Business processing methods, devices, computer equipment and storage media

By adopting a scaling-up approach in the cloud-based business cluster system, the business platform is simultaneously scaled down and expanded, solving the problem of insufficient timeliness in resource management in cloud technology and enabling rapid response to sudden business events and efficient resource utilization.

CN123074933APending Publication Date: 2026-11-10TENCENT TECHNOLOGY (SHENZHEN) CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202510585193.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-05-07
Publication Date
2026-11-10

AI Technical Summary

Technical Problem

In cloud-based business cluster systems, existing technologies suffer from insufficient timeliness in effectively managing the resource demands of multiple business platforms to ensure their normal operation, especially in responding quickly to sudden business disruptions and incremental resource needs.

Method used

The approach of scaling down and expanding simultaneously is adopted to perform synchronous scaling down and expansion on the access business platform. By rapidly releasing resources during the scaling down process, the platform can be expanded to handle sudden business surges. The container orchestration algorithm is used to select the scaling down platform and release resources immediately during the scaling down process. Combined with process signal interception and delayed reclamation of target containers, the rapid scheduling and migration of resources can be achieved.

Benefits of technology

It enabled timely response to unexpected business events, improved resource utilization and the timeliness of handling unexpected business events, and ensured the normal operation of the business.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123074933A_ABST
    Figure CN123074933A_ABST
Patent Text Reader

Abstract

This invention discloses a business processing method, apparatus, computer equipment, and storage medium. The method is applied to a business cluster system, where at least one business platform is connected to the system, and the business of each business platform runs in the system as a business container. The method includes: when a first business platform connected to the business cluster system generates a sudden surge in business, selecting a second business platform from the at least one business platform connected to the system; scaling down the second business platform, and simultaneously scaling up the first business platform using the resources released during the scaling down process; running the sudden surge in business on the expanded first business platform ensures the normal operation of the corresponding business of the business platform.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This application relates to the field of computer technology, and in particular to a business processing method, a business processing apparatus, a computer device, and a storage medium. Background Technology

[0002] With the rise of cloud technology, relying on cloud technology for business processing can effectively improve business processing efficiency. Currently, different business platforms can connect to cloud-based business cluster systems and use the resources of these systems to process their business. Since the number of business platforms connected to a business cluster system is large, and the resource requirements for processing the business of different platforms vary, ensuring the resource requirements of each business platform connected to the cluster system to guarantee the normal operation of their respective services has become a current research hotspot. Summary of the Invention

[0003] This invention provides a business processing method, apparatus, computer equipment, and storage medium, which can ensure the normal operation of the business platform corresponding to the business.

[0004] On one hand, embodiments of the present invention provide a business processing method, which is applied to a business cluster system, wherein at least one business platform is connected to the business cluster system, and the business of each business platform runs in the business cluster system in the form of a business container; the method includes:

[0005] When a sudden service occurs on the first service platform connected to the service cluster system, a second service platform is selected from at least one service platform connected to the service cluster system.

[0006] The second business platform is scaled down, and during the scaling down process, the resources released by the scaling down process are used to simultaneously scale up the first business platform.

[0007] The first business platform, after capacity expansion, will run the sudden business.

[0008] In another aspect, embodiments of the present invention provide a service processing apparatus that runs on a service cluster system, wherein at least one service platform is connected to the service cluster system, and the services of each service platform run in the service cluster system in the form of service containers; the apparatus includes:

[0009] The selection unit is used to select a second business platform from at least one business platform connected to the business cluster system when a sudden business occurs on the first business platform connected to the business cluster system.

[0010] The processing unit is used to scale down the second business platform and, during the scaling down process, use the resources released by the scaling down process to simultaneously scale up the first business platform.

[0011] The processing unit is also used to run sudden business operations on the first business platform after the expansion process.

[0012] In one embodiment, the processing unit is specifically used for:

[0013] Identify the target business containers to be reclaimed in the second business platform;

[0014] Add tag information to the node where the target business container is located. The tag information is used to declare that the resources released in the node are reserved for the first business platform.

[0015] Delete the management metadata corresponding to the target business container; and,

[0016] The target business container is delayed by intercepting process signals.

[0017] In one embodiment, the processing unit is specifically used for:

[0018] Set the running status of the target business container to a waiting state;

[0019] Disable the virtual route address corresponding to the target service container and remove the target service container that is in a waiting state from the endpoint list of the service cluster system;

[0020] An interception strategy for sending process termination signals to the node where the target service container resides. This strategy is used to indicate a delay in the reclamation of the target service container.

[0021] In one embodiment, the processing unit is specifically used for:

[0022] When the target business container is detected to be in a waiting state, the management metadata corresponding to the target business container in the waiting state is deleted from the database of the business cluster system.

[0023] In one embodiment, the processing unit is specifically used for:

[0024] Once the management metadata corresponding to the target business container is deleted, the resources released by the scaling-down process are obtained from the node where the target business container is located.

[0025] Utilize the released resources to create new business containers on the first business platform for running sudden business operations;

[0026] Based on the node's tag information, the released resources are scheduled to the new business container in the first business platform.

[0027] In one embodiment, the released resources include CPU resources unbound from the target business container; the processing unit is specifically used for:

[0028] The new business containers in the first business platform are bound to CPU resources.

[0029] In one embodiment, the released resources also include bandwidth resources occupied by the target service container; the processing unit is specifically used for:

[0030] Schedule new service containers in the first service platform to reuse bandwidth resources.

[0031] In one embodiment, the processing unit is further specifically used for:

[0032] Once a new service container for running sudden business operations is successfully created, the marking information for the node where the target service container is located is removed.

[0033] In one embodiment, the processing unit is further specifically used for:

[0034] During the scaling-down process, monitor the traffic changes of the target service container;

[0035] The capacity of new service containers is dynamically adjusted based on traffic changes.

[0036] When the remaining traffic of the target service container is less than the preset traffic threshold, the target service container is reclaimed and the scaling-down process ends.

[0037] In one embodiment, the processing unit is specifically used for:

[0038] When the remaining traffic of the target service container is less than the preset traffic threshold, an interception and release policy is sent to the node where the target service container is located. The interception and release policy of the process termination signal is used to indicate the reclamation of the target service container.

[0039] Based on the termination signal of the allowed process, reclaim the target business container.

[0040] In one embodiment, the selected unit is specifically used for:

[0041] Receive resource requests from the primary business platform; these resource requests are used to request runtime resources for sudden business events.

[0042] According to resource requirements, a second business platform is selected from at least one business platform connected to the business cluster system using a container orchestration algorithm. The number of second business platforms is one or more. The resources released by the second business platform after scaling down meet the resource requirements, and the remaining resources after scaling down meet the operational requirements of the second business platform.

[0043] In one embodiment, in a business cluster system, the first business platform and the second business platform each occupy different namespaces. The business containers corresponding to the business of the first business platform are deployed in the namespace exclusively occupied by the first business platform, and the business containers corresponding to the business of the second business platform are deployed in the namespace exclusively occupied by the second business platform.

[0044] The scaling down of the second business platform is performed on a per-namespace basis, while the scaling up of the first business platform is performed on a per-namespace basis.

[0045] In another aspect, embodiments of the present invention provide a computer device, including a processor, an input device, an output device, and a memory, wherein the processor, input device, output device, and memory are interconnected, wherein the memory is used to store a computer program that supports the computer device in executing the above-described method, the computer program including program instructions, and the processor is configured to call the program instructions to execute the above-described business processing method.

[0046] In another aspect, embodiments of the present invention provide a computer-readable storage medium storing a computer program, the computer program including program instructions, which, when executed by a processor, cause the processor to perform the aforementioned business processing method.

[0047] In another aspect, embodiments of the present invention provide a computer program product, which includes a computer program and program instructions. When the program instructions are invoked by a processor, the processor executes the aforementioned business processing method.

[0048] In this embodiment, when a sudden surge in service occurs on the first service platform connected to the service cluster system, the second service platform connected to the same system can be scaled down. During this scaling down process, the resources released during the scaling down process are simultaneously used to scale up the first service platform. In other words, this embodiment employs a simultaneous scaling down and scaling up approach, with the scaling down and scaling up processes executed concurrently. This allows for the rapid allocation of resources released from the scaling down of the second service platform to the first service platform where the sudden surge in service occurs. Through this rapid allocation and migration of resources released from scaling down, sudden surges in service scenarios can be addressed promptly, ensuring the normal operation of the surge in service. Attached Figure Description

[0049] Figure 1a This is a schematic diagram of the structure of a service cluster system provided in an embodiment of this application;

[0050] Figure 1b This is a schematic diagram illustrating business platform access based on a business cluster system, provided in an embodiment of this application.

[0051] Figure 2 This is a schematic flowchart of a business processing method provided in an embodiment of this application;

[0052] Figure 3 This is a schematic flowchart of a capacity reduction process provided in an embodiment of this application;

[0053] Figure 4 This is a schematic flowchart of a capacity expansion process provided in an embodiment of this application;

[0054] Figure 5 This is a schematic diagram illustrating a capacity expansion process during a capacity reduction process, provided in an embodiment of this application.

[0055] Figure 6 This is a schematic block diagram of a business processing device provided in an embodiment of the present invention;

[0056] Figure 7 This is a schematic block diagram of a computer device provided in an embodiment of the present invention. Detailed Implementation

[0057] The technical terms used in the embodiments of this application are explained below.

[0058] I. Business Cluster System.

[0059] The business cluster system is a cloud-native container platform used for business platform management and business processing. A cloud-native container platform is a container-centric platform built on cloud-native technology stacks. Cloud-native technology is a methodology and technology system for application design and development based on cloud computing, fully leveraging the characteristics and capabilities of cloud computing to enhance the elasticity, scalability, and flexibility of applications. Unless otherwise specified, in the subsequent embodiments of this application, the business cluster system is described using a Kubernetes (K8s, an open-source container orchestration platform) cluster system as an example.

[0060] The business cluster system allows one or more business platforms to connect. A business platform refers to an application or system that provides business services. For example, if a business platform is a live streaming application, then its business service can be live streaming; if it's a game application, then its business service can be gaming; if it's an instant messaging application, then its business service can be instant messaging, and so on. After a business platform connects to the business cluster system, its services run in the business cluster system as business containers (Pods).

[0061] Please see Figure 1a This is a schematic diagram of a service cluster system provided in an embodiment of this application. Figure 1a As shown, a business cluster system can include a control plane and a physical cluster. The control plane maintains the binding relationship between the physical cluster and the data center (i.e., which data center connects to which physical cluster). The physical cluster is responsible for managing the access of equipment resources in the data center, generating Pod instances, and managing access on a data center basis, etc. In a business cluster system, a data center typically refers to the location where the infrastructure (server hardware and network equipment, etc.) of the business cluster system is stored. The data center contains equipment, including physical or virtual servers running Kubernetes nodes. These nodes can be management nodes (such as master nodes) in the control plane or worker nodes in the physical cluster.

[0062] II. Business Container (Pod).

[0063] A Pod is a collection of containers. A Pod contains one or more containers and is the basic unit of work in a business cluster system. A Pod is managed and scheduled as a whole. After a business platform is connected to the business cluster system, the business platform's services can be containerized into one or more containers, which can be encapsulated in one or more Pods.

[0064] III. Namespace.

[0065] In a business cluster system, namespaces are a mechanism used to logically separate resources within the cluster. Namespaces provide independent isolation environments for resources (such as Pods and Services) in the business cluster system, ensuring that resources under different namespaces do not interfere with each other. Each business platform connected to the business cluster system can have its own dedicated namespace, which allows the business operations of different platforms to be isolated and independent, thus facilitating the management of the business platforms.

[0066] Please see Figure 1b This is a schematic diagram illustrating a business platform access based on a business cluster system, as provided in an embodiment of this application. Figure 1b As shown, the business cluster system contains one or more namespaces (such as namespace1, namespace2). The business cluster system is connected to a first business platform and a second business platform. The first business platform will exclusively occupy namespace1, which contains at least one Pod (Pod1-Podk) used to run the business of the first business platform. The second business platform will exclusively occupy namespace2, which contains at least one Pod (Podk+1-Podn) used to run the business of the second business platform.

[0067] Scaling up and down for each business platform is typically handled on a per-platform basis, within its corresponding namespace. For example, scaling up a business platform can be understood as allocating more resources (e.g., adding Pods) to the namespace exclusively used by that platform. Scaling down a second business platform can be understood as releasing resources (e.g., reducing the number of Pods) from its exclusive namespace. When a sudden surge in service demand occurs on the first business platform in the business cluster system, the system selects a second business platform that can be scaled down, performs the scaling down process on the second platform to release resources, and then allocates these released resources to the first business platform to ensure the smooth operation of the sudden surge in service demand.

[0068] IV. HPA (HorizontalPod Autoscaler)

[0069] In edge cloud scenarios, there are thousands of data centers, with each data center containing approximately 40 to 100 devices. This demonstrates the vast amount of data center equipment resources available in edge cloud environments. After connecting to the business cluster system, each business platform can utilize the resources provided by the data centers at the edge to run its services, enabling rapid response to business requests.

[0070] Due to limited resources in each data center, the business cluster system operates on a Kubernetes-based model, providing HPA (Hyper-Planetary Availability) elastic scaling capabilities. This dynamically adjusts resource allocation across different business platforms based on their actual resource requirements. HPA automatically scales up or down the number of Pods by monitoring specified resource metrics (such as CPU or memory), ensuring high availability under high load and conserving resources under low load.

[0071] For incremental resource demands on business platforms (e.g., the first business platform requests additional resources due to sudden business needs), in scenarios where resources are saturated from the perspective of a single data center, HPA performs a "shrink-then-expand" action between different business platforms. That is, it first selects the second business platform that can be shrunk for shrinking: after waiting for the traffic of the old Pod in the second business platform to drop to a minimum (traffic below a preset traffic threshold) and reclaiming the old Pod, it clears the management metadata of the old Pod to free up resources and ends the shrinking process; then it uses the freed-up resources to expand the first business platform. Practice has revealed that HPA has limitations in terms of timeliness. The timeliness of scaling down is strongly correlated with the business model, making it uncontrollable. For example, old Pods need to wait for traffic to drop before they can exit the business and be recycled. In the business model of CDN traffic-intensive long-connection services, the time for old Pods to exit the business is relatively long, resulting in a long recycling time for old Pods. The scaling down process is too time-consuming, and the scaling up process can only be executed after the scaling down process is completed. The timeliness of scaling up is limited by the timeliness of scaling down, thus making it unable to cope with incremental resource demands with high time requirements. For example, it cannot adjust resources at the minute level to cope with burst traffic at the terabyte (TB) level.

[0072] This application proposes a business processing method applied to a business cluster system. When a sudden surge in business activity occurs on a first business platform connected to the system, such as a sudden increase in the volume of requests from business clients (e.g., software, mini-programs, web pages, etc.) to the first business platform, causing a sudden surge in business activity on the first business platform, the method can scale down a second business platform connected to the system. During this scaling down process, the resources released during the scaling down process are used to simultaneously scale up the first business platform. In other words, this application abandons the "scaling down first, then scaling up" approach and adopts a "scaling down and scaling up simultaneously" approach, with scaling down and scaling up processes executed concurrently. This allows for the rapid allocation of resources released during the scaling down of the second business platform to the first business platform where the sudden surge in business activity occurs. Through the rapid allocation and migration of resources released during scaling down, the method can respond promptly to sudden business scenarios and ensure the normal operation of sudden business activities.

[0073] It should be specifically noted that the relevant data involved in the business processing of this application (such as relevant data when the business platform accesses the business cluster system, business-related data in the business platform, or relevant data involved when scaling down or expanding the business platform, etc.) requires the permission or consent of the target entity when the above embodiments of this application are applied to specific products or technologies. Furthermore, the collection, use, and processing of relevant data must comply with the relevant laws, regulations, and standards of the region, conforming to the principles of legality, legitimacy, and necessity, and not involving the acquisition of data types prohibited or restricted by laws and regulations. In some optional embodiments, the relevant data involved in the embodiments of this application is obtained after separate authorization from the target entity. Additionally, when obtaining separate authorization from the target entity, the purpose of the relevant data is explained to the target entity.

[0074] Please see Figure 2 This is a schematic flowchart illustrating a business processing method proposed in an embodiment of this application. This method is applied to a business cluster system and can be executed by a computer device (such as a master node) used for management and control within the business cluster system. Figure 2 As shown, the method may include:

[0075] S201, when a sudden service occurs on the first service platform connected to the service cluster system, a second service platform is selected from at least one service platform connected to the service cluster system.

[0076] Sudden surges in traffic, also known as sudden traffic spikes, refer to a rapid increase in business requests or traffic within a short period (less than a preset time threshold) due to specific reasons (such as events, promotions, or trending news). When a sudden surge in traffic occurs on the first business platform connected to the business cluster system, it indicates that the first business platform has an incremental resource requirement. The first business platform can then submit this incremental resource requirement to the business cluster system, requesting the allocation of incremental resources to handle the surge. Generally, the business cluster system reserves a certain amount of buffer space for each connected business platform. In one implementation, in response to the incremental resource requirement of the first business platform, the business cluster system can allocate the reserved buffer space resources to the first business platform. In another implementation, in response to the incremental resource requirement of the first business platform, the business cluster system can leverage the scalability of HPA to release resources by scaling down other business platforms (such as a second business platform) and then allocate the released resources to the first business platform.

[0077] In one embodiment, based on the incremental resource requirements of the first service platform, a container orchestration algorithm can be used to select one or more second service platforms from the service cluster system that can undergo scaling down. The container orchestration algorithm is a series of strategies and techniques used in container management and scheduling to automate and optimize container deployment, scaling, management, and network configuration. This container orchestration algorithm may include, but is not limited to, First-Fit, Best-Fit, Random-Fit, load balancing algorithms, scaling and scaling algorithms, etc. This application does not limit the specific algorithm. It is understood that the resources released by the second service platform after scaling down are scheduled to the first service platform, while the remaining resources of the second service platform itself are sufficient to ensure the normal operation of its own services. In other words, HPA ensures that the resources released by the second service platform after scaling down can be scheduled to the first service platform to run bursty services normally, while also ensuring that the remaining resources of the second service platform after scaling down are sufficient to run its own services normally.

[0078] S202, scale down the second business platform.

[0079] S203, during the scaling down process, the resources released during the scaling down process are used to simultaneously expand the capacity of the first business platform.

[0080] The process of scaling down the second business platform can be understood as waiting for the traffic of the target business container (referred to as the old Pod) in the second business platform to reach its limit (remaining traffic is less than a preset traffic threshold), then reclaiming the target business container and freeing up the resources it occupies. The timeliness of this process depends on the business type running by the target business container. To promptly respond to sudden business surges on the first business platform, in this embodiment, once resources are released during the scaling down process, these released resources are immediately scheduled to the first business platform for simultaneous scaling up. In this way, the scaling up process can be performed synchronously with the scaling down process, and the timeliness of responding to sudden business surges is no longer limited by the business type, thereby improving the business cluster system's ability to handle sudden business surges and increasing resource utilization.

[0081] S204, running sudden business on the first business platform after capacity expansion.

[0082] After the expansion process, the first business platform can add new business containers (called new Pods) to run sudden business operations.

[0083] In this embodiment, when a sudden surge in service occurs on the first service platform connected to the service cluster system, the second service platform connected to the same system can be scaled down. During this scaling down process, the resources released during the scaling down process are simultaneously used to scale up the first service platform. In other words, this embodiment employs a "scaling down while scaling up" approach, with scaling down and scaling up occurring concurrently. This allows for the rapid allocation of resources released from the scaling down of the second service platform to the first service platform where the sudden surge in service occurs. Through this rapid allocation and migration of resources released from scaling down, sudden surges in service can be addressed promptly, ensuring the normal operation of the surge in service.

[0084] The following is combined Figure 3 This application provides a detailed description of the process for scaling down the second service platform connected to the service cluster system, as described in the embodiments of this application. Figure 3 As shown, the process of this volume reduction procedure may include:

[0085] S301, Identify the target business container to be reclaimed in the second business platform.

[0086] The second service platform is one or more service platforms selected from the service platforms connected to the service cluster system that can be used for scaling down. This application does not limit the selection method of the second service platform. For example, one or more service platforms can be randomly selected as the second service platform, or service platforms with service traffic below the traffic threshold can be selected as the second service platform, or a container orchestration algorithm can be used to select the second service platform from the service platforms connected to the service cluster system.

[0087] In one embodiment, when selecting a second business platform from among the business platforms connected to the business cluster system using a container orchestration algorithm, the business cluster system may receive resource requests from the first business platform. These resource requests are for runtime resources for sudden service requests. Based on these resource requests, the container orchestration algorithm is used to select a second business platform from at least one business platform connected to the business cluster system. The number of second business platforms selected based on the container orchestration algorithm can be one or more. The resources released by scaling down the second business platform satisfy the resource requests of the first business platform, and the remaining resources after scaling down satisfy the operational requirements of the second business platform.

[0088] The release of resources satisfying the resource requirements of the first business platform means that, after being allocated to the first business platform, the released resources can ensure the normal operation of the first business platform's services (including sudden surges in service). The remaining resources satisfying the operational requirements of the second business platform means that the remaining resources can ensure the normal operation of the second business platform's services. The second business platform undergoing scaling down, as determined according to the embodiments of this application, is a business platform that can simultaneously meet the resource requirements of sudden surges in service and effectively guarantee the operational requirements of the second business platform itself.

[0089] After selecting a second business platform from the business cluster system, target business containers to be reclaimed can be determined from that second business platform. This application embodiment does not limit the method of selecting target business containers to be reclaimed from the second business platform. For example, any one or more business containers in the second business platform can be used as target business containers to be reclaimed; or, the business container with the lowest corresponding business traffic in the second business platform can be used as the target business container to be reclaimed; or, the business container of the second business platform with an earlier creation time (earlier than a preset time threshold) can be used as the target business container to be reclaimed; or, one or more specified business containers (e.g., business containers used to run specified services) can be used as business containers to be reclaimed, and so on.

[0090] S302, add tag information to the node where the target business container is located. The tag information is used to declare that the resources released in the node will be reserved for the first business platform.

[0091] By tagging the nodes where the target service container resides, marking information is added. For example, in an edge cloud scenario, the nodes where the target service container resides include the equipment in the data center running the target service container. The marking information is used to declare that the resources released in the node are reserved for the first service platform. This is equivalent to locking resources for the first service platform in advance. By adding marking information to the nodes where the target service container resides, resources can be reserved for the first service platform, preventing other service platforms from preempting the reserved resources. This allows the service cluster system to allocate the reserved resources to the first service platform in a timely manner according to the marking information to cope with sudden business surges.

[0092] S303, delete the management metadata corresponding to the target business container; and delay the reclamation of the target business container by intercepting process signals.

[0093] Management metadata refers to the relevant data (i.e., pod data) of the target business container recorded in the control plane database (ETCD). After a business container is created in the business cluster system, its pod data must be written to the ETCD database. Each business container in the business cluster system depends on the pod data recorded in the ETCD database to run. By deleting the management metadata of the target business container, the management metadata of the target business container is removed from the Kubernetes scheduling system, thus preventing new business requests from being scheduled to that target business container. In this embodiment, when the scaling down of the second business platform is initiated, the business cluster system can call the forced deletion interface in the business cluster system to delete the management metadata of the target business container from the ETCD, thereby immediately releasing resources.

[0094] In one implementation, when initiating the scaling-down process for the second service platform, the target service container's running state can be set to a terminating state. A terminating state refers to a state awaiting service termination (or a state awaiting recycling). Once the target service container enters the terminating state, the service cluster system can delete the management metadata of the terminating target service container from the service cluster system's database (ETCD). After the management metadata of the target service container is deleted, the target service container will only process currently requested service traffic, and no new service traffic (or new service requests) will be scheduled to the target service container.

[0095] As can be seen, once the target business container enters the waiting state, the business traffic of the target business container will show a downward trend. As the business traffic of the target business container continues to decrease, the second business platform will continuously release resources. The business cluster system can immediately schedule the released resources to the first business platform, thereby realizing the scaling up of the first business platform while scaling down the second business platform, achieving the effect of "scaling up while scaling down".

[0096] In one embodiment, according to the HPA mechanism, once the management metadata of the target service container is deleted, the control plane will trigger a rapid reclamation of the target service container. Due to the influence of the business type running the target service container, the reclamation of the target service container may take too long, thereby affecting resource scheduling and the expansion processing flow of the first service platform. Based on this, the embodiments of this application improve the HPA mechanism by using process signal interception at the kernel level to delay the reclamation process of the target service container; by delaying the reclamation of the target service container, the time spent in the reclamation process is avoided from affecting resource scheduling.

[0097] In one implementation, delaying the reclamation of a target service container by intercepting process signals can be achieved by performing the following operations: ① setting the running state of the target service container to a terminating state; ② disabling the virtual IP address (VIP) corresponding to the target service container and removing the target service container in the terminating state from the endpoints list in the service cluster system; ③ sending a process termination signal interception policy to the node where the target service container resides, the process termination signal interception policy being used to instruct the delayed reclamation of the target service container. For example, the process termination signal can specifically be the SIGKILL signal.

[0098] The Endpoints list records running business container pods. A service load balances the addresses recorded in Endpoints to forward business requests to healthy business container pods for execution. Removing target business containers from the endpoint list disables their real service (Real server, RS). Furthermore, disabling the corresponding Virtual Router Address (VIP) prevents incremental business requests from the second business platform from being routed to these waiting target business containers, thus maintaining a downward trend in business traffic to the target business containers.

[0099] As the service traffic of the target service container decreases, when the service traffic of the target service container drops below a preset traffic threshold, the target service container can be recycled. In one implementation, the process of recycling the target service container may include: when the remaining traffic of the target service container is less than the preset traffic threshold, a process termination signal interception and release policy can be sent to the node where the target service container is located. This process termination signal interception and release policy is used to instruct the recycling of the target service container; the service cluster system recycles the target service container (i.e., destroys the target service container) according to the released process termination signal, thereby completing the scaling down of the second service platform.

[0100] In this embodiment, when scaling down the second business platform in the business cluster system, by adding marking information to the node containing the target business container to be recycled in the second business platform, the resources released by that node can be reserved for the first business platform. The business cluster system can then quickly allocate the released resources to the first business platform based on the marking information to handle sudden business surges, improving response efficiency. When initiating the scaling down process, the management metadata of the target business container is immediately deleted without waiting for the target business container's business traffic to decrease, thus quickly releasing resources and enabling timely allocation to the first business platform. This reduces the startup time of the scaling up process and improves the efficiency of handling sudden business surges. Furthermore, after deleting the management metadata of the target business container, the recycling of the target business container can be delayed through process signal interception, which avoids the time consumed during the recycling process affecting resource scheduling.

[0101] The following is combined Figure 4 This application provides a detailed description of the process in which the first business platform connected to the business cluster system is simultaneously expanded during the scaling-down process in the embodiments of this application. Figure 4 As shown, the process of simultaneously scaling up the first business platform during the scaling-down process may include:

[0102] S401, after the management metadata corresponding to the target business container is deleted, obtain the resources released by the scaling-down process from the node where the target business container is located.

[0103] S402 utilizes the released resources to create a new business container on the first business platform for running bursty business operations.

[0104] Once the management metadata of the target service container is deleted, the node where the target service container resides begins to release resources. The computer devices used for management in the service cluster system (such as the master node) can immediately detect the released resources. Since the node has been marked with tagging information indicating that the released resources are reserved for the first service platform, the service cluster system will schedule the released resources to the first service platform and use the released resources to create a new service container (denoted as a new Pod) on the first service platform to run sudden service requests.

[0105] In one implementation, the business cluster system can determine a node list based on the tagging information. The node list contains one or more nodes where the target business container is located. These nodes are all tagged, and based on the scaling-down process of the target business container, resources have been released in these nodes. The business cluster system can then notify the nodes in the node list to use the released resources to create and deploy the new business container for running bursty business.

[0106] S403, based on the declaration of the node's tag information, schedules the released resources to the new business container in the first business platform.

[0107] As the management metadata of the target business container is deleted, the target business container's VIP and RS are disabled, and the target business container undergoes delayed recycling and other scaling-down operations, the resources occupied by the target business container are gradually released. These released resources can be scheduled to new business containers one after another, thereby achieving the effect of scaling down while expanding, effectively improving the expansion speed.

[0108] In one embodiment, once a new service container for running bursty services is successfully created, the marking information of the node where the target service container is located can be revoked, thereby realizing the reclamation of the marking information and avoiding the long-term occupation of the node's resources by the first service platform.

[0109] In one embodiment, the released resources can be scheduled to a new business container running a sudden surge in business operations. This can be achieved by binding the released resources to the new business container. For example, the released resources can be bound to the new business container in a way that allows for resource overselling and reuse.

[0110] In one implementation, the released resources include CPU resources unbound from the target business container. Then, the new business container in the first business platform can be bound to the CPU resources, thereby scheduling the released resources to the new business container on the first business platform. Furthermore, the business cluster system can also bind the remaining unbound CPU resources in the nodes to the new business container in the first business platform, thereby achieving over-selling and reuse of CPU resources.

[0111] In another implementation, the released resources also include the bandwidth resources occupied by the target service container. The service cluster system can schedule the new service container in the first service platform to reuse the occupied bandwidth resources, so as to schedule the released resources to the new service container in the first service platform.

[0112] In another implementation, the released resources also include other resources (such as memory resources) occupied by the target business container. Similarly, referring to the above method of scheduling CPU resources, the business cluster system can schedule these other resources to the new business container in an over-selling and reuse manner.

[0113] The process of scaling down the second business platform is a continuous release of resources; therefore, the process of simultaneously scaling up the first business platform is also a phased scaling-up process. In one implementation, the initial capacity of the new business container created on the first business platform to run sudden business surges is generally 0. The business cluster system continuously acquires the business traffic changes of the target business container. Since the target business container no longer generates incremental business, its business traffic and CPU overhead continuously decrease. Based on this, the business cluster system can monitor the traffic changes of the target business container in the second business platform during the scaling-down process and dynamically adjust the capacity of the new business container on the first business platform according to these traffic changes, gradually and continuously increasing the capacity of the new business container until it reaches its nominal upper limit capacity, and then ending the scaling-up process. When the remaining traffic of the target business container is less than a preset traffic threshold, the target business container is reclaimed and the scaling-down process ends.

[0114] In this embodiment, during the scaling down process of the second business platform, the business cluster system uses a tagging method to lock the resources released during the scaling down process of the first business platform. These released resources are then used to simultaneously scale up the first business platform, enabling rapid resource scheduling and improving resource utilization. Furthermore, by using a tiered approach to rapidly schedule the continuously released resources from the scaling down process to the new business containers added during the scaling up of the first business platform to handle sudden business surges, resource over-sale and reuse are achieved. This not only achieves the goal of scaling down and scaling up simultaneously but also improves resource utilization.

[0115] In one embodiment, Figure 3 The process of volume reduction shown is similar to Figure 4 The scaling process shown is performed synchronously. In the business cluster system, the first business platform and the second business platform each occupy different namespaces. The business containers corresponding to the services of the first business platform are deployed in the namespace occupied by the first business platform, and the business containers corresponding to the services of the second business platform are deployed in the namespace occupied by the second business platform. Scaling down the second business platform is performed on a per-namespace basis; that is, scaling down the second business platform can mean reducing the number of Pods in the namespace occupied by the second business platform. Scaling up the first business platform is performed on a per-namespace basis; that is, scaling up the first business platform can mean increasing the number of Pods in the namespace occupied by the first business platform.

[0116] The business cluster system contains various components. Figure 3 The following are the relevant steps involved in the volume reduction process shown in the diagram, and Figure 4The relevant steps involved in the capacity expansion process can all be implemented by the various components provided by the business cluster system.

[0117] Used to implement Figure 3 Related steps and Figure 4 The main components involved in the relevant steps include: KubeGOSS components (a management component), a health center, a management interface component service (K8s host API server), a control component (GCController), a physical interface component service (K8s member API server), a workload controller, a scheduler, edge nodes (OC nodes), and the central management component IPList and the time-series data acquisition service (Prometheus). The functional descriptions of these main components are provided below.

[0118] ①KubeGOSS components.

[0119] The KubeGOSS component is a management and control component located in the control plane of the K8s cluster (i.e., the business cluster system) and is used to verify the availability of nodes and services. The scaling down of the second business platform and the scaling up of the first business platform during the scaling down process involved in this application embodiment are achieved collaboratively through the scheduling of other components / services by the KubeGOSS component.

[0120] ②Health Center.

[0121] The health center is used to view the health status of each business container pod in the business cluster system to determine whether each business container can run the business normally, and to send the virtual routing address of the business container that can run the business normally to the upper-layer scheduler so that the scheduler can direct business requests to the healthy business container.

[0122] ③Management interface component service (K8s host API server).

[0123] The management interface component service is the interface service of the upper control plane, and can synchronize its state with the physical interface component service of the lower physical cluster.

[0124] ④ Physical Interface Component Service (K8s member API server).

[0125] The physical interface component service is a service in the multi-physical cluster of this K8s cluster. Each physical cluster has a corresponding interface service (API Server). In other words, the management and control components in the control plane (such as the aforementioned KubeGOSS component) can manage the business containers in the corresponding physical cluster based on the physical interface service of the physical cluster.

[0126] ⑤ Workload controller, scheduler, and edge nodes (OC nodes).

[0127] The workload controller is used to dynamically adjust the state of a business container, which includes one or more of the following: creation (or addition), deletion, replacement, and waiting.

[0128] The scheduler is used for resource scheduling. Edge nodes are the nodes where the target service containers reside.

[0129] ⑥ Central-side control components IPList and time-series data acquisition service (Prometheus).

[0130] The central control component IPList is used to expose the address of the target service container, the address of the new service container, and the capacity of the new service container to the outside world for scheduling decisions, and provide them to the traffic scheduling system for decision-making.

[0131] The time-series data acquisition service (Prometheus) monitors traffic changes on the second business platform.

[0132] The following is combined Figure 5 The process of scaling down the second business platform and scaling up the first business platform by the components provided by the business cluster system is described. The process specifically includes the following stages (1) to (11).

[0133] (1) The KubeGOSS control plane receives a request from the first service platform to use the resources of the second service platform for expansion processing. This requires the second service platform to be scaled down and the resources of the scaled-down process to be used to expand the first service platform.

[0134] (2) The KubeGOSS component in the control plane adds tag information to the node where the target service container is located in the second service platform through the physical interface service component K8s member API server, declaring that the resources released by the node are reserved for the first service platform. This ensures that the scheduler can prioritize the scheduling of the resources released by the node to the sudden services generated by the first service platform.

[0135] The KubeGOSS component in the control plane can also intercept processes that terminate signals to edge nodes.

[0136] (3) The KubeGOSS component in the control plane disables VIP and RS by requesting the health center, so that newly generated business traffic (or incremental business) in the second business platform cannot request the target business container, thus keeping the resource usage of the target business container decreasing.

[0137] (4) After the interception policy is successfully issued, the KubeGOSS control plane can cancel the deployment of the new business container of the second business platform in this node through the management interface service component K8shost API server, thereby avoiding the occupation of resources by the new business generated by the second business platform.

[0138] (4.1) Based on the cancellation of the deployment of new business containers in the node, the control component GC controller can detect the change and issue a deletion command (i.e. delete workload command) to the corresponding physical interface service component K8s member API server. Based on the deletion command, the physical interface service component K8s member API server can adjust the expected state of the target business container recorded in the spec field to the waiting state.

[0139] (4.2) If the Workload Controller obtains a record in the spec field that has changed according to the deletion command issued to the physical interface service component K8s member API server, it can adjust the target business container to a waiting state based on the waiting state described after the spec field change.

[0140] (4.3) Edge nodes (OC nodes) can attempt to delete the target service container after the workload controller adjusts the target service container to a waiting state. However, they will be blocked based on the aforementioned interception strategy, thereby preventing the rapid recycling of the target service container and delaying the issuance of the policy.

[0141] (5) After the target business container of the control plane is adjusted to the waiting state, the KubeGOSS component can call the strong deletion interface to delete the management metadata (i.e., pod data recorded by ETCD) of the target business container, thereby realizing the rapid scaling down of the second business platform.

[0142] (6) The KubeGOSS component in the control plane controls the deployment of new business containers for sudden business events of the first business platform in the first business platform through the management and control interface service.

[0143] (7) Edge nodes can bind resources to new business containers based on the resources released by scaling down the second business platform, in accordance with the principle of resource overselling and reuse, through the physical interface component service of the physical cluster where the first business platform is located;

[0144] (8) After the new service container for the burst service is successfully created, the KubeGOSS component of the control plane removes the tag information added to the node where the target service container is located, and continuously obtains the traffic changes of the second service platform through the time series data acquisition service (Prometheus).

[0145] (9) Based on the obtained traffic changes, the KubeGOSS component in the control plane dynamically adjusts the capacity of the new service container corresponding to the sudden service through the central management component IPList until the maximum capacity is reached.

[0146] (10) When the service traffic of the second service platform drops to a low level, the KubeGOSS component of the control plane performs interception policy recovery, thereby enabling the release of the process termination signal sent by the node where the target service container is located, thereby realizing the recovery of the target service container based on the process termination signal, and stopping the over-selling and reuse of resources on the node where the target service container is located.

[0147] The business processing method of this application embodiment can be applied to various business scenarios, including but not limited to: game scenarios, interactive entertainment scenarios, access scenarios, fault interruption scenarios, cloud platform emergency scenarios, etc.

[0148] In one embodiment, in a gaming scenario, the first business platform refers to the game operating platform. When this platform releases a new game on a massive scale (up to 100TB), it triggers a surge in downloads, leading to a sudden increase in traffic. Based on the business processing method of this embodiment, the business cluster system can quickly release and migrate resources from other business platforms to the first business platform to support the download of the new game and facilitate its rapid growth. Once the download trend stabilizes, the resources of the first business platform can be quickly redistributed to other business platforms.

[0149] In another embodiment, in an interactive entertainment scenario, a sudden event occurs when the first business platform releases a new business. Based on the business processing method of this application embodiment, the business cluster system can quickly release and schedule resources from other business platforms to the first business platform to support the new business. After the event ends, the resources of the first business platform can be quickly returned to other business platforms.

[0150] In another embodiment, in an access scenario, when it is necessary to relocate a designated data center to consider access costs, the resources of the designated data center (such as the nearest edge data center for game release) can be released based on the business processing method of the embodiments of this application, so as to make full use of the 95 space.

[0151] In another embodiment, in a fault-based scenario, when the sudden traffic event of the first business platform is a specific platform (non-accelerated general) event, since the specific platform intervention requires multi-tenant overload first, the resources of other business platforms can be migrated to quickly expand the capacity of the first business platform based on the business processing method of this application embodiment.

[0152] In another embodiment, in a cloud platform emergency scenario, when the business traffic of the first business platform is interrupted due to the normal volume and insufficient resources of the EO (a type of cloud platform) traffic platform, it is difficult to achieve the same configuration for the EO traffic platform and the acceleration general platform. Therefore, it is necessary to quickly expand the EO traffic platform. The business processing method of this application embodiment can be used to achieve rapid expansion of the EO traffic platform.

[0153] Furthermore, the embodiments of this application can also be applied to other business scenarios. For example, when the sudden business of the first business platform is a sudden event of the data center reaching its 95% capacity limit, the business processing method of the embodiments of this application can be used to release the computing resources of some low-capacity, high-QPS (query count) business to make high-bandwidth capacity business reach its capacity limit, etc. This application does not limit the business scenarios.

[0154] In this embodiment, when a sudden surge in service occurs on the first service platform connected to the service cluster system, the second service platform connected to the same system can be scaled down. During this scaling down process, the resources released during the scaling down process are simultaneously used to scale up the first service platform. In other words, this embodiment employs a "scaling down while scaling up" approach, with scaling down and scaling up occurring concurrently. This allows for the rapid allocation of resources released from the scaling down of the second service platform to the first service platform where the sudden surge in service occurs. Through this rapid allocation and migration of resources released from scaling down, sudden surges in service can be addressed promptly, ensuring the normal operation of the surge in service.

[0155] Please see Figure 6 , Figure 6 This is a schematic diagram of a business processing device provided in an embodiment of this application. This business processing device runs on a business cluster system, specifically within a computer device used for management and control within the business cluster system. At least one business platform is connected to the business cluster system, and the services of each business platform run in the business cluster system in the form of a business container. For example... Figure 6As shown, the service processing device may include a selection unit 601 and a processing unit 602.

[0156] The selection unit 601 is used to select a second business platform from at least one business platform connected to the business cluster system when a sudden service occurs on the first business platform connected to the business cluster system.

[0157] The processing unit 602 is used to perform scaling down processing on the second business platform, and during the scaling down process, it uses the resources released by the scaling down process to simultaneously perform scaling up processing on the first business platform.

[0158] The processing unit 602 is also used to run sudden business on the first business platform after the expansion process.

[0159] In one embodiment, the processing unit 602 is specifically used for:

[0160] Identify the target business containers to be reclaimed in the second business platform;

[0161] Add tag information to the node where the target business container is located. The tag information is used to declare that the resources released in the node are reserved for the first business platform.

[0162] Delete the management metadata corresponding to the target business container; and,

[0163] The target business container is delayed by intercepting process signals.

[0164] In one embodiment, the processing unit 602 is specifically used for:

[0165] Set the running status of the target business container to a waiting state;

[0166] Disable the virtual route address corresponding to the target service container and remove the target service container that is in a waiting state from the endpoint list of the service cluster system;

[0167] An interception strategy for sending process termination signals to the node where the target service container resides. This strategy is used to indicate a delay in the reclamation of the target service container.

[0168] In one embodiment, the processing unit 602 is specifically used for:

[0169] When the target business container is detected to be in a waiting state, the management metadata corresponding to the target business container in the waiting state is deleted from the database of the business cluster system.

[0170] In one embodiment, the processing unit 602 is specifically used for:

[0171] Once the management metadata corresponding to the target business container is deleted, the resources released by the scaling-down process are obtained from the node where the target business container is located.

[0172] Utilize the released resources to create new business containers on the first business platform for running sudden business operations;

[0173] Based on the node's tag information, the released resources are scheduled to the new business container in the first business platform.

[0174] In one embodiment, the released resources include CPU resources unbound from the target business container; the processing unit 602 is specifically used for:

[0175] The new business containers in the first business platform are bound to CPU resources.

[0176] In one embodiment, the released resources also include bandwidth resources occupied by the target service container; the processing unit 602 is specifically used for:

[0177] Schedule new service containers in the first service platform to reuse bandwidth resources.

[0178] In one embodiment, the processing unit 602 is further specifically used for:

[0179] Once a new service container for running sudden business operations is successfully created, the marking information for the node where the target service container is located is removed.

[0180] In one embodiment, the processing unit 602 is further specifically used for:

[0181] During the scaling-down process, monitor the traffic changes of the target service container;

[0182] The capacity of new service containers is dynamically adjusted based on traffic changes.

[0183] When the remaining traffic of the target service container is less than the preset traffic threshold, the target service container is reclaimed and the scaling-down process ends.

[0184] In one embodiment, the processing unit 602 is specifically used for:

[0185] When the remaining traffic of the target service container is less than the preset traffic threshold, an interception and release policy is sent to the node where the target service container is located. The interception and release policy of the process termination signal is used to indicate the reclamation of the target service container.

[0186] Based on the termination signal of the allowed process, reclaim the target business container.

[0187] In one embodiment, unit 601 is selected and specifically used for:

[0188] Receive resource requests from the primary business platform; these resource requests are used to request runtime resources for sudden business events.

[0189] According to resource requirements, a second business platform is selected from at least one business platform connected to the business cluster system using a container orchestration algorithm. The number of second business platforms is one or more. The resources released by the second business platform after scaling down meet the resource requirements, and the remaining resources after scaling down meet the operational requirements of the second business platform.

[0190] In one embodiment, in a business cluster system, the first business platform and the second business platform each occupy different namespaces. The business containers corresponding to the business of the first business platform are deployed in the namespace exclusively occupied by the first business platform, and the business containers corresponding to the business of the second business platform are deployed in the namespace exclusively occupied by the second business platform.

[0191] The scaling down of the second business platform is performed on a per-namespace basis, while the scaling up of the first business platform is performed on a per-namespace basis.

[0192] In this embodiment, when a sudden surge in service occurs on the first service platform connected to the service cluster system, the processing unit 602 can scale down the second service platform connected to the same system. During this scaling-down process, the resources released during the scaling-down process are simultaneously used to scale up the first service platform. In other words, this embodiment employs a "scaling down while scaling up" approach, with scaling-down and scaling-up processes executed concurrently. This allows for the rapid allocation of resources released from the scaling down of the second service platform to the first service platform where the sudden surge in service occurs. Through this rapid allocation and migration of resources released from scaling down, the system can respond promptly to sudden service scenarios and ensure the normal operation of the surge in service.

[0193] According to one embodiment of this application, Figure 6 The modules in the illustrated business processing apparatus can be individually or entirely combined into one or more units, or some of these units can be further divided into multiple functionally smaller sub-units to achieve the same operation without affecting the technical effects of the embodiments of this application. The above modules are based on logical functional division; in practical applications, the function of one module can be implemented by multiple units, or the function of multiple modules can be implemented by one unit. In other embodiments of this application, the business processing apparatus may also include other units; in practical applications, these functions can also be implemented with the assistance of other units, and can be implemented collaboratively by multiple units.

[0194] In this application embodiment, the terms "module" or "unit" refer to a computer program or part of a computer program that has a predetermined function and works with other related parts to achieve a predetermined goal, and can be implemented wholly or partially using software, hardware (such as processing circuitry or memory), or a combination thereof. Similarly, a processor (or multiple processors or memory) can be used to implement one or more modules or units. Furthermore, each module or unit can be part of an overall module or unit that includes the functionality of that module or unit.

[0195] According to one embodiment of this application, a computer program capable of executing the steps involved in the corresponding methods shown in the various embodiments of this application can be run on a general-purpose computer device (which may include processing elements and storage elements such as a central processing unit (CPU), random access memory (RAM), and read-only memory (ROM)) to construct, as described in the embodiments of this application. Figure 6 The business processing apparatus shown herein. The aforementioned computer program may be recorded on a computer-readable recording medium, and may be loaded into the aforementioned computer device via the computer-readable recording medium and run therein.

[0196] Please see Figure 7 , Figure 7 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. This computer device is used for management and control within a business cluster system. At least one business platform is connected to the business cluster system, and the services of each business platform run in the business cluster system in the form of business containers. Figure 5 The components / services involved run on or are controlled by this computer device. For example... Figure 7 As shown, the computer device 1000 may include a processor 1001, a network interface 1004, and a memory 1005. In some embodiments, the computer device 1000 may also include a user interface 1003 and at least one communication bus 1002. The communication bus 1002 is used to enable communication between these components. The user interface 1003 may include a display screen and a keyboard; optionally, the user interface 1003 may also include a standard wired interface or a wireless interface. The network interface 1004 may optionally include a standard wired interface or a wireless interface (such as a Wi-Fi interface). The memory 1005 may be high-speed RAM or non-volatile memory, such as at least one disk storage device. Optionally, the memory 1005 may also be at least one storage device located remotely from the aforementioned processor 1001. Figure 7As shown, the memory 1005, which serves as a computer storage medium, may include an operating system, a network communication module, a user interface module, and a device control application program.

[0197] exist Figure 7 In the computer device 1000 shown, the network interface 1004 provides network communication functionality; the user interface 1003 is mainly used to provide an input interface for the user; and the processor 1001 can be used to call the device control application stored in the memory 1005 to achieve:

[0198] When a sudden service occurs on the first service platform connected to the service cluster system, a second service platform is selected from at least one service platform connected to the service cluster system.

[0199] The second business platform is scaled down, and during the scaling down process, the resources released by the scaling down process are used to simultaneously scale up the first business platform.

[0200] The first business platform, after capacity expansion, will run the sudden business.

[0201] In one embodiment, the processor 1001 can be used to invoke a device control application stored in the memory 1005, and also to implement:

[0202] Identify the target business containers to be reclaimed in the second business platform;

[0203] Add tag information to the node where the target business container is located. The tag information is used to declare that the resources released in the node are reserved for the first business platform.

[0204] Delete the management metadata corresponding to the target business container; and,

[0205] The target business container is delayed by intercepting process signals.

[0206] In one embodiment, the processor 1001 can be used to invoke a device control application stored in the memory 1005, and also to implement:

[0207] Set the running status of the target business container to a waiting state;

[0208] Disable the virtual route address corresponding to the target service container and remove the target service container that is in a waiting state from the endpoint list of the service cluster system;

[0209] An interception strategy for sending process termination signals to the node where the target service container resides. This strategy is used to indicate a delay in the reclamation of the target service container.

[0210] In one embodiment, the processor 1001 can be used to invoke a device control application stored in the memory 1005, and also to implement:

[0211] When the target business container is detected to be in a waiting state, the management metadata corresponding to the target business container in the waiting state is deleted from the database of the business cluster system.

[0212] In one embodiment, the processor 1001 can be used to invoke a device control application stored in the memory 1005, and also to implement:

[0213] Once the management metadata corresponding to the target business container is deleted, the resources released by the scaling-down process are obtained from the node where the target business container is located.

[0214] Utilize the released resources to create new business containers on the first business platform for running sudden business operations;

[0215] Based on the node's tag information, the released resources are scheduled to the new business container in the first business platform.

[0216] In one embodiment, the released resources include CPU resources unbound from the target service container; the processor 1001 can be used to call the device control application stored in the memory 1005, and is also used to achieve:

[0217] The new business containers in the first business platform are bound to CPU resources.

[0218] In one embodiment, the released resources also include bandwidth resources occupied by the target service container; the processor 1001 can be used to call the device control application stored in the memory 1005, and is also used to implement:

[0219] Schedule new service containers in the first service platform to reuse bandwidth resources.

[0220] In one embodiment, the processor 1001 can be used to invoke a device control application stored in the memory 1005, and also to implement:

[0221] Once a new service container for running sudden business operations is successfully created, the marking information for the node where the target service container is located is removed.

[0222] In one embodiment, the processor 1001 can be used to invoke a device control application stored in the memory 1005, and also to implement:

[0223] During the scaling-down process, monitor the traffic changes of the target service container;

[0224] The capacity of new service containers is dynamically adjusted based on traffic changes.

[0225] When the remaining traffic of the target service container is less than the preset traffic threshold, the target service container is reclaimed and the scaling-down process ends.

[0226] In one embodiment, the processor 1001 can be used to invoke a device control application stored in the memory 1005, and also to implement:

[0227] When the remaining traffic of the target service container is less than the preset traffic threshold, an interception and release policy is sent to the node where the target service container is located. The interception and release policy of the process termination signal is used to indicate the reclamation of the target service container.

[0228] Based on the termination signal of the allowed process, reclaim the target business container.

[0229] In one embodiment, the processor 1001 can be used to invoke a device control application stored in the memory 1005, and also to implement:

[0230] Receive resource requests from the primary business platform; these resource requests are used to request runtime resources for sudden business events.

[0231] According to resource requirements, a second business platform is selected from at least one business platform connected to the business cluster system using a container orchestration algorithm. The number of second business platforms is one or more. The resources released by the second business platform after scaling down meet the resource requirements, and the remaining resources after scaling down meet the operational requirements of the second business platform.

[0232] In one embodiment, in a business cluster system, the first business platform and the second business platform each occupy different namespaces. The business containers corresponding to the business of the first business platform are deployed in the namespace exclusively occupied by the first business platform, and the business containers corresponding to the business of the second business platform are deployed in the namespace exclusively occupied by the second business platform.

[0233] The scaling down of the second business platform is performed on a per-namespace basis, while the scaling up of the first business platform is performed on a per-namespace basis.

[0234] It should be understood that the computer device 1000 described in the embodiments of this application can execute the business processing methods described in the embodiments of this application, and can also execute the methods described above. Figure 6 The description of the service processing apparatus in the corresponding embodiments will not be repeated here. Furthermore, the beneficial effects of using the same method will also not be repeated.

[0235] Furthermore, it should be noted that this application also provides a computer-readable storage medium storing a computer program. When a processor executes the computer program, it can perform the business processing methods described in the various embodiments of this application; therefore, these will not be repeated here. Additionally, the beneficial effects of using the same method will also not be repeated. For technical details not disclosed in the embodiments of the computer storage medium involved in this application, please refer to the description of the method embodiments of this application.

[0236] As an example, the aforementioned computer program can be deployed and executed on a single computer device, or deployed and executed on multiple computer devices located in one location, or executed on multiple computer devices distributed across multiple locations and interconnected via a communication network. These multiple computer devices distributed across multiple locations and interconnected via a communication network can form a blockchain network.

[0237] The aforementioned computer-readable storage medium can be an internal storage unit of the computer device, such as a hard drive or memory. It can also be an external storage device, such as a plug-in hard drive, smart media card (SMC), secure digital card (SD) card, or flash card. Furthermore, the computer-readable storage medium can include both internal and external storage units of the computer device. This computer-readable storage medium is used to store the computer program and other programs and data required by the computer device. It can also be used to temporarily store data that has been output or will be output.

[0238] This application provides a computer program product comprising a computer program stored in a computer-readable storage medium. A processor of a computer device reads the computer program from the computer-readable storage medium and executes the computer program, causing the computer device to perform the business processing methods described in the embodiments of this application; therefore, these will not be repeated here. Furthermore, the beneficial effects of using the same method will also not be repeated. For technical details not disclosed in the embodiments of the computer-readable storage medium involved in this application, please refer to the description of the method embodiments of this application.

[0239] The terms "first," "second," etc., in the specification, claims, and drawings of this application are used to distinguish different objects, not to describe a specific order. Furthermore, the term "comprising," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, apparatus, product, or device that includes a series of steps or units is not limited to the listed steps or modules, but may optionally include steps or modules not listed, or may optionally include other step units inherent to these processes, methods, apparatuses, products, or devices.

[0240] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this application.

[0241] The above-disclosed embodiments are merely preferred embodiments of this application and should not be construed as limiting the scope of this application. Therefore, any equivalent variations made in accordance with the claims of this application shall still fall within the scope of this application.

Claims

1. A business processing method, characterized in that, The method is applied to a business cluster system, wherein at least one business platform is connected to the business cluster system, and the business of each business platform runs in the business cluster system in the form of a business container; the method includes: When a sudden service occurs on the first service platform connected to the service cluster system, a second service platform is selected from at least one service platform connected to the service cluster system. The second business platform is scaled down, and during the scaling down process, the resources released by the scaling down process are used to simultaneously scale up the first business platform. The first service platform, after the capacity expansion process, runs the burst service.

2. The method as described in claim 1, characterized in that, The scaling down of the second business platform includes: Identify the target business containers to be reclaimed in the second business platform; Add tag information to the node where the target service container is located. The tag information is used to declare that the resources released in the node are reserved for the first service platform. Delete the management metadata corresponding to the target service container; and, The target service container is delayed by intercepting process signals.

3. The method as described in claim 2, characterized in that, The method of delaying the reclamation of the target service container by intercepting process signals includes: Set the running state of the target business container to a waiting state; Disable the virtual routing address corresponding to the target service container and remove the target service container in a waiting state from the endpoint list of the service cluster system; An interception strategy for sending a process termination signal to the node where the target service container is located, wherein the process termination signal interception strategy is used to indicate a delay in the reclamation of the target service container.

4. The method as described in claim 2, characterized in that, The deletion of the management metadata corresponding to the target service container includes: When the target service container is detected to be in a waiting state, the management metadata corresponding to the target service container in the waiting state is deleted from the database of the service cluster system.

5. The method as described in claim 2, characterized in that, During the scaling-down process, the resources released by the scaling-down process are used to simultaneously scale up the first service platform, including: After the management metadata corresponding to the target service container is deleted, the resources released by the scaling-down process are obtained from the node where the target service container is located. Using the released resources, a new business container is created on the first business platform to run the sudden business. Based on the declaration of the node's tag information, the released resources are scheduled to the new service container in the first service platform.

6. The method as described in claim 5, characterized in that, The released resources include the CPU resources unbound from the target service container; the step of allocating the released resources to a new service container in the first service platform includes: The new business container in the first business platform is bound to the CPU resources.

7. The method as described in claim 6, characterized in that, The released resources also include the bandwidth resources occupied by the target service container; the step of allocating the released resources to the new service container in the first service platform further includes: The bandwidth resources are reused by scheduling new service containers in the first service platform.

8. The method as described in claim 5, characterized in that, The process of scaling down, utilizing the resources released during the scaling down process to simultaneously scale up the first service platform, further includes: Once the new service container used to run the burst service is successfully created, the marking information is revoked for the node where the target service container is located.

9. The method as described in claim 5, characterized in that, The process of scaling down, utilizing the resources released during the scaling down process to simultaneously scale up the first service platform, further includes: During the scaling-down process, the traffic changes of the target service container are monitored; The capacity of the new service container is dynamically adjusted based on the changes in traffic. When the remaining traffic of the target service container is less than the preset traffic threshold, the target service container is recycled and the scaling-down process ends.

10. The method as described in claim 9, characterized in that, The step of reclaiming the target service container when the remaining traffic of the target service container is less than a preset traffic threshold includes: When the remaining traffic of the target service container is less than a preset traffic threshold, an interception and release policy is sent to the node where the target service container is located. The interception and release policy of the process termination signal is used to indicate the reclamation of the target service container. Based on the terminate signal of the allowed process, the target service container is reclaimed.

11. The method as described in claim 1, characterized in that, When a sudden service occurs on a first service platform connected to the service cluster system, selecting a second service platform from at least one service platform connected to the service cluster system includes: Receive resource requests from the first service platform, wherein the resource requests are used to request the operating resources for the sudden service. According to the resource requirements, a second business platform is selected from at least one business platform connected to the business cluster system using a container orchestration algorithm. The number of the second business platforms is one or more. The resources released by the second business platform after scaling down meet the resource requirements, and the remaining resources after scaling down meet the operational requirements of the second business platform.

12. The method as described in claim 1, characterized in that, In the business cluster system, the first business platform and the second business platform each occupy different namespaces. The business containers corresponding to the business of the first business platform are deployed in the namespace exclusively occupied by the first business platform, and the business containers corresponding to the business of the second business platform are deployed in the namespace exclusively occupied by the second business platform. The scaling down process for the second business platform is performed on a per-namespace basis, and the scaling up process for the first business platform is performed on a per-namespace basis.

13. A business processing apparatus, wherein the business processing apparatus operates in a business cluster system, at least one business platform is connected to the business cluster system, and the business of each business platform runs in the business cluster system in the form of a business container; The device includes: The selection unit is used to select a second service platform from at least one service platform connected to the service cluster system when a sudden service occurs on the first service platform connected to the service cluster system. The processing unit is used to perform scaling down processing on the second business platform, and during the scaling down process, it uses the resources released by the scaling down process to simultaneously perform scaling up processing on the first business platform. The processing unit is also used to run the burst service on the first service platform after the expansion process.

14. A computer device, characterized in that, The system includes a processor, an input device, an output device, and a memory, wherein the processor, the input device, the output device, and the memory are interconnected, wherein the memory is used to store a computer program, the computer program including program instructions, and the processor is configured to invoke the program instructions to execute the method as described in any one of claims 1 to 12.

15. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, the computer program including program instructions that, when executed by a processor, cause the processor to perform the method as described in any one of claims 1 to 12.

16. A computer program product, characterized in that, The computer program product includes a computer program, which includes program instructions that, when invoked by a processor, cause the processor to perform the method as described in any one of claims 1 to 12.