Hybrid department business processing method and device, equipment, storage medium and program product

Abstract

The invention discloses a hybrid business processing method, a hybrid business processing device, hybrid business processing equipment, a storage medium and a computer program product, and relates to the technical field of computers, and the method comprises the following steps: according to the influence degree of a business system on a company and whether the business system is a key node on a micro-service link, determining whether the business system is a key node on the micro-service link; dividing importance levels of the service systems to obtain service systems with different continuity guarantee levels; for each service system, dividing to obtain different levels of services of each service system according to the resource consumption degree and the time delay sensitivity degree of the services in each service system; and deploying the businesses of different levels into a hybrid container cluster. According to the application, the utilization rate of the container instance resources is improved while the service continuity is guaranteed by defining the division standard of the hybrid services.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a hybrid service processing method, a hybrid service processing apparatus, a hybrid service processing device, a storage medium, and a computer program product. Background Art

[0002] Currently, there are two operating resource allocation modes for colocation services (online and offline services) - exclusive kernel mode and shared kernel mode. In shared kernel mode, colocation services are deployed on the same physical machine. Although this alleviates the problem of operating resource fragmentation, the standard for dividing colocation services is unclear, making it difficult to select the priority that best suits the actual situation. The system tends to select a type with a higher scheduling priority, resulting in low resource utilization of container instances. Summary of the invention

[0003] The main purpose of the present application is to provide a hybrid business processing method, a hybrid business processing apparatus, a hybrid business processing device, a storage medium and a computer program product, aiming to solve the technical problem of low resource utilization of container instances.

[0004] To achieve the above objectives, the present application proposes a hybrid service processing method, the method comprising:

[0005] According to the impact of the business system on the company and whether the business system is a key node in the microservice link, the importance level of the business system is divided to obtain business systems with different continuity assurance levels;

[0006] For each of the business systems, different levels of services of each of the business systems are obtained according to the resource consumption degree and delay sensitivity degree of the services in each of the business systems;

[0007] Deploy the different levels of services into the colocation container cluster.

[0008] In one embodiment, different continuity assurance levels include a core business system of level A, an important business system of level B, a general business system of level C, and business systems other than level A, level B, and level C. For each of the business systems, the steps of obtaining different levels of business of each of the business systems according to the resource consumption degree and delay sensitivity degree of the business in each of the business systems include:

[0009] If the resource consumption reaches a first preset threshold and the delay sensitivity does not reach a second preset threshold, the service is classified as a level 4 offline type service;

[0010] If the resource consumption level does not reach the first preset threshold and the delay sensitivity level reaches the second preset threshold, the service is classified as a level 3 online type service;

[0011] When the continuity guarantee level is A or B, if the resource consumption level does not reach the first preset threshold and the delay sensitivity level does not reach the third preset threshold, the service is classified as a level 2 online type service;

[0012] When the continuity guarantee level is A or B, if the resource consumption level does not reach the first preset threshold and the delay sensitivity level reaches the third preset threshold, the service is classified as a level 1 online type service.

[0013] In one embodiment, at the platform layer, the level of the divided services of each of the service systems includes non-low priority and low priority, and after the step of deploying the services of different levels into the colocation container cluster, the following steps are included:

[0014] Treat the non-low-priority service as an online type service, and treat the low-priority service as an offline type service;

[0015] According to the requirements of the online type services and the offline type services for container instance resources, the container instance resources are allocated to each container in the colocation container cluster where each of the online type services and the offline type services are located;

[0016] When the resource consumption of the container instance of the online type business on the node of each container reaches a preset water level, the container instance resources of the offline type business running on the node are scheduled and suppressed, and / or the container instance of the offline type business running on the node is expelled.

[0017] In one embodiment, the non-low priority includes level 1, level 2, and level 3, and the step of allocating container instance resources to each container in the colocation container cluster where each of the online type services and the offline type services are located according to the requirements of each of the online type services and the offline type services for container instance resources includes:

[0018] The container of the level 1 online type service exclusively occupies the physical core and all logical cores in the physical core;

[0019] The container of the level 2 online type service exclusively occupies the physical core and some logical cores in the physical core, and the remaining logical cores carry the level 4 offline type service;

[0020] The containers of the level 3 online type service and the level 4 offline type service share the remaining physical cores.

[0021] In one embodiment, when the resource consumption of the container instance of the online type service on the node of each container reaches a preset water level, the step of scheduling and suppressing the container instance resources of the node running the offline type service, and / or the step of evicting the container instance of the node running the offline type service includes:

[0022] When the resource consumption of container instances reaches the prohibition level for scheduling, scheduling of container instance resources for offline services is prohibited.

[0023] When the resource consumption of the container instance reaches the resource suppression waterline, a bandwidth threshold is set for the offline type service. If the bandwidth of the offline type service is greater than the bandwidth threshold, packet loss is performed on the offline type service.

[0024] When the resource consumption of the container instance reaches the eviction watermark, the offline type business is evicted, allowing the containers of the level 1 and level 2 online type business to occupy the remaining container instance resources.

[0025] In one embodiment, at the infrastructure layer, different levels of services include online type services and offline type services. After the step of deploying the different levels of services into the colocation container cluster, the step further includes:

[0026] Combining multiple single-availability zone colocation container subclusters into a logical federated cluster, wherein the single availability zone is a subcluster of the colocation container;

[0027] By backing up the online type services in the two sub-clusters of the same logical federated cluster, when one of the sub-clusters is abnormal, the services of the other cluster are used to provide external services.

[0028] In addition, to achieve the above-mentioned purpose, the present application also proposes a hybrid business processing device, which includes: a first division module, which is used to divide the importance level of the business system according to the impact of the business system on the company and whether the business system is a key node on the microservice link, so as to obtain business systems with different continuity assurance levels;

[0029] The second division module is used to divide the services of each service system into different levels according to the resource consumption degree and delay sensitivity degree of the services in each service system;

[0030] The deployment module is used to deploy the services of different levels into the colocation container cluster.

[0031] In addition, to achieve the above-mentioned purpose, the present application also proposes a hybrid service processing device, the device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the hybrid service processing method as described above.

[0032] In addition, to achieve the above-mentioned purpose, the present application also proposes a storage medium, which is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by the processor, the steps of the hybrid service processing method described above are implemented.

[0033] In addition, to achieve the above-mentioned purpose, the present application also provides a computer program product, wherein the computer program product includes a computer program, and when the computer program is executed by a processor, the steps of the hybrid service processing method described above are implemented.

[0034] The hybrid business processing method of the embodiment of the present application divides business systems and different levels of business into different continuity guarantee levels according to the impact of the business system on the company, the criticality of the business on the microservice link, and the resource consumption and latency sensitivity of the business. In addition, after the hybrid business is divided, the different levels of business are deployed to the hybrid container cluster. Compared with the conventional means, the division criteria of the shared kernel mode hybrid business are unclear and it is difficult to select the priority that best suits the actual situation. By considering the key links in the entire microservice call link, the situation in which some applications on the link obtain higher priority scheduling guarantees but are slowed down by the downstream link is avoided, the division criteria of the hybrid business are clarified, and the resource utilization of the container instance is improved while ensuring business continuity. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0036] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0037] Figure 1 A flowchart of the first embodiment of the hybrid service processing method of the present application is provided;

[0038] Figure 2 A flowchart of the second embodiment of the hybrid service processing method of the present application is provided;

[0039] Figure 3A brief flowchart of the third embodiment of the hybrid service processing method of the present application is provided;

[0040] Figure 4 A schematic diagram of a brief process of classifying and classifying hybrid deployment services provided for the hybrid deployment service processing method of this application;

[0041] Figure 5 This is a schematic diagram of the module structure of the hybrid service processing device of the present application;

[0042] Figure 6 This is a schematic diagram of the device structure of the hardware operating environment involved in the hybrid service processing method of this application.

[0043] The purpose, features and advantages of this application will be further described in conjunction with the embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0044] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application and are not used to limit the present application.

[0045] In order to better understand the technical solution of the present application, a detailed description will be given below in conjunction with the accompanying drawings and specific implementation methods.

[0046] It should be noted that the execution subject of this embodiment may be a processor, or a computing service device with data processing, network communication and program running functions, such as a tablet computer, a personal computer, a mobile phone, etc., or an electronic device capable of realizing the above functions, a controller, etc. The following takes a processor as an example to illustrate this embodiment and the following embodiments.

[0047] Based on this, the embodiment of the present application provides a hybrid service processing method, referring to Figure 1 , Figure 1 This is a flowchart of the first embodiment of the hybrid service processing method of the present application.

[0048] In one implementation, the businesses in various embodiments of the present application are all financial businesses.

[0049] In this embodiment, the hybrid service processing method includes steps S10 to S30 at the application layer:

[0050] Step S10, according to the impact of the business system on the company and whether the business system is a key node in the microservice link, the importance level of the business system is divided to obtain business systems with different continuity assurance levels;

[0051] It should be noted that, first of all, business systems need to be divided into four levels according to the degree of impact of the business system on the company and whether the business system is a key node in the microservice link. Business systems of different importance levels have different requirements for high-availability deployment strategies, release process control, monitoring alarm thresholds, application architecture design, and deployment architecture optimization.

[0052] Specifically, the degree of impact on the company refers to the management consideration of factors such as the economic benefits and reputation of the business system on the enterprise; whether it is a key node in the microservice link refers to the technical identification of the key path of the business under the microservice architecture and the important microservice business on the key path, among which the key node refers to the node with a large fan-in / fan-out value.

[0053] As the banking data center business is fully deployed to the cloud, in order to better enable the business to enjoy the dividends of cloud resources, more and more single businesses in the data center are transforming to microservice architecture. The cloud of the microservice architecture can be separated from the technical focus, and more non-functional capabilities can be hosted in the cloud, which greatly improves the scalability, fault tolerance and maintainability of the application. However, the service dependency between services makes the performance change characteristics of microservice business significantly different from those of traditional single business. In the microservice architecture, the performance degradation of a single microservice will cause cascading performance degradation of other services. For example, an online type business contains two microservices, namely microservice A and microservice B. Microservice A is the service directly accessed by users. Microservice A will call microservice B in the process of processing user requests. If the performance of B deteriorates, the time A waits for B's return result will be longer, resulting in a longer response time for A to user requests. In the offline and online hybrid deployment job scheduling algorithm for microservice architecture, it is necessary to consider not only the performance interference caused by the job during hybrid deployment, but also the cascading performance degradation caused by the performance interference. Therefore, in order to ensure that the system classification can accurately reflect the business needs and provide the most appropriate SLA (service level agreement), the calling relationship between microservices within the system should also be fully considered during classification. Identify the key calling links, that is, the key paths through which data and requests flow in the business process.

[0054] Exemplarily, business systems with four different continuity assurance levels can be divided, namely, A-level core business systems, B-level important business systems, C-level general business systems, and D-level; among them, A-level refers to the core business systems on the key microservice paths of important businesses. Abnormalities or failures in A-level business systems will affect the entire business, directly affect user experience and corporate reputation, and cause significant losses; B-level refers to the general basic functional components on the key microservice paths of important businesses, which are called and relied upon by multiple parties. Failures in B-level business systems have a wider impact, indirectly affecting user experience and the economic benefits of the enterprise; C-level refers to single business processing systems or support systems on the key microservice paths of important businesses; D-level refers to other systems except A, B, and C.

[0055] Step S20, for each of the service systems, classifying services of different levels of each of the service systems according to the resource consumption degree and delay sensitivity degree of the services in each of the service systems;

[0056] It should be noted that services are divided into Level 1, Level 2, Level 3, and Level 4 according to their resource consumption and latency sensitivity;

[0057] In addition, it should be noted that it can also be divided according to tidal characteristics. Businesses without tidal characteristics are level 4, and businesses with tidal characteristics are other level 3 businesses. Among them, tidal characteristics refer to the fact that the resource demand of the business changes over time and presents a tidal-like fluctuation. In some periods, the business volume will increase significantly, forming a peak in resource demand; while in other periods, the business volume will decrease significantly, entering the trough of resource demand; and in periods with large business volume and frequent user activities, the system needs to process more requests, data or transactions, so the demand for computing resources (such as CPU, memory), network bandwidth, storage space, etc. increases significantly; tidal characteristics cause the difference in resource demand between peak and trough periods to be very significant. For example, during peak periods, resource demand may reach several times or even dozens of times that of trough periods, which is a challenge for resource management. It is difficult to ensure sufficient resource supply during peak resource demand and avoid over-allocation and waste of resources during low demand periods.

[0058] In one implementation, different continuity assurance levels include A-level core business systems, B-level important business systems, C-level general business systems, and business systems other than A-level, B-level, and C-level. Step S20 includes steps A10 to A40:

[0059] Step A10: if the resource consumption reaches a first preset threshold and the delay sensitivity does not reach a second preset threshold, the service is classified as a level 4 offline service;

[0060] It should be noted that latency sensitivity refers to the sensitivity of the business to response time. Resource consumption refers to the consumption of computing resources (such as CPU, memory), network resources, and storage resources by the container instance when the business is deployed in the container.

[0061] The resource consumption level reaches the first preset threshold value, which means that the consumption of computing resources (such as CPU, memory), network resources and storage resources of the container instance during operation reaches the first preset threshold value (for example, reaching 80%), and the sensitivity of the business to response time does not reach the second preset threshold value, which means that the interruption can be accepted.

[0062] Services that consume a lot of resources and can tolerate interruptions are classified into Level 4 offline services. Level 4 offline services have relatively stable resource requirements, support task and self-task breakpoint retries, allow business containers to be restarted, and have no clear latency requirements for task processing or interface returns.

[0063] Step A20, if the resource consumption level does not reach the first preset threshold and the delay sensitivity level reaches the second preset threshold, the service is classified into a level 3 online type service;

[0064] It should be noted that the resource consumption level not reaching the first preset threshold means that the demand for resources is relatively elastic, and the delay sensitivity level reaching the second preset threshold means that the delay sensitivity level is average and interruption is not accepted.

[0065] Services that have relatively flexible resource requirements, certain requirements for latency sensitivity, and cannot accept interruptions are classified as Level 3 online services.

[0066] Step A30, when the continuity guarantee level is level A or level B, if the resource consumption level does not reach the first preset threshold and the delay sensitivity level does not reach the third preset threshold, the service is classified as a level 2 online type service;

[0067] It should be noted that when the continuity assurance level is A or B, that is, the core business system on the key microservice path and the common basic functional components on the key microservice path, the delay sensitivity does not reach the third preset threshold, indicating that they are sensitive to delay and can tolerate 1% TP99 (the minimum time required to meet 99% of network requests) impact;

[0068] When the continuity guarantee level is A or B, services with relatively flexible resource requirements, sensitive to latency, and able to tolerate 1% TP99 impact are classified as Level 2 online services.

[0069] The degree of latency sensitivity is specifically divided into the following categories: TP99 performance is within 100ms, TP999 performance is within 500ms, resource interference in any dimension is not allowed, and services with glitches are not accepted, which are extremely latency-sensitive services; TP99 performance is within 500ms, TP999 performance is within 1s, occasional glitches are allowed, and services with TP99 interference within 1%-3% are latency-sensitive services; TP99 performance is within 5s, TP99 interference within 3%-5% is acceptable, and glitches have little impact on business logic are general latency services; services that can accept applications being suppressed or expelled and rebuilt when resources are limited are services with no latency requirements.

[0070] Step A40: When the continuity guarantee level is A or B, if the resource consumption level does not reach the first preset threshold and the delay sensitivity level reaches the third preset threshold, the service is classified as a level 1 online type service.

[0071] It should be noted that when the delay sensitivity reaches the third preset threshold, it means that the service is extremely sensitive to delay and cannot tolerate the impact of 1% TP99 (the minimum time required to meet 99% of network requests); when the continuity guarantee level is A or B, the service with more flexible resource requirements, extremely sensitive to delay and cannot tolerate the impact of 1% TP99 is classified as Level 1 online type service.

[0072] For example, level 1 is the highest priority service, which requires extremely low latency to ensure high performance and real-time response, such as Redis service; level 2 is the second highest priority service, which has higher latency requirements but can still tolerate a certain degree of resource interference; level 3 is a lower priority service, which is less sensitive to latency and can handle sudden resource demands, such as web services. Level 4 is a low priority service, which consumes a lot of resources and is less sensitive to latency. It can accept suppression and expulsion in resource scheduling, and is usually executed when resources are sufficient or during non-peak hours, such as big data analysis, batch processing tasks, AI training, and video transcoding.

[0073] In this implementation, when prioritizing colocation services, the key links in the entire microservice call link (including middleware dependencies, etc.) are taken into consideration, thereby avoiding the situation where some applications on the link are given higher priority scheduling guarantees but are slowed down by downstream links, and effectively improving the overall operational stability of microservice architecture applications in the colocation cluster.

[0074] Step S30: deploy the services of different levels into the colocation container cluster.

[0075] At present, for the exclusive kernel mode, different levels of services are deployed in the same container cluster. Although this improves resource utilization to a certain extent, the exclusive kernel mode requires online and offline services to occupy the entire physical machine. When the online service is at a low point, the entire physical machine is freed up for offline services by moving the entire machine. When the online service peaks again during the day, the offline service yields physical machine resources to the online service, and the entire physical machine is marked as used by online or offline services. This will lead to fragmentation and low utilization of operating resources.

[0076] Therefore, in this embodiment, in the scenario where business continuity is extremely high, a shared kernel colocation technology is selected, which has less impact on online and offline application scheduling and more significant resource utilization improvement effect. This allows the business system to maximize the configuration of operating resources while meeting different operating resource requirements, thereby improving resource configuration efficiency, and further improving system stability, thereby improving resource utilization.

[0077] It should be noted that a colocation container cluster refers to a containerized infrastructure that can simultaneously carry different levels of services. Through a flexible resource allocation and scheduling mechanism, the cluster can dynamically adjust resource allocation according to the characteristics and needs of different services, thereby achieving efficient resource utilization.

[0078] In this embodiment, business systems with different continuity assurance levels and businesses of different levels are divided according to the degree of influence of the business system on the company, the criticality of the business on the microservice link, and the degree of resource consumption and latency sensitivity of the business. In addition, after the hybrid business is divided, the different levels of business are deployed to the hybrid container cluster. Compared with the conventional means, the division criteria of the shared kernel mode hybrid business are unclear, and it is difficult to select the priority that best suits the actual situation. By considering the key links in the entire microservice call link, the situation in which some applications on the link obtain higher priority scheduling guarantees but are slowed down by the downstream link is avoided, the division criteria of the hybrid business are clarified, and the resource utilization of the container instance is improved while ensuring business continuity.

[0079] Based on the first embodiment of the present application, in the second embodiment of the present application, the same or similar contents as those in the first embodiment can refer to the above introduction, and will not be repeated in the following. Figure 2 After step S30, the hybrid service processing method further includes steps D10 to D30 at the platform layer:

[0080] Step D10, treating the non-low-priority service as an online type service, and treating the low-priority service as an offline type service;

[0081] It should be noted that online services refer to services that have long running times, tidal characteristics in service traffic and resource utilization, are very sensitive to delays, and require real-time responses, such as message feed services, e-commerce transaction services, and middleware services.

[0082] Offline services refer to tasks that are executed during off-peak hours and have high resource utilization during operation, relatively low latency requirements, and can be performed within time intervals. These tasks include big data processing, batch processing, machine learning training, MapReduce (a programming framework for distributed computing programs) in the Hadoop (a software framework for storage and computing systems) ecosystem, and Spark (a big data processing framework) services.

[0083] Step D20, according to the requirements of the online type services and the offline type services for container instance resources, allocating container instance resources to each container in the colocation container cluster where each of the online type services and the offline type services are located;

[0084] It should be noted that in order to optimize resource usage and costs, data centers or cloud platforms will deploy these different services in a co-location container cluster. Co-location means running different types of services simultaneously in the same physical or logical cluster. Container technology (such as Docker and Kubernetes) provides an isolated operating environment, allowing multiple independent services or applications to run safely on the same host.

[0085] Calculate the specific resource requirements of online and offline services based on their characteristics and current operating status. Optimize resource allocation through algorithms or strategies while ensuring service quality, such as giving priority to the resource requirements of online services while using the flexibility of offline services to fill idle resource periods. Adjust resource allocation in real time based on changes in business load to cope with peak online business and batch processing requirements of offline services. Continuously monitor resource usage and business performance, and collect data feedback for subsequent resource optimization decisions.

[0086] In one embodiment, step D20 includes steps E10 to E30:

[0087] The non-low priority includes level 1, level 2 and level 3;

[0088] Step E10: The container of the level 1 online type service exclusively occupies the physical core and all logical cores in the physical core;

[0089] It should be noted that modern processors usually contain multiple physical cores, each of which can execute instructions independently. Multi-core processors are the basis of modern servers and high-performance computing systems. They can handle multiple tasks at the same time and significantly improve system performance. With the support of hyper-threading technology (such as Intel's Hyper-Threading), one physical core can simulate two or more logical cores, each of which can execute tasks independently. In this way, the processor can process more threads at the same time and improve multi-tasking capabilities. Level 1 online type business has extremely high requirements for system response time and stability, and cannot tolerate any unnecessary delays or interruptions. In order to ensure the performance of this type of business, the system will assign it to an exclusive physical core to run.

[0090] In addition, it should be noted that when an online service is marked as level 1, the system will allocate a physical core to it. This physical core and all logical cores on it will be dedicated to this service and will not be shared by any other tasks or services. This means that the service can exclusively use all computing resources of the core, including CPU computing power, cache, etc., to ensure that its operation is not interfered with by other services and achieve optimal performance and response speed.

[0091] Exemplarily, at the kernel layer, level 1 online type services exclusively occupy the CPU physical cores. According to resource requirements, multiple containers of level 1 online type services are scheduled to a fixed number of physical cores and occupy all logical cores in the same CPU physical core. At the same time, the use of CPU Burst (central processing unit burst time) technology is supported to meet sudden resource requirements; at the container layer, when the physical machine resource consumption reaches the safety waterline, the successfully deployed level 1 online type services will not be evicted, and the amount of resources occupied will remain unchanged; when the level 1 online type services generate traffic spikes and have sudden resource requirements, they can immediately seize the CPU, memory, network bandwidth, disk IO and other resources of low-priority offline type services for scheduling.

[0092] Step E20, the container of the level 2 online type service exclusively occupies the physical core and some logical cores in the physical core, and the remaining logical cores carry the level 4 offline type service;

[0093] It should be noted that although the Level 2 online type business does not have extreme requirements for real-time and response time like the Level 1 online business, it still has high requirements for performance and stability. They may include some background processing, data synchronization tasks, etc., and a certain service quality must be guaranteed. Similar to the Level 1 online business, the Level 2 online business will also be assigned to a physical core to run. However, unlike the former, it does not occupy all the logical cores of the physical core. The system will allocate some logical cores in the physical core to this type of business to ensure that it has stable resources available. After the Level 2 online business occupies some logical cores, the remaining logical cores on the physical core will be allocated to the Level 4 offline type business. This type of business usually does not have high requirements for real-time performance, but may require a large amount of computing resources to process batch tasks, data analysis, etc. By sharing some resources of the same physical core with the online business, the offline business can use the resources in idle or low-load periods without affecting the performance of the online business, thereby improving the resource utilization of the overall system.

[0094] Exemplarily, at the kernel layer, the level 2 online type business occupies the CPU physical core exclusively. According to resource requirements, multiple containers of the level 1 online type business are scheduled on one of the logical cores of each physical core in a fixed number of physical cores, and the other logical core can carry the level 4 offline type business; when the level 2 online type business has sudden resource requirements, it can immediately seize the CPU, memory, network bandwidth, disk IO and other resources of the level 4 offline type business to meet its sensitive latency requirements. At the container layer, when the physical machine resource consumption reaches the safety waterline, the successfully deployed level 2 online type business will not be evicted, and the amount of resources occupied will remain unchanged; when the level 2 online type business generates a traffic spike and has a sudden resource demand, it can immediately seize the CPU, memory, network bandwidth, disk IO and other resources of the level 4 offline type business for scheduling.

[0095] Step E30: The containers of the level 3 online type service and the level 4 offline type service share the remaining physical cores.

[0096] It should be noted that the requirements of level 3 online type services on system response time and performance are relatively low. They may include some non-critical user services, non-real-time data processing or background maintenance tasks. Although their demand for resources is not as urgent as that of level 1 services, they still require certain computing resources to ensure their operation. Level 4 offline type services involve batch processing of large amounts of data, data analysis, model training, etc. They do not require high real-time response, but may require long-term operation and a large amount of computing resources. During non-peak hours when resources are sufficient, such services can make full use of remaining resources and improve overall resource utilization. In the later stage of resource allocation, when sufficient resources have been allocated to level 1 and level 2 services, the remaining computing resources (including physical cores) will be shared by multiple level 3 online type services and level 4 offline type services. This sharing mode allows multiple services to run in parallel, and through dynamic scheduling and resource management, it ensures that all services have the opportunity to obtain the necessary computing resources, while avoiding idleness and waste of resources.

[0097] For example, at the kernel layer, level 3 online services share CPU physical cores, and multiple containers of the services can be scheduled to physical cores where containers of non-level 1 and level 2 services are located; when level 3 online services have sudden resource demands, they can immediately seize the CPU, memory, network bandwidth, disk IO and other resources of level 4 offline services, or use the idle resources of other level 3 online services to meet their more sensitive latency requirements. At the container layer, when the physical machine resource consumption reaches the safety waterline, the successfully deployed level 3 online services will not be evicted; when level 3 online services have traffic spikes and sudden resource demands, they can immediately seize the CPU, memory, network bandwidth, disk IO and other resources of level 4 offline services for scheduling.

[0098] At the kernel layer, level 4 offline services share CPU physical cores, and multiple containers of this service can be scheduled to physical cores other than level 1 and level 2 service containers, as well as idle logical cores of level 3 online services. Since offline services have no latency requirements, when new offline services are to be deployed, they queue up and wait for available resources. At the container layer, when the node resource utilization reaches different thresholds, different scheduling strategies will be triggered to give resources to the highest, second highest, and lower level online services.

[0099] In this embodiment, the resource allocation strategy for level 1 online services under the multi-core processor architecture. In order to ensure high performance and low latency of such services (such as real-time trading systems, etc.), the system will adopt a mechanism called "exclusive core". This exclusive core allocation strategy, especially when processing high loads or critical tasks, can significantly improve the execution efficiency of the business and the overall service quality of the system, and is an advanced optimization method in resource management.

[0100] In a resource-constrained environment, resource allocation strategies are used to meet the needs of services of different priorities. Specifically for level 2 online services and level 4 offline services, the system adopts a "shared core" allocation mechanism, which not only guarantees the performance requirements of online services, but also makes full use of computing resources; in this way, the system can dynamically adjust the allocation of logical cores while ensuring the performance of online services, realize reasonable support for level 2 online services and level 4 offline services, and achieve the purpose of resource optimization and business balance.

[0101] In resource management, once the business requirements of level 1 and level 2 are met, the remaining computing resources will be allocated to level 3 online type business and level 4 offline type business to maximize resource utilization and parallel processing of multiple businesses. The key to the shared resource strategy lies in dynamic scheduling and load balancing. The system will adjust resource allocation in real time according to the current business load and resource usage to ensure that all businesses can be completed within a reasonable time while maximizing resource utilization efficiency.

[0102] Step D30, when the container instance resource consumption of the online type business on the node of each container reaches a preset water level, the container instance resources of the offline type business running on the node are scheduled and suppressed, and / or the container instance of the offline type business running on the node is expelled.

[0103] It should be noted that in resource management, the preset watermark refers to the usage threshold set by the system for various resources. When the resource usage reaches or approaches this threshold, the system will trigger the corresponding resource management policy. The setting of the preset watermark needs to take into account factors such as business needs, resource capacity, and system load to ensure the reasonable allocation and efficient use of resources. When the system monitors that the usage of key resources (such as CPU, memory, network bandwidth, etc.) reaches the preset watermark, the system will take measures to adjust the operation of offline type services. These measures may include: prohibiting scheduling, resource suppression, and container instance eviction.

[0104] In one embodiment, step D30 includes steps F10 to F30:

[0105] Step F10: When the resource consumption of the container instance reaches the scheduling prohibition waterline, scheduling of the container instance resources of the offline type service is prohibited;

[0106] It should be noted that in a cloud computing environment, container instance resources usually refer to the computing resources required to support business operations, including but not limited to CPU computing power, memory, storage, and network bandwidth. The scheduling prohibition watermark is a set resource usage threshold. When the consumption of container instance resources reaches or exceeds this threshold, the system will take strict resource control measures to avoid system performance degradation or service interruption caused by excessive resource consumption.

[0107] For example, as the node resource utilization rate increases, the first thing to be reached is the scheduling prohibition watermark. When the threshold set by the watermark is reached, offline type services will not be able to add new containers on the node.

[0108] When it is detected that the operating resource consumption reaches the prohibited scheduling waterline, the system will automatically execute an emergency resource protection mechanism, that is, stop or refuse to allocate new operating resources for offline type services.

[0109] Step F20, when the resource consumption of the container instance reaches the resource suppression waterline, a bandwidth threshold is set for the offline type service. If the bandwidth of the offline type service is greater than the bandwidth threshold, packet loss processing is performed on the offline type service.

[0110] It should be noted that the resource suppression waterline is a threshold set in a resource management policy. When the consumption of network resources reaches this waterline, the system will take measures to limit or adjust resource allocation.

[0111] When the resource suppression watermark is triggered, the system will limit the network bandwidth usage of offline services and control the network resource consumption of offline services by setting bandwidth thresholds. When the actual network bandwidth usage of offline services exceeds the set bandwidth threshold, the system will execute the packet loss strategy, that is, discard some network data packets to reduce the consumption of network resources.

[0112] Exemplarily, as the node resource utilization rate continues to rise, after exceeding the scheduling prohibition waterline, the next waterline to arrive is the resource suppression waterline. The four resources of CPU, memory, disk io, and network bandwidth all have different resource suppression waterline thresholds. Memory asynchronous recovery waterline: When the node memory resource triggers the asynchronous recovery waterline, the memory resources of the offline type business will be recycled first and given to the highest, second highest, and lower level online type business. Container network speed limit: The offline type business will set the maximum bandwidth, that is, the bandwidth threshold. If the threshold is exceeded, the packet will be dropped to ensure that the highest, second highest, and lower level online type business have sufficient network bandwidth available. At the same time, the offline type business will set the minimum resource guarantee, that is, the minimum amount of resources that can be used, to maintain the normal operation of the offline type business.

[0113] Step F30 , when the resource consumption of the container instance reaches the eviction watermark, the offline type services are evicted, and the containers of the level 1 and level 2 online type services are allowed to occupy the remaining container instance resources.

[0114] It should be noted that the eviction watermark is a resource usage threshold set by the system. When resource consumption reaches or exceeds this threshold, the system will initiate an emergency resource management strategy. When resource consumption reaches the eviction watermark, the system will execute the resource eviction strategy and forcibly migrate some or all offline-type services from the currently running physical resources to release resources. Offline services usually do not have high real-time requirements and can withstand a certain degree of delay or interruption. Therefore, they become the primary target of resource eviction when resources are extremely tight. After resource eviction, the system will give priority to allocating the remaining operating resources to level 1 and level 2 online-type services. Online services, especially level 1 and level 2 services, have high requirements for real-time performance and service quality.

[0115] Exemplarily, when the resource consumption of the physical machine rises to the eviction waterline, the offline type services that have been successfully deployed are first evicted until the node resources do not exceed the eviction waterline; within the 4-level offline services, the service eviction priority can be further defined, and the services are evicted in order of priority; when the resource demand of the online type services of levels 1, 2, and 3 shows a continuous downward trend, and the resource consumption level of the nodes where the online type services of levels 1, 2, and 3 are located falls below the suppression waterline, the resource supply of the 4-level offline services that have not been evicted on the node is slowly restored; the 4-level offline services that have not been evicted gradually occupy the idle resources of the online type services, and maintain the node resource consumption below the suppression waterline, wherein the 4-level offline services that have not been evicted are the remaining 4-level offline services after the node resources have fallen below the eviction waterline after some offline services have been evicted).

[0116] In this embodiment, when resources are extremely tight, the resource scheduling of offline services is restricted or prohibited to give priority to online services. The reason for adopting this strategy is that offline services have lower requirements for real-time response time and can withstand a certain degree of delay or interruption, while online services require continuous and stable resource support to ensure service quality and user experience. By restricting the resource allocation of offline services when resources are tight, data centers or cloud platforms can more effectively cope with resource pressure and ensure the stable operation of key services.

[0117] In addition, bandwidth is a key performance indicator in network resource management, especially under high load and resource constraints. By setting network bandwidth thresholds to control the network resource usage of offline services, it is possible to prioritize the network resource requirements of online services when resources are tight, while reasonably controlling the network resource consumption of offline services to avoid system performance degradation or service interruption caused by excessive resource consumption.

[0118] By implementing a resource management strategy based on the eviction watermark, the system can release resources by evicting offline businesses when resources are extremely tight, giving priority to the operation of key online businesses, which helps to achieve reasonable allocation and efficient use of resources when resources are limited.

[0119] In another embodiment, in terms of operating mode, since level 1 and level 2 core-binding services exclusively occupy a group of CPU physical cores, when implementing such application admission to the mixed-deployment container cluster, it is necessary to consider the overall utilization rate of the cluster and control the proportion of level 1 and level 2 service CPU resource applications to the total CPU resources of the entire physical machine mixed-deployment container cluster; if level 1 and level 2 core-binding services account for a large proportion, when these services are in off-peak periods, more idle resources will be generated, resulting in a lower average resource utilization rate for the cluster.

[0120] In terms of technology selection, a non-invasive and scalable mode can be selected to install the hybrid function plug-in on the private cloud's self-developed container cluster scheduler, thereby achieving unified scheduling of online and offline containers with a relatively low modification cost, and being compatible with the original private cloud container cluster version.

[0121] In terms of supporting measures, observable indicators such as the kernel layer and container cluster layer are used in conjunction to timely monitor the interference of online applications at the CPU, memory, network and IO levels. When the threshold is reached, offline application resources can be automatically preempted for recovery, and business administrators can be notified to observe whether the results meet expectations.

[0122] In this embodiment, in resource management, the operation strategy of offline type services is dynamically adjusted according to resource usage. Based on the dynamic resource management strategy of the preset water level, the system can give priority to the operation of high priority services when resources are tight, and reasonably arrange offline services to ensure efficient use of resources and smooth operation of services. This resource management strategy helps to improve resource utilization, reduce costs, and ensure stable operation and efficient processing of services.

[0123] Based on the first embodiment of the present application, in the third embodiment of the present application, the same or similar contents as those in the first embodiment can refer to the above introduction, and will not be repeated later. Figure 3Different levels of services include online services and offline services. After step S30, the hybrid service processing method further includes steps T10 to T20 at the infrastructure layer:

[0124] Step T10, combining multiple single-availability zone colocation container subclusters into a logical federated cluster, wherein the single availability zone is a subcluster of the colocation container;

[0125] It should be noted that in cloud computing, an availability zone (AZ) refers to a physically independent data center area with independent power supply, cooling and network facilities. A single availability zone refers to an independently operated, resource-isolated subcluster area created by a management node in a container cluster. It can be regarded as an abstraction and simulation of the concept of a physical availability zone in a containerized environment.

[0126] A subcluster of a colocation container cluster refers to an environment where both online and offline services are deployed and run in a container cluster. Online services usually have high requirements for real-time performance and service quality, while offline services have high requirements for resource consumption and running time flexibility. A logical federation cluster refers to a unified cluster that crosses physical boundaries by logically connecting multiple physically independent single-availability zone subclusters through a network and management software. This cluster structure allows resources to be dynamically scheduled and load balanced across multiple physical areas, while providing a unified management and access interface to enhance the flexibility and availability of the system. A single-availability zone subcluster refers to one or more container clusters running in a physical availability zone. Each single-availability zone subcluster has independent resources and network environment to achieve resource isolation and fault domain division.

[0127] By using two or more single-availability zone co-location container subclusters as logical federated clusters, resource isolation and optimized scheduling can be achieved to improve availability.

[0128] Step T20, by backing up the online type services in the two sub-clusters of the same logical federated cluster, when one of the sub-clusters is abnormal, the services of the other cluster are used to provide external services.

[0129] It should be noted that online services usually have high requirements for real-time performance and service quality. In order to ensure high availability and data security, a backup strategy needs to be implemented in the logical federation cluster. This backup can be data backup to ensure redundant data storage; it can also be service backup to achieve service redundancy and failover by deploying the same or similar online services in multiple single-availability zone subclusters. Under the framework of the logical federation cluster, the implementation of the backup strategy can cross physical boundaries and improve the utilization efficiency of backup resources and the reliability of backup.

[0130] The subclusters of each federated cluster are deployed across multiple cabinets, and the three or more management nodes of each subcluster are placed in different cabinets to avoid the impact of cabinet-level failures.

[0131] In this embodiment, different types of services are managed and deployed through container technology to achieve efficient use of resources and flexible scheduling of services. Containerization and clustering are key technologies for achieving resource isolation, elastic expansion and efficient utilization. By creating a single availability zone in a container cluster, the system can achieve refined management and scheduling of resources, ensuring the optimization of resource isolation and scheduling strategies for online and offline services. By building a logical federation cluster and implementing a backup strategy for online services, the system can achieve cross-regional scheduling and optimal utilization of resources while providing high availability and data security. In the event of a local failure or fluctuation in resource demand, the logical federation cluster can quickly adjust resource allocation to ensure the continuous operation and service quality of online services. In the two subclusters of the same federation, each publishes a copy of the same application to back up each other. When one of the clusters is abnormal, the application of the other cluster can provide services to the outside world normally, which can provide a larger resource pool and higher fault tolerance.

[0132] For example, please refer to Figure 4 , Figure 4 A schematic diagram of a hybrid deployment service classification process of a hybrid deployment service processing method is provided, specifically:

[0133] The hybrid services are classified according to whether the services are insensitive to latency and whether they can accept interruptions. If they are insensitive to latency and can accept interruptions, they are classified as level 4 offline services. If they are sensitive to latency and cannot accept interruptions, the importance level is determined to be A or B. If they are neither A nor B, they are level 3 online services without core binding. If they are A or B, it is determined whether they can tolerate a 1% TP99 impact. If they cannot tolerate it, the platform administrator approves the issuance of operating resources and classifies them as level 1 online services with core binding. If they can tolerate it, the platform administrator approves the issuance of operating resources and classifies them as level 2 online services with core binding.

[0134] It should be noted that the above examples are only used to understand the present application and do not constitute a limitation on the hybrid service processing method of the present application. More simple transformations based on this technical concept are all within the protection scope of the present application.

[0135] This application also provides a hybrid service processing device, please refer to Figure 5 , the hybrid service processing device includes:

[0136] The first division module 10 is used to divide the importance level of the business system according to the impact of the business system on the company and whether the business system is a key node in the microservice link, so as to obtain business systems with different continuity assurance levels;

[0137] The second division module 20 is used to divide the services of each service system into different levels according to the resource consumption degree and delay sensitivity degree of the services in each service system;

[0138] The deployment module 30 is used to deploy the services of different levels into the colocation container cluster.

[0139] Optionally, the second classification module 20 is further configured to classify the service into a level 4 offline type service if the resource consumption level in the service characteristics reaches a first preset threshold and the delay sensitivity level does not reach a second preset threshold;

[0140] If the resource consumption level in the service characteristics does not reach the first preset threshold and the delay sensitivity level reaches the second preset threshold, the service is classified as a level 3 online type service;

[0141] When the continuity guarantee level is A or B, if the resource consumption level in the service characteristics does not reach the first preset threshold and the delay sensitivity level does not reach the third preset threshold, the service is classified as a level 2 online type service;

[0142] When the continuity guarantee level is A or B, if the resource consumption level in the service characteristics does not reach the first preset threshold and the delay sensitivity reaches the third preset threshold, the service is classified as a level 1 online type service.

[0143] Optionally, the deployment module 30 is further configured to treat the non-low-priority service as an online type service and treat the low-priority service as an offline type service;

[0144] According to the requirements of the online type services and the offline type services for container instance resources, the container instance resources are allocated to each container in the colocation container cluster where each of the online type services and the offline type services are located;

[0145] When the resource consumption of the container instance of the online type business on the node of each container reaches a preset water level, the container instance resources of the offline type business running on the node are scheduled and suppressed, and / or the container instance of the offline type business running on the node is expelled.

[0146] Optionally, the deployment module 30 is also used for the container of the level 1 online type service to exclusively occupy the physical core and all logical cores in the physical core;

[0147] The container of the level 2 online type service exclusively occupies the physical core and some logical cores in the physical core, and the remaining logical cores carry the level 4 offline type service;

[0148] The containers of the level 3 online type service and the level 4 offline type service share the remaining physical cores.

[0149] Optionally, the deployment module 30 is further configured to prohibit scheduling of container instance resources for offline services when the consumption of container instance resources reaches a scheduling prohibition waterline;

[0150] When the resource consumption of the container instance reaches the resource suppression waterline, a bandwidth threshold is set for the offline type service. If the bandwidth of the offline type service is greater than the bandwidth threshold, packet loss is performed on the offline type service.

[0151] When the resource consumption of the container instance reaches the eviction watermark, the offline type business is evicted, allowing the containers of the level 1 and level 2 online type business to occupy the remaining container instance resources.

[0152] Optionally, the deployment module 30 is further configured to form a logical federated cluster from multiple colocation container subclusters in a single availability zone, wherein the single availability zone is a subcluster of the colocation container;

[0153] By backing up the online type services in the two sub-clusters of the same logical federated cluster, when one of the sub-clusters is abnormal, the services of the other cluster are used to provide external services.

[0154] The hybrid service processing device provided by the present application adopts the hybrid service processing method in the above embodiment, which can solve the technical problem of low operating resource utilization of the hybrid service. Compared with the prior art, the beneficial effects of the hybrid service processing device provided by the present application are the same as the beneficial effects of the hybrid service processing method provided by the above embodiment, and other technical features in the hybrid service processing device are the same as the features disclosed in the above embodiment method, which will not be repeated here.

[0155] The present application provides a hybrid service processing device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the hybrid service processing method in the first embodiment described above.

[0156] Reference below Figure 6, which shows a schematic diagram of the structure of a hybrid service processing device suitable for implementing the embodiment of the present application. The hybrid service processing device in the embodiment of the present application may include but is not limited to mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions), PMPs (Portable Media Players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 6 The hybrid service processing device shown is only an example and should not limit the functions and scope of use of the embodiments of the present application.

[0157] like Figure 6 As shown, the hybrid service processing device may include a processing device 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM: Read Only Memory) 1002 or a program loaded from a storage device 1003 to a random access memory (RAM: Random Access Memory) 1004. In RAM1004, various programs and data required for the operation of the hybrid service processing device are also stored. The processing device 1001, ROM1002, and RAM1004 are connected to each other through a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Generally, the following systems can be connected to the I / O interface 1006: an input device 1007 including, for example, a touch screen, a touchpad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; an output device 1008 including, for example, a liquid crystal display (LCD: Liquid Crystal Display), a speaker, a vibrator, etc.; a storage device 1003 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 1009. The communication device 1009 may allow the hybrid service processing device to communicate with other devices wirelessly or by wire to exchange data. Although the figure shows a hybrid service processing device with various systems, it should be understood that it is not required to implement or have all the systems shown. More or fewer systems may be implemented or have alternatively.

[0158] In particular, according to the embodiments disclosed in the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present application include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through a communication device, or installed from a storage device 1003, or installed from a ROM 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the method of the embodiment disclosed in the present application are executed.

[0159] The hybrid service processing device provided by the present application adopts the hybrid service processing method in the above embodiment, which can solve the technical problem of low operating resource utilization of the hybrid service. Compared with the prior art, the beneficial effects of the hybrid service processing device provided by the present application are the same as the beneficial effects of the hybrid service processing method provided by the above embodiment, and other technical features in the hybrid service processing device are the same as the features disclosed in the method of the previous embodiment, which will not be repeated here.

[0160] It should be understood that the various parts disclosed in this application can be implemented by hardware, software, firmware or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

[0161] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0162] The present application provides a computer-readable storage medium having computer-readable program instructions (ie, computer programs) stored thereon, and the computer-readable program instructions are used to execute the hybrid service processing method in the above embodiment.

[0163] The computer-readable storage medium provided in the present application may be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems or devices, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM: Random Access Memory), a read-only memory (ROM: Read Only Memory), an erasable programmable read-only memory (EPROM: Erasable Programmable Read Only Memory or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM: CD-Read Only Memory), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system or device. The program code contained on the computer-readable storage medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (Radio Frequency: Radio Frequency), etc., or any suitable combination of the above.

[0164] The computer-readable storage medium may be included in the hybrid service processing device; or may exist independently without being installed in the hybrid service processing device.

[0165] The computer-readable storage medium carries one or more programs. When the one or more programs are executed by the hybrid business processing device, the hybrid business processing device: classifies the importance of the business system according to the impact of the business system on the company and whether the business system is a key node on the microservice link, and obtains business systems with different continuity assurance levels;

[0166] For each of the service systems, different levels of services of each of the service systems are obtained according to the resource consumption degree and delay sensitivity degree of the services in each of the service systems;

[0167] Deploy the different levels of services into the colocation container cluster.

[0168] Computer program code for performing the operations of the present application may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0169] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present application. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0170] The modules involved in the embodiments described in this application may be implemented by software or hardware, wherein the name of the module does not constitute a limitation on the unit itself in some cases.

[0171] The readable storage medium provided in the present application is a computer-readable storage medium, which stores computer-readable program instructions (i.e., computer programs) for executing the above-mentioned hybrid service processing method, and can solve the technical problem of low operating resource utilization of hybrid services. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in the present application are the same as the beneficial effects of the hybrid service processing method provided in the above-mentioned embodiment, and will not be repeated here.

[0172] The present application also provides a computer program product, including a computer program, which implements the steps of the above-mentioned hybrid service processing method when executed by a processor.

[0173] The computer program product provided in this application can solve the technical problem of low resource utilization of the operation of the colocation service. Compared with the prior art, the beneficial effects of the computer program product provided in this application are the same as the beneficial effects of the colocation service processing method provided in the above embodiment, which will not be repeated here.

[0174] The above descriptions are only some embodiments of the present application, and are not intended to limit the patent scope of the present application. All equivalent structural changes made using the contents of the present application specification and drawings under the technical concept of the present application, or direct / indirect applications in other related technical fields are included in the patent protection scope of the present application.

Claims

1. A method for processing hybrid services, characterized in that: At the application layer, the method includes: According to the impact of the business system on the company and whether the business system is a key node in the microservice link, the importance level of the business system is divided to obtain business systems with different continuity assurance levels; For each of the business systems, different levels of services of each of the business systems are obtained according to the resource consumption degree and delay sensitivity degree of the services in each of the business systems; Deploy the different levels of services into the colocation container cluster.

2. The method according to claim 1, characterized in that Different continuity assurance levels include A-level core business systems, B-level important business systems, C-level general business systems, and business systems other than A-level, B-level, and C-level. For each of the business systems, the steps of obtaining different levels of business in each of the business systems according to the resource consumption degree and delay sensitivity of the business in each of the business systems include: If the resource consumption reaches a first preset threshold and the delay sensitivity does not reach a second preset threshold, the service is classified as a level 4 offline type service; If the resource consumption level does not reach the first preset threshold and the delay sensitivity level reaches the second preset threshold, the service is classified as a level 3 online type service; When the continuity guarantee level is A or B, if the resource consumption level does not reach the first preset threshold and the delay sensitivity level does not reach the third preset threshold, the service is classified as a level 2 online type service; When the continuity guarantee level is A or B, if the resource consumption level does not reach the first preset threshold and the delay sensitivity level reaches the third preset threshold, the service is classified as a level 1 online type service.

3. The method according to claim 1, characterized in that At the platform layer, the level of the divided services of each of the service systems includes non-low priority and low priority. After the step of deploying the services of different levels into the colocation container cluster, the following steps are included: Treat the non-low-priority service as an online type service, and treat the low-priority service as an offline type service; According to the requirements of the online type services and the offline type services for container instance resources, the container instance resources are allocated to each container in the colocation container cluster where each of the online type services and the offline type services are located; When the resource consumption of the container instance of the online type business on the node of each container reaches a preset water level, the container instance resources of the offline type business running on the node are scheduled and suppressed, and / or the container instance of the offline type business running on the node is expelled.

4. The method according to claim 3, characterized in that The non-low priority includes level 1, level 2 and level 3. The step of allocating container instance resources to each container in the colocation container cluster where each of the online type services and the offline type services are located according to the requirements of each of the online type services and the offline type services for container instance resources includes: The container of the level 1 online type service exclusively occupies the physical core and all logical cores in the physical core; The container of the level 2 online type service exclusively occupies the physical core and some logical cores in the physical core, and the remaining logical cores carry the level 4 offline type service; The containers of the level 3 online type service and the level 4 offline type service share the remaining physical cores.

5. The method according to claim 3, characterized in that When the resource consumption of the container instance of the online type service on the node of each container reaches a preset water level, the step of scheduling and suppressing the container instance resources of the offline type service running on the node, and / or expelling the container instance of the offline type service running on the node includes: When the resource consumption of container instances reaches the prohibition level for scheduling, scheduling of container instance resources for offline services is prohibited. When the resource consumption of the container instance reaches the resource suppression waterline, a bandwidth threshold is set for the offline type service. If the bandwidth of the offline type service is greater than the bandwidth threshold, packet loss is performed on the offline type service. When the resource consumption of the container instance reaches the eviction watermark, the offline type business is evicted, allowing the containers of the level 1 and level 2 online type business to occupy the remaining container instance resources.

6. The method according to claim 1, characterized in that At the infrastructure layer, the different levels of services include online type services and offline type services. After the step of deploying the different levels of services into the colocation container cluster, the following step is further included: Combining multiple single-availability zone colocation container subclusters into a logical federated cluster, wherein the single availability zone is a subcluster of the colocation container; By backing up the online type services in the two sub-clusters of the same logical federated cluster, when one of the sub-clusters is abnormal, the services of the other cluster are used to provide external services.

7. A hybrid service processing device, characterized in that: The hybrid service processing device includes: The first classification module is used to classify the importance of business systems according to the impact of the business systems on the company and whether the business systems are key nodes in the microservice link, and obtain business systems with different continuity assurance levels; The second division module is used to divide the services of each service system into different levels according to the resource consumption degree and delay sensitivity degree of the services in each service system; The deployment module is used to deploy the services of different levels into the colocation container cluster.

8. A hybrid service processing device, characterized in that: The device comprises: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program is configured to implement the steps of the hybrid service processing method according to any one of claims 1 to 6.

9. A storage medium, characterized in that: The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, the steps of the hybrid service processing method according to any one of claims 1 to 6 are implemented.

10. A computer program product, characterized in that The computer program product includes a computer program, and when the computer program is executed by a processor, the steps of the hybrid service processing method according to any one of claims 1 to 6 are implemented.

Images