Method and system for self-adaptive elastic expansion strategy of Android distributed edge container

Abstract

The invention relates to a method and a system for an adaptive elastic expansion strategy of an Android distributed edge container. Monitoring the resource use condition of each Android container in real time; wherein the resource use condition of the Android container is determined; according to the resource use condition of each Android container obtained by the resource monitoring module, a preset calculation resource is dynamically allocated to each preset Android container by applying a preset algorithm; and according to different service flow characteristics and specifications of resources required by services, adjusting preset rearranged available container resources, or calling elastic resources to carry out resource expansion. The resources are flexibly allocated according to actual resource requirements and business load conditions of the Android containers, the problem of unbalanced resource allocation and utilization is solved, container expansion can be quickly and flexibly performed when the business volume is increased, manual intervention is not needed, the response time is shortened, the quick increase requirement of the business for the container resources is met in time, and the business efficiency is improved. And the service continuity is ensured.

Description

Technical Field

[0001] The present application belongs to the field of edge computing and Android system technology, and in particular, to a method and system for an Android distributed edge container adaptive elastic expansion strategy. Background Art

[0002] With the rapid development of Internet technology, cloud computing, virtualization technology and broadband networks, cloud gaming business has emerged. Cloud gaming allows games to be run on remote servers and transmitted to user terminals in real time. The use of ARM architecture servers can effectively simulate the operating environment of mobile devices, while reducing the hardware cost required for users to enjoy high-quality gaming experience, while maintaining low operating costs and energy consumption.

[0003] However, the current Android distributed edge container technology in cloud gaming scenarios still faces many challenges. In terms of resource allocation, static allocation is often used. For example, when the system is initialized, resources such as CPU and memory are allocated to each Android container according to a preset fixed ratio. In the subsequent operation process, the actual business load of the container is completely ignored, and the resource allocation remains unchanged. This static mode is very likely to cause some container resources to be idle while some container resources are tight when the business load is unbalanced, greatly reducing the overall performance of the system.

[0004] In addition, when business volume fluctuates and containers need to be expanded, the existing expansion mechanism has obvious defects. For example, the manually expanded Android edge container system relies on manual operation when the business volume increases. Operation and maintenance personnel need to manually allocate new resource areas to new containers or adjust existing container resources according to business needs. This method is extremely inefficient. It not only consumes a lot of time and manpower, but also easily causes configuration errors due to human operation errors, which has a serious adverse impact on the normal operation of the system and business continuity. More efficient and intelligent solutions are urgently needed to overcome these problems in order to promote the further development and optimization of Android distributed edge container technology in cloud gaming business. Summary of the invention

[0005] In view of this, the purpose of the present invention is to provide a method and system for an Android distributed edge container adaptive elastic expansion strategy to solve the technical problems existing in the prior art. According to a first aspect of an embodiment of the present invention, a method for adaptive elastic expansion strategy of an Android distributed edge container is provided, comprising: Real-time monitoring of resource usage of each Android container; wherein the resource usage of the Android container includes: CPU utilization, memory usage, and business load status; Using the resource usage of each Android container obtained by the resource monitoring module, the preset computing resources are dynamically allocated to each preset Android container using a preset algorithm; According to different business traffic characteristics and the specifications of the resources required by the business, the preset available container resources are re-arranged and adjusted, or elastic resources are called for resource expansion.

[0006] Furthermore, the real-time monitoring of resource usage of each Android container includes: By interacting with the preset board hardware interface and the preset Android container management interface, the resource usage of each Android container is monitored in real time.

[0007] Furthermore, the resource usage of each Android container obtained by the resource monitoring module is used to dynamically allocate the preset computing resources to each preset Android container using a preset algorithm, including: Using the resource usage of each Android container obtained by the resource monitoring module, container nodes are allocated through a preset load balancing strategy to obtain a first processing result; Using the first processing result and combining the business resource demand characteristics, the preset computing resources are dynamically allocated to the preset Android containers.

[0008] Furthermore, the steps of implementing the re-arrangement of available container resources include: Obtain historical statistical data for real-time monitoring of resource usage of each Android container within a preset period of time, and mark the business type or application name of the Android container; Removing outliers from the historical statistical data, filling missing values ​​using a preset algorithm, and converting the data into a time series format suitable for modeling using a preset time series conversion algorithm to obtain first processed data; Input the first processed data into a preset trained processing model to output the demand for Android containers of different specifications in a future period of time; wherein the demand for Android containers includes: the number of CPU cores and the memory size; According to the demand for Android containers of different specifications in the future period and the preset allocation strategy, a resource allocation plan is obtained; According to the resource allocation plan, using the demand for Android containers of different specifications in the future period and the preset allocation strategy, dynamically create a required number of new Android containers, and configure the newly created new Android containers with appropriate hardware specifications; If there are containers that are running but no longer meet the above allocation strategy, they are smoothly transferred to other locations through hot migration technology.

[0009] Furthermore, the method further comprises: When a business request ends, the associated Android container resources are recovered, temporary files are deleted, and the container status table entry in the database is updated synchronously.

[0010] Furthermore, the preset re-arrangement of available container resources is adjusted according to different business traffic characteristics and the specifications of resources required by the business, or elastic resources are called for resource expansion, including: Obtain historical data of the load of each Android container, and use a preset machine learning algorithm to predict the business traffic trend in the future period based on the historical data of the load of each Android container; wherein the business traffic trend in the future period includes: a peak time period and a trough time period; The peak time period, the trough time period, and the preset business characteristics are used to predict the peak, and the capacity expansion process is started according to the predicted result of the peak.

[0011] Furthermore, the peak time period, the trough time period, and the preset service characteristics predict the peak, and start the expansion process according to the predicted result of the peak, including: According to the predicted result of the wave peak and the currently available fixed resources, if the currently available fixed resources are sufficient to cope with the peak traffic predicted according to the predicted result of the wave peak, the hardware specifications of the re-arranged available container resources are adjusted, and a new Android container is created according to business needs or the resources of the existing Android container are expanded; If the currently available fixed resources are insufficient to cope with the peak traffic predicted based on the peak prediction results, the elastic resource pool is called according to business needs.

[0012] According to a second aspect of an embodiment of the present invention, a system for an Android distributed edge container adaptive elastic expansion strategy is provided, which is applied to any of the above methods for an Android distributed edge container adaptive elastic expansion strategy, including: The resource monitoring module is used to monitor the resource usage of each Android container in real time; wherein the resource usage of the Android container includes: CPU utilization, memory usage, and business load status; A resource allocation module, used to utilize the resource usage of each Android container obtained by the resource monitoring module, and dynamically allocate the preset computing resources to each preset Android container using a preset algorithm; The container expansion module is used to adjust the preset re-arrangement of available container resources or call elastic resources for resource expansion according to different business traffic characteristics and the specifications of resources required by the business.

[0013] The technical solution provided by the embodiments of the present invention may have the following beneficial effects: By utilizing the resource usage of each Android container obtained by the resource monitoring module, the preset computing resources are dynamically allocated to each preset Android container using a preset algorithm, thereby realizing dynamic allocation of Android container computing resources, flexibly allocating resources according to the actual resource requirements and business load of each Android container, solving the problem of unbalanced resource allocation and utilization, and adjusting the preset re-arrangement of available container resources or calling elastic resources for resource expansion according to the specifications of resources required by the business according to different business traffic characteristics. Through the dynamic resource allocation mechanism, the computing resources of the 865 board can be dynamically allocated according to the actual business load, avoiding idle resources and resource shortages, significantly improving resource utilization efficiency, and improving the overall performance of the system; when the business volume increases, the container can be expanded quickly and flexibly without manual intervention, shortening the response time, and promptly meeting the rapidly growing business demand for container resources, ensuring business continuity.

[0014] It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0016] Figure 1 It is a schematic diagram of the steps of a method for adaptive elastic expansion strategy of an Android distributed edge container according to an exemplary embodiment; Figure 2 It is a schematic diagram of the system composition of an Android distributed edge container adaptive elastic expansion strategy according to an exemplary embodiment. DETAILED DESCRIPTION

[0017] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. Instead, they are merely examples of devices and methods consistent with some aspects of the present invention as detailed in the appended claims.

[0018] Embodiment 1 See also Figure 1 , Figure 1 The present invention is a schematic diagram of steps of a method for adaptive elastic expansion strategy of an Android distributed edge container according to an exemplary embodiment, the method comprising: S1. Real-time monitoring of resource usage of each Android container; wherein the resource usage of the Android container includes: CPU utilization, memory usage, and business load status; S2. Using the resource usage of each Android container obtained by the resource monitoring module, the preset algorithm is used to dynamically allocate the preset computing resources to each preset Android container; S3. According to different business traffic characteristics and the specifications of the resources required by the business, adjust the preset re-arrangement of available container resources, or call elastic resources for resource expansion.

[0019] In a specific implementation, as described in step S1, the real-time monitoring of resource usage of each Android container includes: interacting with a preset board hardware interface and a preset Android container management interface to monitor the resource usage of each Android container in real time.

[0020] During the specific implementation, the resource usage (including CPU utilization, memory usage, etc.) and business load status (such as number of requests, processing time, etc.) of each Android container are monitored in real time to provide data support for subsequent resource allocation and container expansion decisions.

[0021] After the system is started, the resource monitoring module starts working and continuously collects resource usage data and business load data of each Android container.

[0022] By interacting with the board hardware interface and the Android container management interface, it obtains accurate information such as CPU utilization, memory usage, number of requests, processing time, etc., and transmits this information to the resource allocation module and the adaptive adjustment module in real time.

[0023] In the specific implementation, as described in step S2, the resource usage of each Android container obtained by the resource monitoring module is used to dynamically allocate the preset computing resources to each preset Android container using a preset algorithm, including: Based on the information fed back by the resource monitoring module, intelligent algorithms (such as load balancing algorithms, priority algorithms, etc.) are used to dynamically allocate computing resources to ensure the reasonable allocation of resources among various Android containers and maximize resource utilization.

[0024] More specifically, the implementation steps are as follows: S2.1. Receiving monitoring information Receive real-time monitoring information, continuously track and collect the operating status data of each node in the system (such as x86 servers, ARM boards), including but not limited to key performance indicators such as CPU usage, memory usage, disk I / O speed, etc. These data are important basis for resource allocation decisions.

[0025] S2.2. Applying the preset algorithm After receiving the latest monitoring information, the data is processed according to the pre-set resource allocation algorithm. The commonly used strategy is based on load balancing, that is, to ensure that the load between different nodes is distributed as evenly as possible to avoid the situation where some nodes are overloaded while other nodes are idle.

[0026] S2.3. Analyze resource demand characteristics When allocating resources, we not only rely on current performance data, but also consider the specific business resource demand characteristics. We need to understand the specific resource requirements of each application or service. For example, a popular mobile game may have high requirements for graphics processing capabilities, or other applications may focus more on network bandwidth rather than computing resources. Through in-depth analysis of such characteristics, we can allocate the most appropriate physical resources for different types of applications.

[0027] S2.4. Determine container configuration Finally, based on the results of the above two steps, we will decide how to select the appropriate hardware specifications for each Android container to be created. This involves determining the number of CPU cores and memory size required for each container to ensure that they can meet the performance requirements of their respective applications while making full use of existing resources, reducing waste and improving overall efficiency.

[0028] In practice, when a new group of gamers are about to enter the cloud gaming platform, the above resource allocation method is used to carry out a series of key operations in an orderly manner. The first step is to conduct a detailed assessment of the current situation, and use the powerful monitoring function of the preset resource monitoring module to accurately and comprehensively grasp the load details of each node. This includes important information such as the CPU usage rate, memory usage, network bandwidth consumption, and GPU busyness of each node, thereby providing a detailed and reliable data basis for subsequent decision-making.

[0029] Next, we enter the key link of selecting the best solution. In this process, advanced and intelligent load balancing algorithms are cleverly used to carefully screen out nodes with relatively low loads. At the same time, the differences in game types that new players are keen on are fully taken into account, because different game types have very different requirements for resources. For example, some large 3A games or games with extremely high requirements for picture rendering often require more powerful GPU resources to ensure their exquisite picture quality and smooth operation; while some strategy or text games may have more prominent requirements for CPU and memory resources. By comprehensively weighing these factors, ensure that the most suitable node resources are selected for new players.

[0030] The last step is to carefully configure the container parameters. According to the specific needs of each new player, an Android container of the right size is allocated to them. When determining the container parameters, we should not only focus on satisfying the new players to be able to start the game journey smoothly and enjoy the game experience without any lag, clear picture quality and responsiveness, but also take into account the efficient and stable operation of the entire cloud gaming system from a macro perspective. By reasonably allocating resources, we can avoid system performance bottlenecks or failures caused by excessive load on individual nodes, ensure that the entire cloud gaming platform can still maintain the best operating state after accepting new players, and continue to provide high-quality cloud gaming services to the majority of players.

[0031] It should be noted that for container resources released after the business request is completed, the system will rearrange the available container resources based on the demand ratio of container resources of different specifications predicted based on historical statistical data.

[0032] In the specific implementation, the specific steps for re-arranging available container resources are as follows: 1. Data collection and preprocessing 1.1 Collecting Historical Data Data source: Obtain usage records of all containers in the past period of time (such as one week or one month) from the preset resource monitoring module, including but not limited to performance indicators such as CPU utilization, memory usage, and disk I / O speed.

[0033] Labeled data: For each container, mark its corresponding business type or application name so that it can be classified according to specific business characteristics during subsequent analysis.

[0034] 1.2 Data cleaning and organization Outlier removal: Identify and remove data points that significantly deviate from the normal range to avoid negatively affecting model training.

[0035] Fill missing values: Use reasonable strategies to fill in any data gaps that may exist, such as using forward filling, backward filling, or interpolation.

[0036] -Time series conversion: Convert the original timestamp data into a time series format suitable for modeling, facilitating subsequent trend analysis and periodic pattern mining.

[0037] 2. Demand forecasting model construction Feature Engineering 2.1 Extract features: Based on the sorted data, extract key features that help predict future demand, such as weekdays / weekends, time periods, and influencing factors of specific holidays.

[0038] 2.2 Create new features: Generate new derivative features based on business logic, such as average response time, number of concurrent users, etc. This information can help capture business fluctuations more accurately.

[0039] 2.3 Model Selection and Training Choose the right algorithm: Considering the characteristics of cloud gaming services, you can choose machine learning or deep learning algorithms such as support vector machine (SVM), random forest, and long short-term memory network (LSTM) to build a prediction model.

[0040] Divide the training set and test set: Split the data set according to a certain ratio (such as 70% for training and 30% for validation), and ensure the time sequence consistency of the two parts of data.

[0041] - Model training and optimization: Use the training set to repeatedly adjust the model parameters until the best performance is achieved; evaluate the performance of the model on the test set through methods such as cross-validation to ensure generalization ability.

[0042] 3. Resource allocation decision Analyze the prediction results Determine demand ratio: Based on the output of the forecasting model, calculate the approximate demand for containers of different specifications (CPU cores, memory size) in the future. This step requires comprehensive consideration of multiple factors, such as business growth trends, seasonal changes, etc.

[0043] - Formulate a resource allocation strategy: Based on the total amount of physical resources that can be allocated and the status of existing containers (idle, busy), plan an optimal resource allocation plan. This plan should not only meet current and future business needs, but also minimize the waste caused by idle resources.

[0044] 4. Re-arrange container resources Release unused containers Clean up useless containers: When a business request ends, immediately recycle the associated container resources, disconnect the network connection, stop the process, and delete temporary files.

[0045] Update status information: Synchronously update the container status table entries in the database to mark which containers have become "idle" and can be rescheduled.

[0046] Creating or resizing a new container Create containers on demand: Dynamically create the required number of new containers based on the demand ratio determined previously, and configure them with appropriate hardware specifications (number of CPU cores, memory size).

[0047] Migrate existing containers: For old containers that are still running but no longer meet the current ratio requirements, hot migration technology can be used to smoothly transfer them to a more suitable location, while adjusting their resource configuration to adapt to the new load level.

[0048] In the specific implementation, as described in step S3, the preset re-arrangement of available container resources is adjusted according to different business traffic characteristics and the specifications of the resources required by the business, or the elastic resources are called for resource expansion, including: Obtain historical data of the load of each Android container, and use a preset machine learning algorithm to predict the business traffic trend in the future period based on the historical data of the load of each Android container; wherein the business traffic trend in the future period includes: a peak time period and a trough time period; The peak time period, the trough time period, and the preset business characteristics are used to predict the peak, and the capacity expansion process is started according to the predicted result of the peak.

[0049] Furthermore, the peak time period, the trough time period, and the preset service characteristics predict the peak, and start the expansion process according to the predicted result of the peak, including: According to the predicted result of the wave peak and the currently available fixed resources, if the currently available fixed resources are sufficient to cope with the peak traffic predicted according to the predicted result of the wave peak, the hardware specifications of the re-arranged available container resources are adjusted, and a new Android container is created according to business needs or the resources of the existing Android container are expanded; If the currently available fixed resources are insufficient to cope with the peak traffic predicted based on the peak prediction results, the elastic resource pool is called according to business needs.

[0050] In the specific implementation, the system continuously tracks the load of each node through the resource monitoring module, including key performance indicators such as CPU utilization and memory usage.

[0051] Afterwards, based on the historical data accumulated over a long period of time, machine learning algorithms (such as time series analysis, regression models, etc.) are used to predict the business traffic trend in the future, especially to identify the specific time periods of peaks (high demand periods) and troughs (low demand periods).

[0052] Based on the characteristics of different businesses (such as the differences in resource requirements for different types of games), the forecast results are further refined to develop personalized resource allocation plans for each specific business.

[0053] When the system predicts an upcoming business peak, it will start the capacity expansion process in advance to ensure that there are sufficient resources to meet the upcoming traffic growth. According to the predicted demand scale, the system will re-arrange the available container resources through detection, and use the results of the detection and re-arrangement of available container resources to accurately adjust the hardware specifications of the container (such as the number of CPU cores and memory size), including: creating an appropriate number of new containers or adjusting the resource configuration of existing containers to meet the expected workload.

[0054] If the currently available fixed resources are insufficient to cope with the expected peak traffic, the system will automatically obtain additional computing power from a pre-prepared elastic resource pool, which can be temporarily rented cloud instances or other forms of spare capacity.

[0055] In the specific implementation, by utilizing the resource usage of each Android container obtained by the resource monitoring module, the preset computing resources are dynamically allocated to each preset Android container using a preset algorithm, thereby realizing dynamic allocation of Android container computing resources, and flexibly allocating resources according to the actual resource requirements and business load of each Android container, solving the problem of unbalanced resource allocation and utilization, and adjusting the preset re-arrangement of available container resources according to the specifications of resources required by the business according to different business traffic characteristics, or calling elastic resources for resource expansion. Through the dynamic resource allocation mechanism, the computing resources of the 865 board can be dynamically allocated according to the actual business load, avoiding idle resources and resource shortages, significantly improving resource utilization efficiency, and improving the overall performance of the system; when the business volume increases, the container can be expanded quickly and flexibly without manual intervention, shortening the response time, and promptly meeting the rapidly growing business demand for container resources, ensuring business continuity.

[0056] In one embodiment, it also includes: if the expanded resources use the elastic resource pool, during the business downturn period, the system will unsubscribe offline idle resources according to a preset strategy.

[0057] Even after capacity expansion, the system will continue to closely monitor changes in actual business traffic to ensure that allocated resources always match actual demand.

[0058] Specifically, valley period identification: Once it is detected that the business traffic has entered a valley period, that is, user activity has significantly decreased, the system begins to consider whether resources need to be reduced to save costs.

[0059] Preset policy execution: According to the pre-set policy, the system decides which resources can be safely released. This usually involves a series of conditional judgments, such as a container has almost no active tasks in the past X minutes, or the overall system load level has fallen below a certain threshold.

[0060] Orderly offline: For idle resources that are determined to be released, the system will take a gradual approach to gradually reduce their load until they completely stop the service. For example, new requests will be directed to other active containers first, then the network connection will be disconnected, and finally the container and all its associated resources will be formally deleted.

[0061] Protection measures: To prevent the accidental deletion of resources in use, the system may set a certain delay period, during which it will continue to observe whether there are new tasks. The unsubscription operation will only be finally executed when it is confirmed that there is no activity.

[0062] In one embodiment, in the operation and management process of a large-scale cloud gaming platform, the scientific planning and effective implementation of resource scheduling strategies, like the core gears in a precision instrument, play a decisive and key role in ensuring the efficient and stable operation of the platform and achieving precise cost control.

[0063] Take the response measures for the daily peak of player activity as an example. When the time quietly approaches 7 pm, the key node when the player activity is about to rise rapidly, around 6 pm, the resource scheduling system will start the action mode quickly and methodically like a well-trained commander. It will first dig deep and carefully analyze the massive historical data resources. These data are like an information library containing endless treasures, which records in detail the change curve of the number of players' logins in various time periods in the past, the distribution of game time, the peak and average of resource consumption generated by different game types, and other multi-dimensional and valuable information. With the help of in-depth analysis of these rich data, the system can accurately understand the traffic fluctuation rules during the peak period of business and the resource demand characteristics that closely correspond to it. At the same time, the system will also closely combine real-time monitoring information, just like a pair of sharp eyes for decision-making, always keeping a close eye on the key indicators of the current platform, including the load status of the server, the real-time occupancy rate of the network bandwidth, and the online density of players in each game area. Through the comprehensive integration and precise analysis of historical data and real-time monitoring information, the system can predict the scale and resource demand intensity of the upcoming business peak with extremely high accuracy, and then start the expansion process in advance. During the expansion process, the system will intelligently allocate various resources, such as increasing the number of servers, and reasonably allocating new servers to different game areas or business modules based on the predicted player distribution; dynamically adjust the configuration parameters of the server, optimize the allocation strategy of core resources such as CPU, memory, GPU, etc., to ensure that during the peak period of player activity, every player can enjoy a smooth, non-stuttering, beautiful and responsive gaming experience, laying a solid foundation for the stable operation of the platform during the business peak period.

[0064] In terms of dealing with the business trough, when the time passes to 2 a.m., the user activity of the platform decreases significantly like a tide. The intelligent monitoring module in the system will detect this trend of change in a timely and accurate manner like a keen sentinel, and then trigger the automatic reduction mechanism. Under the orderly operation of this mechanism, the system will gradually release the container resources that were heavily occupied during the peak period but are no longer needed at this moment. It will first conduct a comprehensive scan and evaluation of the usage status of each container, determine which game processes in which containers have ended or have been idle for a long time, and then, according to the preset priority rules and resource recovery strategies, methodically close these containers and recycle the memory, CPU and other resources they occupy. Through this precise resource recovery and integration, the platform can effectively reduce operating costs, avoid idleness and waste of resources, and enable the platform to maintain stable operation in the most streamlined and efficient resource allocation mode during the business trough, accumulating energy and making full preparations for the next business peak period.

[0065] It should be noted that during the entire resource scheduling process of expansion or reduction, the system should try to keep the capacity level of the system's overall resource pool relatively stable while ensuring stable business operations, to ensure that the resource margin is not too much to cause cost waste, or too little to affect the elasticity of the entire system.

[0066] See also Figure 2 , Figure 2 The system composition diagram of an Android distributed edge container adaptive elastic expansion strategy according to an exemplary embodiment is shown, and the system includes: The resource monitoring module 21 is used to monitor the resource usage of each Android container in real time; wherein the resource usage of the Android container includes: CPU utilization, memory usage, and business load status; A resource allocation module 22, used to utilize the resource usage of each Android container obtained by the resource monitoring module, and dynamically allocate the preset computing resources to each preset Android container using a preset algorithm; The container expansion module 23 is used to adjust the preset re-arranged available container resources or call elastic resources for resource expansion according to different business traffic characteristics and the specifications of the resources required by the business.

[0067] It can be understood that the same or similar parts of the above embodiments can be referenced to each other, and the contents not described in detail in some embodiments can refer to the same or similar contents in other embodiments.

[0068] It should be noted that, in the description of the present invention, the terms "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or implying relative importance. In addition, in the description of the present invention, unless otherwise specified, the meaning of "plurality" refers to at least two.

[0069] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code that includes one or more executable instructions for implementing the steps of a specific logical function or process, and the scope of the preferred embodiments of the present invention includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by those skilled in the art to which the embodiments of the present invention belong.

[0070] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0071] A person skilled in the art may understand that all or part of the steps in the method for implementing the above-mentioned embodiment may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, which, when executed, includes one or a combination of the steps of the method embodiment.

[0072] In addition, each functional unit in each embodiment of the present invention may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0073] The storage medium mentioned above can be a read-only memory, a magnetic disk or an optical disk, etc.

[0074] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0075] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations of the present invention. A person skilled in the art may change, modify, replace and vary the above embodiments within the scope of the present invention.

Claims

1. A method for adaptive elastic expansion strategy of Android distributed edge container, characterized in that: The method comprises: Real-time monitoring of resource usage of each Android container; wherein the resource usage of the Android container includes: CPU utilization, memory usage, and business load status; Using the resource usage of each Android container obtained by the resource monitoring module, the preset computing resources are dynamically allocated to each preset Android container using a preset algorithm; According to different business traffic characteristics and the specifications of the resources required by the business, the preset available container resources are re-arranged and adjusted, or elastic resources are called for resource expansion.

2. The method according to claim 1, characterized in that: The real-time monitoring of resource usage of each Android container includes: By interacting with the preset board hardware interface and the preset Android container management interface, the resource usage of each Android container is monitored in real time.

3. The method according to claim 1, characterized in that The resource usage of each Android container obtained by the resource monitoring module is used to dynamically allocate the preset computing resources to each preset Android container using a preset algorithm, including: Using the resource usage of each Android container obtained by the resource monitoring module, container nodes are allocated through a preset load balancing strategy to obtain a first processing result; Using the first processing result and combining the business resource demand characteristics, the preset computing resources are dynamically allocated to the preset Android containers.

4. The method according to claim 1, characterized in that The steps of implementing the re-arrangement of available container resources include: Obtain historical statistical data for real-time monitoring of resource usage of each Android container within a preset period of time, and mark the business type or application name of the Android container; Removing outliers from the historical statistical data, filling missing values ​​using a preset algorithm, and converting the data into a time series format suitable for modeling using a preset time series conversion algorithm to obtain first processed data; Input the first processed data into a preset trained processing model to output the demand for Android containers of different specifications in a future period of time; wherein the demand for Android containers includes: the number of CPU cores and the memory size; According to the demand for Android containers of different specifications in the future period and the preset allocation strategy, a resource allocation plan is obtained; According to the resource allocation plan, using the demand for Android containers of different specifications in the future period and the preset allocation strategy, dynamically create a required number of new Android containers, and configure the newly created new Android containers with appropriate hardware specifications; If there are containers that are running but no longer meet the above allocation strategy, they are smoothly transferred to other locations through hot migration technology.

5. The method according to claim 4, characterized in that The method further comprises: When a business request ends, the associated Android container resources are recovered, temporary files are deleted, and the container status table entry in the database is updated synchronously.

6. The method according to claim 1, characterized in that The method of adjusting the preset re-arranged available container resources or calling elastic resources for resource expansion according to different business traffic characteristics and the specifications of the resources required by the business includes: Obtain historical data of the load of each Android container, and use a preset machine learning algorithm to predict the business traffic trend in the future period based on the historical data of the load of each Android container; wherein the business traffic trend in the future period includes: a peak time period and a trough time period; The peak time period, the trough time period, and the preset business characteristics are used to predict the peak, and the capacity expansion process is started according to the predicted result of the peak.

7. The method according to claim 6, characterized in that The peak time period, the trough time period, and the preset business characteristics are used to predict the peak, and the expansion process is started according to the predicted result of the peak, including: According to the predicted result of the wave peak and the currently available fixed resources, if the currently available fixed resources are sufficient to cope with the peak traffic predicted according to the predicted result of the wave peak, the hardware specifications of the re-arranged available container resources are adjusted, and a new Android container is created according to business needs or the resources of the existing Android container are expanded; If the currently available fixed resources are insufficient to cope with the peak traffic predicted based on the peak prediction results, the elastic resource pool is called according to business needs.

8. A system for adaptive elastic expansion strategy of Android distributed edge container, applied to the method for adaptive elastic expansion strategy of Android distributed edge container as described in any one of claims 1-7, characterized in that: The system comprises: The resource monitoring module is used to monitor the resource usage of each Android container in real time; wherein the resource usage of the Android container includes: CPU utilization, memory usage, and business load status; A resource allocation module, used to utilize the resource usage of each Android container obtained by the resource monitoring module, and dynamically allocate the preset computing resources to each preset Android container using a preset algorithm; The container expansion module is used to adjust the preset re-arrangement of available container resources or call elastic resources for resource expansion according to different business traffic characteristics and the specifications of resources required by the business.

Images