Distributed management system and method for cloud container cluster

Abstract

The invention relates to the technical field of cloud computing, and discloses a distributed management system and method for a cloud container cluster, and the system comprises a data collection module which is used for collecting the resource remaining amount of each node in the cloud container cluster and the resource demand of a tenant, and providing real-time data updating support; the state modeling module is used for establishing a state variable model for describing the dynamic change of tenant resources based on the resource data acquired by the data acquisition module, and state variables comprise the resource occupation amount, the resource application amount and the resource release rate of the tenants; the scheduling optimization module is used for generating a resource allocation strategy based on the tenant resource dynamic change characteristics and the multi-tenant competition relationship; according to the invention, through system architectures of data acquisition, state modeling, scheduling optimization and the like which are modularly designed, dynamic acquisition frequency and resource dynamic change modeling are introduced, the data acquisition real-time performance and the resource allocation accuracy are improved, and the problems of low resource utilization rate and slow scheduling response in the prior art are solved.

Description

Technical Field

[0001] The present invention relates to the technical field of cloud computing, and in particular to a distributed management system and method for a cloud container cluster. Background Art

[0002] The distributed management system of cloud container clusters is widely used in microservice architecture. It optimizes the management process of containerized applications through automated deployment, expansion, monitoring and operation and maintenance. It is widely used in microservice architecture, DevOps, elastic computing, edge computing, high availability and hybrid cloud scenarios. Its core functions include container orchestration, resource management, network and services, monitoring logs, security authentication, etc., which can effectively improve the availability and scalability of the system. For example, enterprises can use the system to achieve rapid application delivery, dynamic resource allocation, multi-cloud deployment and fault recovery to adapt to high concurrency or complex distributed requirements. However, it should be noted that the learning curve of the system is high, involving complex fields such as network, security, and scheduling, requiring professional knowledge support and effective troubleshooting capabilities.

[0003] Existing cloud container management systems have many deficiencies in dynamic scheduling and resource optimization. Most systems use static scheduling strategies or rule-based resource allocation methods, lack the ability to dynamically respond to load fluctuations, and are unable to meet the needs of sudden tasks and complex scenarios. Scheduling strategies are usually based on single-node or local optimization, and fail to optimize resource allocation from a global perspective, resulting in uneven resource utilization and even serious waste. In addition, the predefined resource quota method is prone to idle resources, and the system's limited ability to recycle and reallocate resources further exacerbates resource waste.

[0004] In summary, the distributed management system of cloud container clusters has insufficient service quality assurance capabilities in a multi-tenant environment. High-priority tasks may be delayed due to insufficient resources, and low-priority tasks are frequently interrupted. Traditional system scheduling strategies lack intelligent optimization capabilities, cannot learn and optimize based on historical operation data, and are difficult to balance multiple objective requirements such as task delay, energy efficiency, and resource utilization. Summary of the invention

[0005] In view of the shortcomings of the prior art, the present invention provides a distributed management system and method for a cloud container cluster. Through the combination of modular design, dynamic mathematical modeling and intelligent optimization algorithm, it solves the problem of lack of dynamic response capability to load fluctuations and difficulty in meeting the needs of sudden tasks and complex scenarios.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: A distributed management system for a cloud container cluster, comprising: The data collection module is used to collect the remaining resources of each node in the cloud container cluster and the resource requirements of tenants, and provide real-time data update support; A state modeling module, used to establish a state variable model describing the dynamic changes of tenant resources based on the resource data collected by the data collection module, wherein the state variables include the tenant's resource occupancy, resource application amount and resource release rate; The scheduling optimization module is used to generate resource allocation strategies through optimization algorithms based on the dynamic change characteristics of tenant resources and the competition relationship among multiple tenants; A scheduling execution module, used to dynamically adjust the allocation of cloud container resources according to the resource allocation strategy generated by the scheduling optimization module, including allocating computing resources and adjusting container deployment between nodes; The feedback optimization module is used to monitor the scheduling execution effect in real time and adjust the optimization strategy according to the operating status of resource allocation and scheduling results.

[0007] Preferably, the state variable model established by the state modeling module describes the tenant resource changes with a dynamic equation, that is, the tenant's resource occupancy changes over time and is jointly determined by the tenant's resource application amount and release rate. The model includes the correlation between the tenant's resource occupancy and resource request behavior.

[0008] Preferably, the core function of the scheduling optimization module is to generate a resource allocation strategy based on the data and utility function provided by the state modeling module. The scheduling optimization module aims to maximize the tenant utility and adopts the following optimization function: , in, is the cumulative utility of tenant i, represents the benefit of resource allocation, is the resource application cost, dt represents the differential of time, and T is the optimization period.

[0009] Preferably, the resource allocation strategy is subject to the following constraints: The total resource allocation of a cluster cannot exceed its total available resources; The amount of resources requested by a single tenant must not exceed its maximum request capacity; Resource allocation must meet tenant service quality requirements and ensure priority allocation when resources are insufficient.

[0010] Preferably, the scheduling optimization module generates a resource allocation strategy through the Hamilton-Jacobi-Bellman equation, which is: , in, is the tenant’s value function, represents the benefit coefficient of resource use, represents the cost coefficient of resource application, is the current resource usage of tenant i, represents the rate at which the value function of tenant i changes over time, is the partial derivative of the value function with respect to the state variable, is the dynamic change of state variables, To optimize the resource application strategy, is the benefit item of resource allocation, The cost item requested for a resource.

[0011] Preferably, the feedback optimization module updates the resource allocation strategy through a deep Q network of reinforcement learning, and the update rule is: , in, is the value function of the current state and action, represents the state of the system k at the current moment, represents the action taken by the system at the current moment k, For instant rewards, is the discount factor, is the maximum estimate of future returns.

[0012] Preferably, the scheduling execution module includes the following functions: Dynamically adjust tenants' computing and memory resource allocations based on resource allocation policies; Redeploy containers across nodes to balance the computing load of the cluster; When resources are tight, the priority management mechanism is triggered to ensure resource allocation for high-priority tasks first; Provides resource recovery function, which can dynamically release redundant resources to improve the overall resource utilization of the cluster by identifying idle resources or resources occupied by low-priority tasks.

[0013] Preferably, the data acquisition module obtains the following information through real-time monitoring: The remaining amount of resources at each node; Resource usage of each tenant; The overall cluster load status and performance indicators.

[0014] Preferably, the container network optimization function of the scheduling execution module includes the following steps: Dynamically adjust data paths based on real-time container communication requirements and network topology to optimize network performance; Prevent unauthorized access by defining network security rules based on container namespaces and tenant isolation; Use bandwidth allocation optimization and priority scheduling methods to reduce network delays and congestion; By real-time monitoring of the latency and packet loss performance indicators of inter-container communication, network optimization strategies are automatically triggered.

[0015] Preferably, the present invention also provides a distributed management method for a cloud container cluster, comprising the following steps: Collect resource usage of each node in the cloud container cluster and tenants' resource demand information in real time; Based on the collected resource data, a state variable model is constructed, in which the dynamic changes of tenant resources depend on resource application behavior and task completion release; Generate a resource allocation strategy through a utility function optimization method, wherein the strategy is based on a benefit and cost balance and is subject to resource allocation constraints; Dynamically adjust the computing resource allocation, deployment location, and load balancing of containers based on the generated resource allocation strategy; Through reinforcement learning technology, the resource allocation strategy is updated according to the scheduling effect of real-time monitoring, and dynamically adapts to load changes.

[0016] The present invention provides a distributed management system for a cloud container cluster. It has the following beneficial effects: 1. The present invention achieves the technical effect of efficiently managing cloud container resources through modularly designed data collection, state modeling, scheduling optimization, scheduling execution and feedback optimization system architecture. Compared with the technical solution in the prior art that lacks clear coordination between modules, it solves the problems of low resource utilization and slow scheduling response speed;

[0017] 2. The present invention achieves the technical effect of flexibly adjusting the data collection frequency according to the system load by dynamically adjusting the collection frequency through Kubernetes Metrics Server or Prometheus API integration. Compared with the solution of fixed collection cycle in the prior art, it solves the problems of low collection efficiency and insufficient real-time performance;

[0018] 3. The present invention achieves the technical effect of accurately describing the changes in tenant resources by introducing the resource dynamic change equation in the state modeling module and adopting a comprehensive modeling method of task requirements, historical resource usage patterns and resource release rates. Compared with the models in the prior art that lack dynamic change descriptions, the present invention solves the problem that the resource allocation model cannot adapt to the dynamic environment.

[0019] 4. The present invention achieves the technical effect of generating a global optimal resource allocation strategy by using the Hamilton-Jacobi-Bellman equation for optimization and solution in the scheduling optimization module, combining the utility function of revenue and cost. Compared with the technical solution in the prior art that only relies on a simple heuristic algorithm, the present invention solves the problems of optimization target limitations and lack of globality.

[0020] 5. The present invention achieves the technical effects of resource allocation, container migration and dynamic load balancing by combining the Kubernetes API in the scheduling execution module. Compared with the solution of fixed resource allocation and slow load balancing response in the prior art, it solves the problems of low efficiency of system resource utilization and inability to guarantee service quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 This is a system architecture diagram of the present invention.

[0022] Figure 2 The figure is a flow chart of the method steps of the present invention. DETAILED DESCRIPTION

[0023] The technical solution of the present invention will be described clearly and completely below in conjunction with the accompanying drawings of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0024] Please see attached Figure 1 The embodiment of the present invention provides a distributed management system for cloud container clusters, which aims at the resource scheduling problem in the distributed management of cloud container clusters. It includes five modules: data collection, state modeling, scheduling optimization, scheduling execution, and feedback optimization. These modules are based on data collection, realize efficient resource allocation through mathematical models and optimization algorithms, and combine dynamic feedback mechanisms to ensure real-time optimization of scheduling strategies.

[0025] Data acquisition module The main function of the data acquisition module is to provide real-time operation data support for the state modeling module. Specifically, it includes: The current resource usage of each node, such as CPU, memory, and network bandwidth, to ensure the accuracy of global resource distribution. Resource demand data of each tenant, including the resource type and magnitude required for the task and the expected completion time of the task. The overall load of the cluster, including the average load of the node and the length of the task queue. As an option, the data collection module achieves seamless integration with the cloud container platform through a standardized interface (KubernetesMetricsServer). It should be noted that in order to ensure the real-time nature of the data, the collection period is usually set to 5 seconds, and this parameter can be dynamically adjusted according to the actual load conditions.

[0026] Specifically, when the load is low, the collection period can be extended to 10 seconds to reduce system overhead; when the load is high, the collection period can be shortened to 2 seconds to provide more accurate operation data in a timely manner.

[0027] It should be noted that the output of the data acquisition module not only serves the state modeling module, but can also be directly passed to the scheduling optimization module to quickly respond to tenants' resource needs under special circumstances.

[0028] Exemplarily, the collected data will be directly passed to the state modeling module to generate a mathematical model that describes the dynamic changes of resources, providing accurate basic data support for subsequent scheduling optimization.

[0029] State Modeling Module The state modeling module is used to describe the dynamic process of tenant resource application, occupation and release, providing an accurate mathematical basis for scheduling optimization.

[0030] In one implementation, the state modeling module builds a model based on the following dynamic change equation: ,

[0031] in, is the rate of change of tenant resource usage, is the resource usage of the ith tenant at time t, is the amount of resources requested by the ith tenant at time t, is the resource release rate of the ith tenant, is the amount of resources released by the i-th tenant due to task completion at time t.

[0032] It should be noted that the amount of resource application It can be expressed by the following formula: ,

[0033] in, : The current task requirement of tenant i, describing the resource requirement at time t, : Historical resource usage pattern of tenant i.

[0034] As an option, the state modeling module can adjust model parameters for different types of tenant tasks to more accurately describe resource change behavior. The resource release rate of high-priority real-time video processing tasks is usually set higher so that resources can be released quickly after the task is completed, while the resource release rate of low-priority batch processing tasks can be set lower to reduce frequent resource reallocation operations.

[0035] High-priority task modeling: For tasks with high real-time requirements, the amount of resource application The changing trend of can be described by the following formula: .

[0036] Low-priority task modeling: For delayed tasks, the resource request calculation formula can introduce a lag factor ; ,

[0037] in, represents the cumulative delay of tenant i’s high-priority task at time t, represents the unit resource efficiency of tenant i in executing high-priority tasks, To ensure the resource application Not a negative value, represents the maximum delay that tenant i can tolerate for low-priority tasks.

[0038] For example, for mixed task scenarios with more complex task types, the state modeling module can build a hierarchical model in combination with multi-task characteristics, classify and model the tasks by priority, and finally generate an overall resource change description.

[0039] Scheduling Optimization Module The scheduling optimization module generates the optimal resource allocation strategy through the utility function and constraints. Specifically: Utility function construction: In one possible implementation, the tenant's utility function is defined as follows: , The goal is to maximize the total utility of tenants while satisfying the following constraints: ,

[0040] The total amount of cluster resources cannot exceed its available capacity.

[0041] ,

[0042] in, is the cumulative utility of tenant i, represents the benefit of resource allocation, is the resource application cost, dt represents the differential of time, T is the optimization cycle, It represents the total resource usage of all n tenants in the system at time t, and n represents the total number of tenants participating in resource competition in the system. represents the amount of resources allocated by the system to tenant i at time t, Indicates the total available resources of the system. Indicates the resource application amount of tenant i It cannot be a negative value, indicating that the resource application amount of tenant i at time t cannot exceed its maximum application capacity , Indicates the maximum amount of resources that tenant i can apply for at a single point in time.

[0043] The amount of resources requested by a single tenant must not exceed its maximum request capacity.

[0044] The scheduling optimization module uses the Hamilton-Jacobi-Bellman equation to solve the resource allocation strategy: , in, is the tenant’s value function, represents the benefit coefficient of resource use, represents the cost coefficient of resource application, is the current resource usage of tenant i, represents the rate at which the value function of tenant i changes over time, is the partial derivative of the value function with respect to the state variable, is the dynamic change of state variables, To optimize the resource application strategy, is the benefit item of resource allocation, The cost item requested for a resource.

[0045] As an option, the scheduling optimization module can adjust the optimization goal based on the resource allocation requirements of different scenarios. For example, in the edge computing scenario, the optimization goal of the resource allocation strategy can be changed from maximizing tenant utility to minimizing task delay. Specifically: Delay optimization goal: In delay-sensitive scenarios, the optimization goal can be defined as: ,

[0046] Among them, min means that the optimization goal is to minimize the value of the objective function. represents the cumulative execution delay of tasks of tenant i under the current resource allocation strategy, Indicates the contribution of the unit resource allocated to tenant i to the delay reduction, Represents the latency penalty caused by insufficient resource allocation.

[0047] Energy efficiency optimization goal: For energy-sensitive scenarios, the scheduling optimization module can add energy cost items to the utility function: ,

[0048] in, represents the total utility of tenant i in the time interval, represents the energy cost coefficient, is the energy consumption of task execution, It represents the energy consumption cost of tenant i using resources at time T.

[0049] Exemplarily, the scheduling optimization module may use a reinforcement learning algorithm instead of a traditional dynamic programming method to achieve more efficient resource allocation optimization in a high-dimensional state space.

[0050] It should be noted that no matter what optimization goal is adopted, the resource allocation strategy finally generated must meet the total resource constraints of the cluster and the service quality requirements of the tenants.

[0051] Scheduling execution module The scheduling execution module allocates resources and implements specific operations according to the optimization strategy. Its functions include: Dynamically allocate CPU and memory resources to tenants, and update resource limits in real time through the Kubernetes quota mechanism; When the node load is unbalanced, container migration is triggered to schedule tasks from high-load nodes to low-load nodes; Prioritize the allocation of resources to high-priority tasks, and recycle resources from low-priority tasks when resources are insufficient.

[0052] The scheduling execution module is integrated with the Kubernetes API to automate resource allocation and container migration through the scheduler interface.

[0053] As an option, the scheduling execution module can be combined with service grid technology to achieve dynamic load balancing between containers. The introduction of service grid can significantly improve the fault tolerance and communication efficiency of the system.

[0054] Specifically: Through the service mesh control plane, traffic distribution is dynamically adjusted according to the resource usage of each container to avoid performance degradation due to overload of a single container.

[0055] Utilize the data plane of the service grid to encrypt and isolate the communication data of different tenants to ensure security in a multi-tenant environment.

[0056] For example, when migrating containers between nodes, the scheduling execution module can combine the traffic routing capabilities of the service grid to achieve seamless migration of tasks and avoid interruptions during the migration process.

[0057] It should be noted that the functional expansion of the scheduling execution module does not affect the implementation of the core resource allocation strategy, but provides the system with higher reliability and flexibility.

[0058] Feedback Optimization Module The feedback optimization module dynamically adjusts the resource allocation strategy through reinforcement learning, which is implemented as follows: the current system status includes the resource allocation amount, task completion time and the overall cluster load.

[0059] Instant Rewards Based on resource utilization and task delay calculation, the goal is to maximize resource utilization efficiency and minimize task delay.

[0060] Deep Q network update: Optimize the Q network parameters through the following update formula: ,

[0061] in, is the value function of the current state and action, represents the state of the system k at the current moment, represents the action taken by the system at the current moment k, is the discount factor, is the maximum estimate of future returns, For instant rewards, resource utilization, task completion time and service quality are comprehensively considered, and the specific definition is: ,

[0062] in, represents the resource utilization efficiency of the system at time step k, represents the time it takes for a task to arrive at the system and be completed at time step k, represents the quality loss of the system at time step k due to failure to meet the tenant service level agreement, , , : Weight parameter, used to balance the priorities of different goals.

[0063] As an option, the feedback optimization module can dynamically adjust the resource allocation ratio between edge nodes and central nodes based on the special needs of edge computing scenarios.

[0064] Specifically: Center-edge collaborative optimization: The feedback optimization module can dynamically adjust the resource allocation strategy of edge nodes according to the geographical distribution of tasks and data processing requirements. For example, for tasks that require local processing, edge node resources are allocated first; for tasks that require cross-node collaboration, central node resources are allocated first.

[0065] Dynamic training reinforcement learning model: The feedback optimization module can use distributed reinforcement learning methods to introduce real-time data from edge nodes into the training process, thereby improving the generalization ability of the Q network.

[0066] Exemplarily, when processing delay-sensitive tasks, the feedback optimization module can combine the prediction capability of the distributed Q network to adjust the resource allocation strategy in advance to cope with sudden loads.

[0067] It is understandable that the design of the feedback optimization module is not only suitable for centralized cloud computing scenarios, but also can meet the needs of distributed and edge computing scenarios.

[0068] It should be noted that the feedback optimization module works together with the scheduling optimization module to dynamically update the resource allocation strategy to ensure that the system maintains the optimal state during load changes.

[0069] Please see attached Figure 2 The present invention also provides a distributed management method for a cloud container cluster, which combines multiple technical means such as dynamic resource monitoring, mathematical modeling, optimization algorithms, and reinforcement learning to ensure that resources in the cloud container cluster are securely shared and efficiently managed. The following will describe in detail the various steps of the method and its implementation, which can be used to achieve efficient resource utilization and service quality optimization in the dynamic allocation of cloud computing resources and task scheduling.

[0070] S1. Data collection provides real-time monitoring of cloud container cluster resources In this step, the data collection process can use Kubernetes Metrics Server and Prometheus API to obtain the usage status of node resources in real time, including key indicators such as CPU, memory, and network bandwidth. It should be noted that the real-time nature of data collection is controlled by the collection cycle, which is set to 5 seconds by default, and the specific interval can be dynamically adjusted according to the system load.

[0071] Specifically, the collection of node resource usage includes the current number of remaining CPU cores, available memory capacity, and network bandwidth usage. The tenant's resource demand information includes the resource type and resource demand of the currently running task, as well as the task priority and expected completion time.

[0072] In some embodiments, the collected tenant demand information will be updated in real time to the system's state modeling module to ensure that tenant resource demand fluctuations can be captured in a timely manner. It should be noted that this process ensures that the input data for scheduling optimization is always consistent with the actual operating status of the system.

[0073] S2. State modeling provides cloud container dynamic resource state modeling In this step, when constructing the state variable model, the dynamic characteristics of resource application behavior and task completion release are combined. As a possible implementation method, the state variable model describes the mapping relationship between the tenant's resource usage and its resource demand.

[0074] Specifically: Resource application behavior reflects the resource usage needs of tenants, which are affected by task type, runtime and priority. For example, tasks with strong real-time characteristics will show a higher frequency of resource application, while resource application for non-real-time tasks is more predictable. Task completion release describes the process of releasing occupied resources after the tenant completes the task. As an option, the release process can dynamically adjust the release rate according to the task completion rate. For example, for high-priority tasks, the resource release rate is higher in order to quickly free up resources for other tasks. It should be noted that in the process of constructing the state variable model, the historical resource usage pattern of the tenant is combined to further improve the accuracy of the prediction of future resource demand. Exemplarily, the output of the model is the dynamically changing data of resource usage, which is directly used by subsequent optimization modules.

[0075] S3. Generate resource allocation strategy In this step, resource allocation strategy generation is based on the utility function optimization method. The relevant formulas and algorithm logic have been described in detail in the system part and will not be repeated here. The generated resource allocation strategy can be further refined into multi-layer strategies, such as node allocation strategy and container allocation strategy. At the same time, the formulation of the strategy can be based on a multi-objective optimization model, taking into account the resource allocation utility and load balancing goals. In edge computing or multi-cloud environments, resource allocation strategies can be extended to cross-regional or cross-platform resource coordination to ensure unified scheduling in multiple scenarios.

[0076] S4: Dynamically adjust container computing resource allocation, deployment location, and load balancing In this step, the resource allocation and deployment adjustment operations of the container have been described in detail in the system section and will not be repeated here. When adjusting the deployment location, the node with the lowest latency can be selected to deploy computing-intensive tasks, and storage-intensive tasks can be deployed to high-storage nodes first, based on the resource sensitivity of the task and the network latency requirements. In some embodiments, the service grid technology can be combined to optimize the routing of traffic between containers, thereby reducing the communication overhead of task execution.

[0077] S5. Dynamically update resource allocation strategy through reinforcement learning technology In this step, the dynamic update of the resource allocation strategy through reinforcement learning technology has been described in detail in the system part and will not be repeated here. The reinforcement learning module can be combined with the historical data analysis module to enhance the initial performance of the model through offline learning, and adapt to real-time changes in the environment through online learning. In some embodiments, the input state of reinforcement learning can be expanded to a multi-dimensional state space, for example, including more dynamic parameters such as network bandwidth, task priority, container operating environment, etc., thereby further improving the intelligent level of scheduling.

[0078] The present invention provides a distributed management method for cloud container clusters, which realizes efficient resource scheduling management by collecting resource usage and demand information in real time, building a dynamic state model, optimizing and generating resource allocation strategies based on utility functions, dynamically adjusting container resource allocation and deployment locations, and updating scheduling strategies using reinforcement learning technology. Each step cooperates with each other to ensure the real-time and high efficiency of resource allocation in a dynamic load change environment, effectively improve resource utilization and service quality, and enhance the system's adaptability and optimization performance.

[0079] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A distributed management system for a cloud container cluster, characterized in that: include; The data collection module is used to collect the remaining resources of each node in the cloud container cluster and the resource requirements of tenants, and provide real-time data update support; A state modeling module, used to establish a state variable model describing the dynamic changes of tenant resources based on the resource data collected by the data collection module, wherein the state variables include the tenant's resource occupancy, resource application amount and resource release rate; The scheduling optimization module is used to generate resource allocation strategies through optimization algorithms based on the dynamic change characteristics of tenant resources and the competition relationship among multiple tenants; A scheduling execution module, used to dynamically adjust the allocation of cloud container resources according to the resource allocation strategy generated by the scheduling optimization module, including allocating computing resources and adjusting container deployment between nodes; The feedback optimization module is used to monitor the scheduling execution effect in real time and adjust the optimization strategy according to the operating status of resource allocation and scheduling results.

2. A distributed management system for a cloud container cluster according to claim 1, characterized in that: The state variable model established by the state modeling module describes the tenant resource changes with a dynamic equation, that is, the tenant's resource occupancy changes over time, which is jointly determined by the tenant's resource application amount and release rate. The model includes the correlation between the tenant's resource occupancy and resource request behavior.

3. A distributed management system for a cloud container cluster according to claim 1, characterized in that: The core function of the scheduling optimization module is to generate resource allocation strategies based on the data and utility function provided by the state modeling module. The scheduling optimization module aims to maximize tenant utility and adopts the following optimization function: ， in, is the cumulative utility of tenant i, represents the benefit of resource allocation, is the resource application cost, dt represents the differential of time, and T is the optimization period.

4. A distributed management system for a cloud container cluster according to claim 3, characterized in that: The resource allocation strategy is subject to the following constraints: The total resource allocation of a cluster cannot exceed its total available resources; The amount of resources requested by a single tenant must not exceed its maximum request capacity; Resource allocation must meet tenant service quality requirements and ensure priority allocation when resources are insufficient.

5. A distributed management system for a cloud container cluster according to claim 1, characterized in that: The scheduling optimization module generates a resource allocation strategy through the Hamilton-Jacobi-Bellman equation, which is: , in, is the tenant’s value function, represents the benefit coefficient of resource use, represents the cost coefficient of resource application, is the current resource usage of tenant i, represents the rate at which the value function of tenant i changes over time, is the partial derivative of the value function with respect to the state variable, is the dynamic change of state variables, To optimize the resource application strategy, is the benefit item of resource allocation, The cost item requested for a resource.

6. A distributed management system for a cloud container cluster according to claim 1, characterized in that: The feedback optimization module updates the resource allocation strategy through the deep Q network of reinforcement learning, and the update rule is: ， in, is the value function of the current state and action, represents the state of the system k at the current moment, represents the action taken by the system at the current moment k, For instant rewards, is the discount factor, is the maximum estimate of future returns.

7. A distributed management system for a cloud container cluster according to claim 1, characterized in that: The scheduling execution module includes the following functions: Dynamically adjust tenants' computing and memory resource allocations based on resource allocation policies; Redeploy containers across nodes to balance the computing load of the cluster; When resources are tight, the priority management mechanism is triggered to ensure resource allocation for high-priority tasks first; Provides resource recovery function, which can dynamically release redundant resources to improve the overall resource utilization of the cluster by identifying idle resources or resources occupied by low-priority tasks.

8. A distributed management system for a cloud container cluster according to claim 1, characterized in that: The data acquisition module obtains the following information through real-time monitoring: The remaining amount of resources at each node; Resource usage of each tenant; The overall cluster load status and performance indicators.

9. A distributed management system for a cloud container cluster according to claim 1, characterized in that: The container network optimization function of the scheduling execution module includes the following steps: Dynamically adjust data paths based on real-time container communication requirements and network topology to optimize network performance; Prevent unauthorized access by defining network security rules based on container namespaces and tenant isolation; Use bandwidth allocation optimization and priority scheduling methods to reduce network delays and congestion; By real-time monitoring of the latency and packet loss performance indicators of inter-container communication, network optimization strategies are automatically triggered.

10. A distributed management method for a cloud container cluster, applied to the system according to any one of claims 1 to 9, characterized in that: The following steps are involved: Collect resource usage of each node in the cloud container cluster and tenants' resource demand information in real time; Based on the collected resource data, a state variable model is constructed, in which the dynamic changes of tenant resources depend on resource application behavior and task completion release; Generate a resource allocation strategy through a utility function optimization method, wherein the strategy is based on a benefit and cost balance and is subject to resource allocation constraints; Dynamically adjust the computing resource allocation, deployment location, and load balancing of containers based on the generated resource allocation strategy; Through reinforcement learning technology, the resource allocation strategy is updated according to the scheduling effect of real-time monitoring, and dynamically adapts to load changes.

Images