点的网络情况(不管公网还是私网)。其中，k8s也称为Kubernetes，其是一个可移植的、可扩展的开源平台，用于管理容器化的工作负载和服务，可促进声明式配置和自动化，Kubernetes拥有一个庞大且快速增长的生态系统，Kubernetes的服务、支持和工具广泛可用。
在一个实施例中，边缘代理设备中部署有目标资源(DeamonSet)，目标资源用于为边缘代理设备关联预设标签，在一个实施例中，该预设标签例如可以是tencent.cr/ec-role:ec-kafka-proxy，那么，当物理边缘区域中的其他设备被添加预设标签后，被添加预设标签的其他设备被作为边缘代理设备，那么基于在边缘代理设备中部署的目标资源，使得边缘集群中的kafka log proxy组件会自动部署到被部署了目标资源的边缘代理设备中，且与该边缘代理设备关联的边缘服务器也将自动准备好接受位于中心集群的调度服务，从而实现对边缘点(包括边缘服务器和关联的边缘代理设备)的搭建和维护的简化，使边缘点能被自动发现并接入中心集群。
下面，结合图5c，并以目标边缘服务器和该目标边缘服务器关联的边缘代理设备构成目标边缘点为例，对目标边缘点的自动发现流程进行说明。
步骤S501，边缘管理设备对边缘代理设备进行监测。
在一个实施例中，上述提及的第一通信链路和第二通信链路均是在目标边缘点被标记为可用状态后建立的，而目标边缘点是被边缘管理设备标记为可用状态的，那么，该边缘管理设备在将目标边缘点标记为可用状态时，可先由边缘管理设备获取目标边缘点的边缘代理设备的资源对象创建信息。
步骤S502，边缘代理设备创建资源对象。
在一个实施例中，该边缘代理设备创建的资源对象pod为一个或多个容器集合。
步骤S503，边缘管理设备获取对象标识。
在本申请实施例中，在边缘管理设备根据资源对象创建信息确定边缘代理设备创建了资源对象(pod)后，获取边缘代理设备的设备标识及边缘代理设备所在物理边缘区域的区域(zone)标识。
步骤S504，边缘管理设将创建的对象资源的设备标识发送至容器操作平台k8s。
步骤S505，边缘管理设将区域标识发送至k8s。
步骤S506，边缘管理设备在中心集群关联存储设备标识和区域标识。
步骤S507，边缘管理设备标记目标边缘点的可用性。
在一个实施例中，由于边缘代理设备创建的资源对象pod为一个或多个容器集合，那么该边缘管理设备则可在中心集群关联存储设备标识和区域标识后，将目标边缘点标记为可用状态。其中，边缘代理设备的网络类型为宿主机网络类型(即为Host Network)，被配置为宿主机网络类型的边缘代理设备的设备标识(即节点IP，Node IP)，与在被配置为宿主机网络类型的边缘代理设备中创建的资源对象的对象标识一致；那么，该边缘管理设备在获取边缘代理设备的设备标识时，则可获取在边缘代理设备中创建的资源对象的对象标识(即为pod IP)，并将获取的对象标识作为边缘代理设备的设备标识，那么，在边缘管理设备获取到的区域标识也即是相应的边缘服务器所支持的客户端对应终端设备的物理区域(或物理边缘区域)范围，那么，在边缘管理设备将设备标识和区域标识进行关联存储后，即完成了新增边缘点(如上述的目标边缘点)的发现过程。
在一个实施例中，目标边缘点也是由边缘管理设备删除的，那么，边缘管理设备可在边缘管理设备获取到针对目标边缘点的删除触发指令时，停止运行边缘代理设备，并删除在中心集群中关联存储的设备标识和区域标识，而在设备标识和区域标识被删除后，目标边缘点为不可用状态。在一个实施例中，目标云应用被运行在不同边缘点包括的边缘服务器中，一个边缘点对应一个物理边缘区域；任一边缘点的边缘服务器根据相应物理边缘区域的区域标识，将目标云应用的应用画面输出显示到处于物理边缘区域内的客户端中。
S404，通过第二通信链路从中心集群获取响应第二类数据后生成的反馈数据。
S405，采用反馈数据更新应用画面，并在目标云应用的客户端中显示更新后的应用画面。
在步骤S404和步骤S405中，中心集群提供的多种服务至少包括运行分析服务和基础指令服务，第一类数据为支持对目标云应用的运行情况进行分析的数据，第二类数据则为针对目标云应用的基础指令数据。那么当第一类数据为支持对目标云应用的运行情况进行分析的数据时，与第一类数据匹配的服务为运行分析服务；当第二类数据为基础指令数据时，与第二类数据匹配的服务为基础指令服务。在目标边缘服务器将第一类数据通过第一通信链路发送给中心集群，并在中心集群通过运行分析服务对目标云应用的运行情况进行分析后，该目标边缘服务器同样可通过该第一通信链路获取到针对该目标云应用的分析结果，并将获取到的分析结果输出显示到质量监管设备中，或者，针对该目标云应用的质量监管设备也可直接接入到中心集群，并从该中心集群中获取针对该目标云应用的分析结果。
在一个实施例中，目标边缘服务器通过第二通信链路向中心集群发送的实时性要求较高的第二类数据，是该目标边缘服务器通过在客户端显示目标云应用的应用画面后获取到的，那么在目标边缘服务器通过第二通信链路将第二类数据发送给中心集群，并在中心集群通过基础指令服务对该第二类数据进行响应后，该目标边缘服务器将通过第二通信链路从中心集群获取响应该第二类数据后生成的反馈数据，在实现中，该目标边缘服务器可通过目标边缘服务器中的代理容器从中心集群的边缘管理设备中获取反馈数据。
在目标边缘服务器获取到针对该第二类数据的反馈数据后，则可采用反馈数据更新应用画面，并在客户端中显示更新后的应用画面，在一个实施例中，该目标边缘服务器可将反馈数据发送给目标边缘服务器中的运行容器，以在运行容器中采用反馈数据更新应用画面，那么，更新后的应用画面在运行容器中被编码后，通过目标边缘服务器中的传输容器发送到客户端，并在客户端显示更新后的应用画面。也就是说，云处理系统中的中心集群和边缘点(包括边缘服务器和关联的边缘代理设备)在部署时，中心集群中的公共服务可集中管理，如可集中部署在互联网数据中心(IDC，Internet Data Center) 中，边缘点则可选择部署在物理位置与操作对象较近区域的机房，比如在华南地区的中心集群可部署在深圳的互联网数据中心，而边缘点则可分别在惠州和东莞进行部署，以专门针对惠州和东莞的用户提供更加优质和流程的云应用接入体验。可以理解，边缘点在部署时要尽量轻量，边缘点内仅包括支持出流实例运行的出流节点，其他服务将附属于中心集群而存在，减轻边缘点的搭建和维护成本，而在中心集群中同样将部署出流节点，以使该中心集群覆盖的操作对象通过中心集群进行云应用的接入。
在一个实施例中，目标云应用包括目标云游戏，目标云应用的客户端包括目标云游戏对应的游戏客户端；支持对目标云应用的运行情况进行分析的第一类数据包括目标云游戏的日志数据，针对目标云应用的基础指令数据包括游戏基础指令，游戏基础指令包括以下任意一种或多种：存储指令和鉴权指令等。下面结合图5d对目标云游戏的运行过程进行说明。
该云游戏是基于云游戏处理系统(即上述的云处理系统)运行的，该云游戏处理系统包含中心集群和多个边缘服务器，该多个边缘服务器共享由该中心集群提供的质量分析服务531和游戏基础服务532，该目标云游戏将基于与各边缘服务器之间的距离，接入相应距离最近的边缘服务器作为目标边缘服务器，该目标边缘服务器则可通过目标云游戏的游戏实例包括的传输容器与该游戏客户端建立连接，并触发在游戏容器中运行该目标云游戏。在游戏容器中运行目标游戏过程中，该游戏容器将通过传输容器持续发送音视频流到游戏客户端中，并在该游戏客户端中显示相应的游戏画面，此外，在传输容器向游戏客户端中发送音视频流所产生的摘要数据(或运行该云游戏所产生的日志数据)由目标边缘服务器采集到后，通过该目标边缘服务器关联的边缘代理设备与中心集群中的中心代理设备发送到中心集群中，以对该目标云游戏的运行过程进行质量分析。
在目标边缘服务器的游戏容器运行该目标云游戏时，游戏操作用户可通过游戏客户端所在的终端设备发送游戏基础指令，该游戏基础指令则可通过该目标边缘服务器与中心集群中边缘管理设备之间的长连接，将该游戏基础指令发送到中心集群中，并可从该边缘管理设备中获取针对该游戏基础指令的反馈数据，该反馈数据可在游戏容器中对当前显示的游戏画面进行更新后，由游戏容器通过传输容器将更新后的游戏画面下发到游戏客户端进行显示。此外，该目标云游戏处理系统还包括游戏虚拟手柄接入网关(GamePad GW)。其中，该游戏虚拟手柄是为了解决没有实体手柄的时候，游戏操作对象可将终端设备作为接入手柄的实体载体，并通过在终端屏幕上生成一个触屏版的游戏手柄，另外，该中心集群还可提供直播接入的接口，以使在云游戏过程中直接基于中心集群继而到直播平台进行游戏直播。
在本申请实施例中，目标云应用可在云处理系统中运行时，由与该目标云应用对应客户端最近的目标边缘服务器的运行容器执行运行，且该目标边缘服务器是与多个其他边缘服务器共享由中心集群提供的多种服务，从而可保证边缘服务器的轻量部署，并基于对中心集群中各服务的共享复用，可提升中心集群中的资源利用率。而在目标边缘服务器获取到针对目标云应用的运行数据后，在该运行数据为第一类数据时，将该第一类数据通过第一通信链路发送给中心集群进行响应，并该运行数据为第二类数据时，将该第二类数据通过第二通信链路发送给中心集群进行响应，而后则可通过该相应的通信链路获取针对该运行数据的反馈数据，由于云处理系统中的边缘点在部署时，可实现边缘点的自动发现，因此，也就可提升边缘点的部署进度和边缘点发现效率，而基于将数据传输通道的分离，则可有效提升云应用的数据安全性，以及提升云应用的操作流畅度。
基于上述数据处理方法实施例的描述，本申请实施例还提出了一种数据处理装置，应用于云处理系统，所述云处理系统包含中心集群和多个边缘服务器，所述多个边缘服务器共享由所述中心集群提供的多种服务，所述数据处理装置可以是运行于上述目标边 缘服务器中的一个计算机程序(包括程序代码)，所述目标边缘服务器包括所述多个边缘服务器中，与目标云应用的客户端之间的物理距离最近的任一个边缘服务器。该数据处理装置可用于执行如图2和图4所述的数据处理方法，请参见图6，该数据处理装置包括：获取单元601和发送单元602。
获取单元601，配置为获取目标云应用的运行数据；
发送单元602，配置为当所述运行数据为第一类数据时，将所述运行数据通过所述目标边缘服务器与所述中心集群之间的第一通信链路发送给所述中心集群，以使所述中心集群采用与所述第一类数据匹配的服务对所述第一类数据进行响应；所述第一通信链路是与所述第一类数据的实时性要求相匹配的链路；
所述发送单元602，还配置为当所述运行数据为第二类数据时，将所述运行数据通过第二通信链路发送给所述中心集群，以使所述中心集群采用与所述第二类数据匹配的服务对所述第二类数据进行响应，其中，所述第二通信链路是与所述第二类数据的实时性要求相匹配的链路，且所述第一类数据和所述第二类数据的实时性要求不同。
在一个实施例中，所述中心集群包括中心代理设备和边缘管理设备，所述目标边缘服务器与边缘代理设备相关联；若所述第一类数据的实时性要求低于所述第二类数据的实时性要求，所述边缘代理设备与所述中心代理设备之间的通信链路为所述第一通信链路；所述目标边缘服务器与所述边缘管理设备之间的通信链路为所述第二通信链路。
在一个实施例中，所述目标边缘服务器包括运行容器、传输容器及代理容器；所述目标云应用运行在所述目标边缘服务器的运行容器中，所述传输容器用于分别与所述客户端及所述运行容器建立连接，以对所述运行容器生成的所述目标云应用的应用画面进行编码，并将编码后的应用画面发送给所述客户端进行显示；所述代理容器用于为所述传输容器和所述运行容器建立通信连接，并与所述中心集群中的边缘管理设备建立通信连接，其中，所述第二通信链路是指代理容器和所述边缘管理设备建立通信连接后的通信链路。
在一个实施例中，若所述运行数据是由所述目标边缘服务器的代理容器获取到的，所述运行数据为所述第二类数据；所述发送单元602，配置为：当所述运行数据为第二类数据时，将所述运行数据通过所述代理容器与所述中心集群中的边缘管理设备之间的通信连接，发送给所述中心集群中的边缘管理设备。
在一个实施例中，所述中心集群提供的多种服务至少包括运行分析服务和基础指令服务，所述第一类数据为支持对所述目标云应用的运行情况进行分析的数据，所述第二类数据则为针对所述目标云应用的基础指令数据；其中，当所述第一类数据为支持对所述目标云应用的运行情况进行分析的数据时，与所述第一类数据匹配的服务为所述运行分析服务；当所述第二类数据为所述基础指令数据时，与所述第二类数据匹配的服务为所述基础指令服务。
在一个实施例中，所述基础指令数据是通过在所述目标云应用的客户端显示所述目标云应用的应用画面后获取到的；所述装置还包括处理单元603。
所述获取单元601，还配置为通过所述第二通信链路从所述中心集群获取响应所述第二类数据后生成的反馈数据；
处理单元603，配置为采用所述反馈数据更新所述应用画面，将更新后的应用画面发送到所述目标云应用的客户端，以在所述目标云应用的客户端中显示更新后的应用画面。
在一个实施例中，所述获取单元601，还配置为：通过所述目标边缘服务器中的代理容器从所述中心集群的边缘管理设备中获取所述反馈数据；
所述处理单元603，还配置为：将所述反馈数据发送给所述目标边缘服务器中的运 行容器，以在所述运行容器中采用所述反馈数据更新所述应用画面；更新后的应用画面在所述运行容器中被编码后，通过所述目标边缘服务器中的传输容器发送到所述目标云应用的客户端，并在所述目标云应用的客户端显示所述更新后的应用画面。
在一个实施例中，所述目标云应用包括目标云游戏，所述目标云应用的客户端包括所述目标云游戏对应的游戏客户端；支持对所述目标云应用的运行情况进行分析的第一类数据包括所述目标云游戏的日志数据，针对所述目标云应用的基础指令数据包括游戏基础指令，所述游戏基础指令包括以下至少一种：存储指令和鉴权指令。
在一个实施例中，所述目标边缘服务器和所述目标边缘服务器关联的边缘代理设备构成目标边缘点，所述第一通信链路和所述第二通信链路均是在所述目标边缘点被标记为可用状态后建立的；所述目标边缘点是被边缘管理设备标记为可用状态的，所述边缘管理设备将所述目标边缘点标记为可用状态的方式时，由所述边缘管理设备获取所述目标边缘点关联的边缘代理设备的资源对象创建信息，并在所述边缘管理设备根据所述资源对象创建信息确定所述边缘代理设备创建了资源对象后，获取所述边缘代理设备的设备标识及所述边缘代理设备所在物理边缘区域的区域标识；在所述中心集群关联存储所述设备标识和所述区域标识后，将所述目标边缘点标记为可用状态。
在一个实施例中，所述边缘代理设备的网络类型为宿主机网络类型，被配置为所述宿主机网络类型的边缘代理设备的设备标识，与在被配置为所述宿主机网络类型的边缘代理设备中创建的资源对象的对象标识一致；所述获取所述边缘代理设备的设备标识时，获取在所述边缘代理设备中创建的资源对象的对象标识，并将获取的对象标识作为所述边缘代理设备的设备标识。
在一个实施例中，所述边缘代理设备中部署有目标资源，所述目标资源用于为所述边缘代理设备关联预设标签；当所述物理边缘区域中的其他设备被添加所述预设标签后，被添加所述预设标签的其他设备被作为所述边缘代理设备。
在一个实施例中，所述目标边缘点也是由所述边缘管理设备删除的，由所述边缘管理设备删除所述目标边缘点的方式为，当所述边缘管理设备获取到针对所述目标边缘点的删除触发指令时，停止运行所述边缘代理设备，并删除在所述中心集群中关联存储的所述设备标识和所述区域标识；在所述设备标识和所述区域标识被删除后，所述目标边缘点为不可用状态。
在一个实施例中，所述目标云应用被运行在不同边缘点包括的边缘服务器中，一个边缘点对应一个物理边缘区域；任一边缘点的边缘服务器根据相应物理边缘区域的区域标识，将所述目标云应用的应用画面输出显示到处于所述物理边缘区域内的客户端中。
在本申请实施例中，基于云处理系统对提供公共服务的中心集群的集中管理，该云处理系统中的任一边缘服务器可通过对中心集群中相应服务组件的调用进行公共服务的共享复用，从而可避免对公共组件的重复部署，提升中心集群的资源利用率。由于该云处理系统中的边缘服务器与中心集群之间的通信链路分别包括第一通信链路和第二通信链路，那么，如果获取单元601获取到目标云应用的运行数据后，发送单元602可在确定该运行数据为第一类数据时，通过该第一通信链路将该运行数据发送给中心集群，并在中心集群中通过调用相应的服务对该第一类数据进行响应，而如果获取单元601获取到针对该目标云应用的运行数据为第二类数据，则发送单元602可将该第二类数据通过与中心集群之间的第二通信链路发送给中心集群，并由中心集群调用相应的服务进行响应，由于在对中心集群中提供的公共服务进行调用时，将通过不同的通信链路来实现，也就是说，与中心集群之间的不同业务交互和数据传输是基于不同的通信链路来实现的，也就在基于不同业务对中心集群中的公共服务进行调用时，不同业务的数据之间 的数据通路是相互独立且互不影响的，从而可实现对不同业务的并发执行，并可提升各业务数据传输的安全性。
请参见图7，是本申请实施例提供的一种边缘服务器(如上述的目标边缘服务器)的结构示意性框图。其中，目标边缘服务器与其他多个边缘服务器共享由数据处理系包含的中心集群提供的多种服务，该目标边缘服务器为与目标云应用的客户端之间的物理距离最近的任一个边缘服务器，如图7所示的本实施例中的边缘服务器可包括：一个或多个处理器701；一个或多个输入设备702，一个或多个输出设备703和存储器704。上述处理器701、输入设备702、输出设备703和存储器704通过总线705连接。存储器704用于存储计算机程序，所述计算机程序包括程序指令，处理器701用于执行所述存储器704存储的程序指令。
所述存储器704可以包括易失性存储器(Volatile Memory)，如随机存取存储器(RAM，Random-Access Memory)；存储器704也可以包括非易失性存储器(non-volatile memory)，如快闪存储器(Flash Memory)，固态硬盘(SSD，Solid-State Drive)等；存储器704还可以包括上述种类的存储器的组合。
所述处理器701可以是CPU。所述处理器701还可以进一步包括硬件芯片。上述硬件芯片可以是专用集成电路(ASIC，Application-Specific Integrated Circuit)，可编程逻辑器件(PLD，Programmable Logic Device)等。该PLD可以是现场可编程逻辑门阵列(FPGA，Field-Programmable Gate array)，通用阵列逻辑(GAL，Generic Array Logic)等。所述处理器701也可以为上述结构的组合。
本申请实施例中，所述存储器704用于存储计算机程序，所述计算机程序包括程序指令，处理器701用于执行存储器704存储的程序指令，用来实现上述如图2和图4中相应方法的步骤。
在一个实施例中，所述处理器701被配置调用所述程序指令，用于执行：获取目标云应用的运行数据；当所述运行数据为第一类数据时，将所述运行数据通过所述目标边缘服务器与所述中心集群之间的第一通信链路发送给所述中心集群，以使所述中心集群采用与所述第一类数据匹配的服务对所述第一类数据进行响应；所述第一通信链路是与所述第一类数据的实时性要求相匹配的链路；当所述运行数据为第二类数据时，将所述运行数据通过第二通信链路发送给所述中心集群，以使所述中心集群采用与所述第二类数据匹配的服务对所述第二类数据进行响应，其中，所述第二通信链路是与所述第二类数据的实时性要求相匹配的链路，且所述第一类数据和所述第二类数据的实时性要求不同。
在一个实施例中，所述中心集群包括中心代理设备和边缘管理设备，所述目标边缘服务器与边缘代理设备相关联；若所述第一类数据的实时性要求低于所述第二类数据的实时性要求，所述边缘代理设备与所述中心代理设备之间的通信链路为所述第一通信链路；所述目标边缘服务器与所述边缘管理设备之间的通信链路为所述第二通信链路。
在一个实施例中，所述目标边缘服务器包括运行容器、传输容器及代理容器；所述目标云应用运行在所述目标边缘服务器的运行容器中，所述传输容器用于分别与所述客户端及所述运行容器建立连接，以对所述运行容器生成的所述目标云应用的应用画面进行编码，并将编码后的应用画面发送给所述客户端进行显示；所述代理容器用于为所述传输容器和所述运行容器建立通信连接，并与所述中心集群中的边缘管理设备建立通信连接，其中，所述第二通信链路是指代理容器和所述边缘管理设备建立通信连接后的通信链路。
在一个实施例中，若所述运行数据是由所述目标边缘服务器的代理容器获取到的，所述运行数据为所述第二类数据；所述处理器701被配置调用所述程序指令，用于执行： 当所述运行数据为第二类数据时，将所述运行数据通过所述代理容器与所述中心集群中的边缘管理设备之间的通信连接，发送给所述中心集群中的边缘管理设备。
在一个实施例中，所述中心集群提供的多种服务至少包括运行分析服务和基础指令服务，所述第一类数据为支持对所述目标云应用的运行情况进行分析的数据，所述第二类数据则为针对所述目标云应用的基础指令数据；其中，当所述第一类数据为支持对所述目标云应用的运行情况进行分析的数据时，与所述第一类数据匹配的服务为所述运行分析服务；当所述第二类数据为所述基础指令数据时，与所述第二类数据匹配的服务为所述基础指令服务。
在一个实施例中，所述基础指令数据是通过在所述目标云应用的客户端显示所述目标云应用的应用画面后获取到的；所述处理器701被配置调用所述程序指令，用于执行：通过所述第二通信链路从所述中心集群获取响应所述第二类数据后生成的反馈数据；采用所述反馈数据更新所述应用画面，将更新后的应用画面发送到所述目标云应用的客户端，以在所述目标云应用的客户端中显示更新后的应用画面。
在一个实施例中，所述处理器701被配置调用所述程序指令，用于执行：通过所述目标边缘服务器中的代理容器从所述中心集群的边缘管理设备中获取所述反馈数据；
所述处理器701被配置调用所述程序指令，用于执行：将所述反馈数据发送给所述目标边缘服务器中的运行容器，以在所述运行容器中采用所述反馈数据更新所述应用画面；更新后的应用画面在所述运行容器中被编码后，通过所述目标边缘服务器中的传输容器发送到所述目标云应用的客户端，并在所述目标云应用的客户端显示所述更新后的应用画面。
在一个实施例中，所述目标云应用包括目标云游戏，所述目标云应用的客户端包括所述目标云游戏对应的游戏客户端；
支持对所述目标云应用的运行情况进行分析的第一类数据包括所述目标云游戏的日志数据，针对所述目标云应用的基础指令数据包括游戏基础指令，所述游戏基础指令包括以下至少一种：存储指令和鉴权指令。
在一个实施例中，所述目标边缘服务器和所述目标边缘服务器关联的边缘代理设备构成目标边缘点，所述第一通信链路和所述第二通信链路均是在所述目标边缘点被标记为可用状态后建立的；所述目标边缘点是被边缘管理设备标记为可用状态的，所述处理器701被配置调用所述程序指令，用于执行：由所述边缘管理设备获取所述目标边缘点关联的边缘代理设备的资源对象创建信息，并在所述边缘管理设备根据所述资源对象创建信息确定所述边缘代理设备创建了资源对象后，获取所述边缘代理设备的设备标识及所述边缘代理设备所在物理边缘区域的区域标识；在所述中心集群关联存储所述设备标识和所述区域标识后，将所述目标边缘点标记为可用状态。
在一个实施例中，所述边缘代理设备的网络类型为宿主机网络类型，被配置为所述宿主机网络类型的边缘代理设备的设备标识，与在被配置为所述宿主机网络类型的边缘代理设备中创建的资源对象的对象标识一致；所述处理器701被配置调用所述程序指令，用于执行：获取在所述边缘代理设备中创建的资源对象的对象标识，并将获取的对象标识作为所述边缘代理设备的设备标识。
在一个实施例中，所述边缘代理设备中部署有目标资源，所述目标资源用于为所述边缘代理设备关联预设标签；当所述物理边缘区域中的其他设备被添加所述预设标签后，被添加所述预设标签的其他设备被作为所述边缘代理设备。
在一个实施例中，所述目标边缘点也是由所述边缘管理设备删除的，所述处理器701被配置调用所述程序指令，用于执行：当所述边缘管理设备获取到针对所述目标边缘点的删除触发指令时，停止运行所述边缘代理设备，并删除在所述中心集群中关联存 储的所述设备标识和所述区域标识；在所述设备标识和所述区域标识被删除后，所述目标边缘点为不可用状态。
在一个实施例中，所述目标云应用被运行在不同边缘点包括的边缘服务器中，一个边缘点对应一个物理边缘区域；任一边缘点的边缘服务器根据相应物理边缘区域的区域标识，将所述目标云应用的应用画面输出显示到处于所述物理边缘区域内的客户端中。
本申请实施例提供了一种计算机程序产品或计算机程序，该计算机程序产品或计算机程序包括计算机指令，该计算机指令存储在计算机可读存储介质中。计算机设备的处理器从计算机可读存储介质读取该计算机指令，处理器执行该计算机指令，使得该计算机设备执行上述如图2或图4所示的方法实施例。其中，所述的计算机可读存储介质可为磁碟、光盘、只读存储记忆体(Read-Only Memory，ROM)或随机存储记忆体(Random Access Memory，RAM)等。
以上所揭露的仅为本申请的局部实施例而已，当然不能以此来限定本申请之权利范围，本领域普通技术人员可以理解实现上述实施例的全部或局部流程，并依本申请权利要求所作的等同变化，仍属于发明所涵盖的范围。

Claims (18)

  1. 一种数据处理方法，应用于云处理系统，所述云处理系统包含中心集群和多个边缘服务器，所述多个边缘服务器共享由所述中心集群提供的多种服务，所述方法由目标边缘服务器执行，所述目标边缘服务器包括所述多个边缘服务器中，与目标云应用的客户端之间物理距离最近的任一个边缘服务器，所述方法包括：
    获取所述目标云应用的运行数据；
    当所述运行数据为第一类数据时，将所述运行数据通过所述目标边缘服务器与所述中心集群之间的第一通信链路发送给所述中心集群，以使所述中心集群采用与所述第一类数据匹配的服务对所述第一类数据进行响应；所述第一通信链路是与所述第一类数据的实时性要求相匹配的链路；
    当所述运行数据为第二类数据时，将所述运行数据通过第二通信链路发送给所述中心集群，以使所述中心集群采用与所述第二类数据匹配的服务对所述第二类数据进行响应，其中，所述第二通信链路是与所述第二类数据的实时性要求相匹配的链路，且所述第一类数据和所述第二类数据的实时性要求不同。
  2. 根据权利要求1所述的方法，其中，所述中心集群包括中心代理设备和边缘管理设备，所述目标边缘服务器与边缘代理设备相关联；
    若所述第一类数据的实时性要求低于所述第二类数据的实时性要求，所述边缘代理设备与所述中心代理设备之间的通信链路为所述第一通信链路；
    所述目标边缘服务器与所述边缘管理设备之间的通信链路为所述第二通信链路。
  3. 根据权利要求1所述的方法，其中，所述目标边缘服务器包括运行容器、传输容器及代理容器；
    所述目标云应用运行在所述目标边缘服务器的运行容器中，所述传输容器用于分别与所述客户端及所述运行容器建立连接，以对所述运行容器生成的所述目标云应用的应用画面进行编码，并将编码后的应用画面发送给所述客户端进行显示；
    所述代理容器用于为所述传输容器和所述运行容器建立通信连接，并与所述中心集群中的边缘管理设备建立通信连接，其中，所述第二通信链路是所述代理容器和所述边缘管理设备建立通信连接后的通信链路。
  4. 根据权利要求3所述的方法，其中，若所述运行数据是由所述目标边缘服务器的代理容器获取到的，所述运行数据为所述第二类数据；所述当所述运行数据为第二类数据时，将所述运行数据通过第二通信链路发送给所述中心集群，包括：
    当所述运行数据为第二类数据时，将所述运行数据通过所述代理容器与所述中心集群中的边缘管理设备之间的通信连接，发送给所述中心集群中的边缘管理设备。
  5. 根据权利要求1所述的方法，其中，所述中心集群提供的多种服务至少包括运行分析服务和基础指令服务，所述第一类数据为支持对所述目标云应用的运行情况进行分析的数据，所述第二类数据为针对所述目标云应用的基础指令数据；
    其中，当所述第一类数据为支持对所述目标云应用的运行情况进行分析的数据时，与所述第一类数据匹配的服务为所述运行分析服务；当所述第二类数据为所述基础指令数据时，与所述第二类数据匹配的服务为所述基础指令服务。
  6. 根据权利要求5所述的方法，其中，所述基础指令数据是通过在所述目标云应用的客户端显示所述目标云应用的应用画面后获取到的；所述方法还包括：
    通过所述第二通信链路从所述中心集群获取响应所述第二类数据生成的反馈数据；
    采用所述反馈数据更新所述应用画面，将更新后的应用画面发送到所述目标云应用的客户端，以在所述目标云应用的客户端中显示更新后的应用画面。
  7. 根据权利要求6所述的方法，其中，所述通过所述第二通信链路从所述中心集群获取响应所述第二类数据生成的反馈数据，包括：通过所述目标边缘服务器中的代理容器从所述中心集群的边缘管理设备中获取所述反馈数据；
    所述采用所述反馈数据更新所述应用画面，将更新后的应用画面发送到所述目标云应用的客户端，以在所述目标云应用的客户端中显示更新后的应用画面，包括：将所述反馈数据发送给所述目标边缘服务器中的运行容器，以在所述运行容器中采用所述反馈数据更新所述应用画面；更新后的应用画面在所述运行容器中被编码后，通过所述目标边缘服务器中的传输容器发送到所述目标云应用的客户端，以在所述目标云应用的客户端显示所述更新后的应用画面。
  8. 根据权利要求5所述的方法，其中，所述目标云应用包括目标云游戏，所述目标云应用的客户端包括所述目标云游戏对应的游戏客户端；
    支持对所述目标云应用的运行情况进行分析的第一类数据包括所述目标云游戏的日志数据，针对所述目标云应用的基础指令数据包括游戏基础指令，所述游戏基础指令包括以下至少一种：存储指令和鉴权指令。
  9. 根据权利要求1所述的方法，其中，所述目标边缘服务器和所述目标边缘服务器关联的边缘代理设备构成目标边缘点，所述第一通信链路和所述第二通信链路均是在所述目标边缘点被标记为可用状态后建立的；所述目标边缘点是被边缘管理设备标记为可用状态的，所述边缘管理设备将所述目标边缘点标记为可用状态的方式包括：
    由所述边缘管理设备获取所述目标边缘服务器关联的边缘代理设备的资源对象创建信息，并在所述边缘管理设备根据所述资源对象创建信息确定所述边缘代理设备创建了资源对象后，获取所述边缘代理设备的设备标识及所述边缘代理设备所在物理边缘区域的区域标识；
    在所述中心集群关联存储所述设备标识和所述区域标识后，将所述目标边缘点标记为可用状态。
  10. 根据权利要求9所述的方法，其中，所述边缘代理设备的网络类型为宿主机网络类型，被配置为所述宿主机网络类型的边缘代理设备的设备标识，与在被配置为所述宿主机网络类型的边缘代理设备中创建的资源对象的对象标识一致；所述获取所述边缘代理设备的设备标识，包括：
    获取在所述边缘代理设备中创建的资源对象的对象标识，并将获取的对象标识作为所述边缘代理设备的设备标识。
  11. 根据权利要求9所述的方法，其中，所述边缘代理设备中部署有目标资源，所述目标资源用于为所述边缘代理设备关联预设标签；
    当所述物理边缘区域中的其他设备被添加所述预设标签后，被添加所述预设标签的其他设备被作为所述边缘代理设备。
  12. 根据权利要求9所述的方法，其中，所述目标边缘点也是由所述边缘管理设备删除的，由所述边缘管理设备删除所述目标边缘点的方式包括：
    当所述边缘管理设备获取到针对所述目标边缘点的删除触发指令时，停止运行所述边缘代理设备，并删除在所述中心集群中关联存储的所述设备标识和所述区域标识；
    在所述设备标识和所述区域标识被删除后，所述目标边缘点为不可用状态。
  13. 根据权利要求9所述的方法，其中，所述目标云应用被运行在不同边缘点包括的边缘服务器中，一个边缘点对应一个物理边缘区域；
    任一边缘点的边缘服务器根据相应物理边缘区域的区域标识，将所述目标云应用的应用画面输出显示到处于所述物理边缘区域内的客户端中。
  14. 一种云处理系统，所述云处理系统包含中心集群和多个边缘服务器，所述多个 边缘服务器共享由所述中心集群提供的多种服务，所述多个边缘服务器包括目标边缘服务器，所述目标边缘服务器包括所述多个边缘服务器中，与目标云应用的客户端之间的物理距离最近的任一个边缘服务器；
    所述目标边缘服务器配置为获取目标云应用的运行数据，并在所述运行数据为第一类数据时，通过与所述中心集群之间的第一通信链路，将所述运行数据发送给所述中心集群；以及，在所述运行数据为第二类数据时，通过与所述中心集群之间的第二通信链路，将所述运行数据发送给所述中心集群；所述第一通信链路是与所述第一类数据的实时性要求相匹配的链路，所述第二通信链路是与所述第二类数据的实时性要求相匹配的链路，且所述第一类数据和所述第二类数据的实时性要求不同；
    所述中心集群配置为接收由所述目标边缘服务器发送的运行数据，并在所述运行数据为第一类数据时，通过与所述第一类数据匹配的服务对所述第一类数据进行响应；以及，所述中心集群还在所述运行数据为第二类数据时，通过与所述第二类数据匹配的服务对所述第二类数据进行响应。
  15. 一种数据处理装置，应用于云处理系统，所述云处理系统包含中心集群和多个边缘服务器，所述多个边缘服务器共享由所述中心集群提供的多种服务，所述数据处理装置是运行于目标边缘服务器中的一个计算机程序，所述目标边缘服务器包括所述多个边缘服务器中，与目标云应用的客户端之间的物理距离最近的任一个边缘服务器，所述装置包括：
    获取单元，配置为获取目标云应用的运行数据；
    发送单元，配置为当所述运行数据为第一类数据时，将所述运行数据通过所述目标边缘服务器与所述中心集群之间的第一通信链路发送给所述中心集群，以使所述中心集群采用与所述第一类数据匹配的服务对所述第一类数据进行响应；所述第一通信链路是与所述第一类数据的实时性要求相匹配的链路；
    所述发送单元，还配置为当所述运行数据为第二类数据时，将所述运行数据通过第二通信链路发送给所述中心集群，以使所述中心集群采用与所述第二类数据匹配的服务对所述第二类数据进行响应，其中，所述第二通信链路是与所述第二类数据的实时性要求相匹配的链路，且所述第一类数据和所述第二类数据的实时性要求不同。
  16. 一种边缘服务器，包括处理器、输入设备、输出设备和存储器，所述处理器、所述输入设备、所述输出设备和所述存储器相互建立有通信连接，其中，所述存储器用于存储计算机程序，所述计算机程序包括程序指令，所述处理器被配置为调用所述程序指令，执行如权利要求1至13任一项所述的方法。
  17. 一种计算机可读存储介质，所述计算机可读存储介质存储有计算机程序，所述计算机程序包括程序指令，所述程序指令当被处理器执行时使所述处理器执行如权利要求1至13任一项所述的方法。
  18. 一种计算机程序产品，所述计算机产品包括计算机程序或计算机指令，所述计算机程序或所述计算机指令在被处理器执行时用于实现如权利要求1至13任一项所述的方法。
PCT/CN2022/117512 2021-11-02 2022-09-07 数据处理方法、系统、相关设备、存储介质及产品 Ceased WO2023077952A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
EP22888983.8A EP4379545B1 (en) 2021-11-02 2022-09-07 Data processing method and system, related device, storage medium and product
US18/200,437 US12452339B2 (en) 2021-11-02 2023-05-22 Data processing method and system, related device, storage medium, and product for an edge server in cloud computing system

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202111286046.0A CN113722077B (zh) 2021-11-02 2021-11-02 数据处理方法、系统、相关设备、存储介质及产品
CN202111286046.0 2021-11-02

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US18/200,437 Continuation US12452339B2 (en) 2021-11-02 2023-05-22 Data processing method and system, related device, storage medium, and product for an edge server in cloud computing system

Publications (1)

Publication Number Publication Date
WO2023077952A1 true WO2023077952A1 (zh) 2023-05-11

Family

ID=78686412

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/117512 Ceased WO2023077952A1 (zh) 2021-11-02 2022-09-07 数据处理方法、系统、相关设备、存储介质及产品

Country Status (4)

Country Link
US (1) US12452339B2 (zh)
EP (1) EP4379545B1 (zh)
CN (1) CN113722077B (zh)
WO (1) WO2023077952A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117939503A (zh) * 2024-03-21 2024-04-26 中国电子科技集团公司第二十八研究所 基于云边协同架构的低空空域信息汇集处理方法及系统

Families Citing this family (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113722077B (zh) * 2021-11-02 2022-03-15 腾讯科技（深圳）有限公司 数据处理方法、系统、相关设备、存储介质及产品
CN114513524B (zh) * 2022-02-15 2023-08-29 北京百度网讯科技有限公司 一种资源同步方法、装置、电子设备和存储介质
CN114710485B (zh) * 2022-02-21 2023-10-27 联想(北京)有限公司 处理方法和处理装置
CN116684351B (zh) * 2022-02-22 2026-02-17 腾讯科技（深圳）有限公司 数据传输方法、装置、设备、存储介质及计算机程序产品
CN114567648B (zh) * 2022-03-24 2024-07-02 阿里巴巴（中国）有限公司 分布式云系统
CN114760307B (zh) * 2022-03-31 2024-06-21 度小满科技(北京)有限公司 数据处理方法、装置、存储介质和处理器
CN114528225B (zh) * 2022-04-22 2022-06-21 深圳市思迅网络科技有限公司 一种断网可续用的SaaS软件开发方法、装置及系统
CN114860349B (zh) * 2022-07-06 2022-11-08 深圳华锐分布式技术股份有限公司 数据加载方法、装置、设备及介质
CN115277569B (zh) * 2022-07-21 2025-03-21 Oppo广东移动通信有限公司 多媒体通信方法、系统、相关设备及存储介质
CN116932501A (zh) * 2022-09-09 2023-10-24 深圳前海环融联易信息科技服务有限公司 日志处理系统、方法、装置、设备、介质和产品
CN116009769B (zh) * 2022-12-05 2025-11-14 平安银行股份有限公司 分片系统及其流量数据分发方法
CN115964181B (zh) * 2023-03-10 2023-06-06 之江实验室 一种数据处理的方法、装置、存储介质及电子设备
KR20260017467A (ko) * 2023-06-05 2026-02-05 알리바이 엘엘씨 불변 디지털 증언을 제공하는 시스템, 방법, 및 컴퓨터 프로그램 제품
CN117407357A (zh) * 2023-09-26 2024-01-16 中邮理财有限责任公司 系统间的文件交互方法、文件交换平台与电子设备
CN117793210B (zh) * 2023-11-22 2024-08-09 江苏云工场信息技术有限公司 一种边缘云数据处理方法及系统
CN117579679B (zh) * 2024-01-15 2024-04-19 海马云（天津）信息技术有限公司 信令交互方法和装置、电子设备及存储介质

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104780164A (zh) * 2015-03-25 2015-07-15 腾讯科技（深圳）有限公司 一种运行游戏客户端的方法和装置
CN105681423A (zh) * 2016-01-21 2016-06-15 山东省计算中心（国家超级计算济南中心） 一种基于多路径传输的农机调度方法及监控终端
CN111882626A (zh) * 2020-08-06 2020-11-03 腾讯科技（深圳）有限公司 图像处理方法、装置、服务器及介质
US20200346109A1 (en) * 2019-05-02 2020-11-05 Sony Interactive Entertainment Inc. Method of and System for Controlling the Rendering of a Video Game Instance
CN112448858A (zh) * 2021-02-01 2021-03-05 腾讯科技（深圳）有限公司 网络通信控制方法及装置、电子设备和可读存储介质
CN113018851A (zh) * 2021-05-28 2021-06-25 腾讯科技（深圳）有限公司 游戏处理方法、装置、计算机设备及存储介质
CN113722077A (zh) * 2021-11-02 2021-11-30 腾讯科技（深圳）有限公司 数据处理方法、系统、相关设备、存储介质及产品

Family Cites Families (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10038721B2 (en) * 2015-02-16 2018-07-31 International Business Machines Corporation Enabling an on-premises resource to be exposed to a public cloud application securely and seamlessly
US10079877B2 (en) * 2016-02-04 2018-09-18 Citrix Systems, Inc. System and method for cloud aware application delivery controller
US11016824B1 (en) * 2017-06-12 2021-05-25 Pure Storage, Inc. Event identification with out-of-order reporting in a cloud-based environment
CN110557419A (zh) * 2018-06-01 2019-12-10 杭州海康威视数字技术股份有限公司 一种任务处理方法、装置及云计算系统
EP3591938A1 (en) * 2018-07-03 2020-01-08 Electronics and Telecommunications Research Institute System and method to control a cross domain workflow based on a hierachical engine framework
US11206195B2 (en) * 2019-10-04 2021-12-21 Verizon Patent And Licensing Inc. Cloud computing environment with split connectivity and application interfaces that enable support of separate cloud services
US11489914B2 (en) * 2020-01-19 2022-11-01 Bernt Erik Bjontegard Contextual intelligence of every-thing communication platform integrating edge computation, augmented intelligence and distributed ledger data storage nodes and tokens for real world data interactions across internet of things with payment integration
CN111629024A (zh) * 2020-04-02 2020-09-04 北京大米科技有限公司 一种数据传输控制方法、装置、存储介质及电子设备
CN111698470B (zh) * 2020-06-03 2021-09-03 中科民盛安防(河南)有限公司 一种基于云边协同计算的安防视频监控系统及其实现方法
CN112925646A (zh) * 2021-03-12 2021-06-08 威胜信息技术股份有限公司 一种电力数据边缘计算系统以及计算方法
US12307303B2 (en) * 2021-04-09 2025-05-20 Oracle International Corporation Cloud edge device virtualization
CN113018871A (zh) * 2021-04-19 2021-06-25 腾讯科技（深圳）有限公司 业务处理方法、装置及存储介质

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104780164A (zh) * 2015-03-25 2015-07-15 腾讯科技（深圳）有限公司 一种运行游戏客户端的方法和装置
CN105681423A (zh) * 2016-01-21 2016-06-15 山东省计算中心（国家超级计算济南中心） 一种基于多路径传输的农机调度方法及监控终端
US20200346109A1 (en) * 2019-05-02 2020-11-05 Sony Interactive Entertainment Inc. Method of and System for Controlling the Rendering of a Video Game Instance
CN111882626A (zh) * 2020-08-06 2020-11-03 腾讯科技（深圳）有限公司 图像处理方法、装置、服务器及介质
CN112448858A (zh) * 2021-02-01 2021-03-05 腾讯科技（深圳）有限公司 网络通信控制方法及装置、电子设备和可读存储介质
CN113018851A (zh) * 2021-05-28 2021-06-25 腾讯科技（深圳）有限公司 游戏处理方法、装置、计算机设备及存储介质
CN113722077A (zh) * 2021-11-02 2021-11-30 腾讯科技（深圳）有限公司 数据处理方法、系统、相关设备、存储介质及产品

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP4379545A4

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117939503A (zh) * 2024-03-21 2024-04-26 中国电子科技集团公司第二十八研究所 基于云边协同架构的低空空域信息汇集处理方法及系统
CN117939503B (zh) * 2024-03-21 2024-05-31 中国电子科技集团公司第二十八研究所 基于云边协同架构的低空空域信息汇集处理方法及系统

Also Published As

Publication number Publication date
US20230300203A1 (en) 2023-09-21
US12452339B2 (en) 2025-10-21
EP4379545A1 (en) 2024-06-05
EP4379545A4 (en) 2024-12-04
EP4379545B1 (en) 2026-03-25
CN113722077A (zh) 2021-11-30
CN113722077B (zh) 2022-03-15

Similar Documents

Publication Publication Date Title
WO2023077952A1 (zh) 数据处理方法、系统、相关设备、存储介质及产品
US20220224619A1 (en) Api dependency error and latency injection
US11418613B2 (en) Systems and methods for recording metadata about microservices for requests to the microservices
US10983769B2 (en) Systems and methods for using a call chain to identify dependencies among a plurality of microservices
CN111309374B (zh) 一种微服务系统和微服务系统中的服务调用方法
US11057271B2 (en) Systems and method updating ADC configuration with intended state using desired state API
US11032396B2 (en) Systems and methods for managing client requests to access services provided by a data center
CN105681462B (zh) 一种基于消息路由的集群系统及数据通信中转方法
CN101969391B (zh) 一种支持融合网络业务的云平台及其工作方法
CN115580667B (zh) 数据传输方法、装置、设备及存储介质
CN105577801B (zh) 一种业务加速方法及装置
CN102523207A (zh) 基于虚拟网络计算机的远程资源访问方法及代理设备
US20210377294A1 (en) Constraining resource allocation rate for stateful multi-tenant http proxies and denial-of-service attack prevention
US20240106761A1 (en) Adaptive scaling of buffers for communication sessions
US20210329542A1 (en) System and methods for multi-links sd-wan using cellular device connections
US20210185130A1 (en) Systems and methods for reducing connection setup latency
CN111538932B (zh) 一种跨平台、终端及内外网的外事管理方法
CN112702415B (zh) 一种将Kafka长连接消费转换成服务的方法
Gourko WebSocket Communication between Multiple Users in Scalable Web-application Environment
US10303516B1 (en) Allocating computing resources for providing various datasets to client devices
CN121330139A (zh) 一种基于中间件的云渲染方法及中间件系统
CN120066750A (zh) 基于边缘云的安卓容器动态协同调度方法、系统及设备
HK40080386A (zh) 数据传输方法、装置、设备及存储介质
HK40080386B (zh) 数据传输方法、装置、设备及存储介质
CN118612216A (zh) 一种云边融合方法、装置、电子设备及介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22888983

Country of ref document: EP

Kind code of ref document: A1

WWE Wipo information: entry into national phase

Ref document number: 2022888983

Country of ref document: EP

ENP Entry into the national phase

Ref document number: 2022888983

Country of ref document: EP

Effective date: 20240228

NENP Non-entry into the national phase

Ref country code: DE

WWG Wipo information: grant in national office

Ref document number: 2022888983

Country of ref document: EP