Cloud workload classification processing method and device

Abstract

The invention discloses a cloud workload classification processing method and device. The method comprises the following steps: acquiring passively uploaded static data in a container arrangement system and a cloud platform management system according to static element tags; the method comprises the following steps: collecting actively uploaded static data when a host, a virtual machine Agent and a container micro-isolation POD run through a static data collection component; the working load is labeled through a rule matching assembly by utilizing a set label matching rule and combining the passive uploading static data and the active uploading static data collected by the static data collection assembly, and a first labeling result is obtained; utilizing an active reporting mechanism of a host, a virtual machine Agent and a container micro-isolation POD to acquire dynamic label elements during operation through a dynamic data acquisition component; labeling the workload through a rule clustering component in combination with dynamic label elements acquired by a dynamic data acquisition component to obtain a second labeling result; and performing strategy configuration on the workload according to a labeling result. The stability and the performance of the system are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of workload classification, and in particular relates to a cloud workload classification processing method and device. Background Art

[0002] Micro-isolation is achieved by dividing all services within the data center into several tiny network nodes according to specific principles, performing access control on these nodes based on dynamic policy analysis, logically isolating these nodes, and restricting user lateral movement.

[0003] At present, micro-isolation grouping often needs to consider multiple factors, such as the correlation between workloads, resource utilization of computing nodes, etc. Quickly determine the grouping based on the communication relationship between captured workloads to adapt to the rapidly changing cloud business system. In a hybrid architecture with a large number of workloads and frequent changes, neither the static element-based division method nor the rule-based dynamic element division method is sufficient to complete accurate division. Summary of the invention

[0004] To this end, the present invention provides a cloud workload classification processing method and device to solve the problem of poor classification accuracy of traditional technologies under a hybrid architecture with a large amount of workload and frequent changes.

[0005] In order to achieve the above object, the present invention provides the following technical solution: a cloud workload classification processing method, comprising:

[0006] Calling the API interface of the container orchestration system and the cloud platform management system through the static data acquisition component to obtain the static element tags in the container orchestration system and the cloud platform management system, and obtaining the passively uploaded static data in the container orchestration system and the cloud platform management system according to the static element tags;

[0007] Utilizing the active reporting mechanism of the host agent, virtual machine agent and container micro-isolation POD, the static data actively uploaded by the host agent, the virtual machine agent and the container micro-isolation POD during operation is collected through the static data collection component;

[0008] The workload is labeled by using a set label matching rule through a rule matching component in combination with the passively uploaded static data and the actively uploaded static data collected by the static data collection component to obtain a first labeling result;

[0009] Utilize the active reporting mechanism of the host agent, virtual machine agent and container micro-isolation POD, and collect the dynamic label elements of the host agent, virtual machine agent and container micro-isolation POD during operation through the dynamic data collection component;

[0010] Labeling the workload by combining the rule clustering component with the dynamic label elements collected by the dynamic data collection component to obtain a second labeling result;

[0011] The workload is policy-configured according to the first labeling result and the second labeling result.

[0012] As a preferred solution for the cloud workload classification processing method, the static data collection component uses Kubernetes' RESTful API or kubectl command line tool to pull static data from the container orchestration system;

[0013] The static data collection component uses the cloud vendor SDK or RESTful API to pull static data from the cloud platform management system.

[0014] As a preferred solution of the cloud workload classification processing method, the static element tag includes configuration information, resource allocation status and environment variables; the storage format of the static element tag is JSON or YAML;

[0015] The passively uploaded static data includes hardware resource usage and software level information.

[0016] As a preferred solution of the cloud workload classification processing method, it also includes modeling the static element label as a data union of multiple data sources:

[0017] D={Dk,Dc}

[0018] Where D is the uniformly stored dataset; Dk is the dataset from the container orchestration system; Dc

[0019] It is a data set from the cloud platform management system;

[0020] The method further includes normalizing the collected tag values ​​of the static element tags, wherein the normalization formula is:

[0021] X norm =(XX min ) / (X max -X min )

[0022] Where, X norm represents the normalized label value, X represents the original label value, and X min , X max Represents the minimum and maximum values ​​in the data set respectively.

[0023] As a preferred solution of the cloud workload classification processing method, it also includes: analyzing the correlation of the static element labels by establishing a correlation matrix, and the formula of the correlation matrix is:

[0024]

[0025] In the formula, r ij is the Pearson correlation coefficient between label i and label j; X i is the value of label i, X j is the value of label j; Cov(X i , X j ) is the variable X i and variable X j The covariance of Var(X i ) is the variable X i The variance of Var(X j ) is the variable X j The variance of .

[0026] As a preferred solution of the cloud workload classification processing method, the tag matching rule is defined by a logical expression and a regular expression;

[0027] The rule clustering component labels the workload using a set clustering algorithm.

[0028] The present invention also provides a cloud workload classification processing device, comprising:

[0029] A first static data acquisition module is used to call the API interface of the container orchestration system and the cloud platform management system through the static data acquisition component to obtain the static element tags in the container orchestration system and the cloud platform management system, and obtain the passively uploaded static data in the container orchestration system and the cloud platform management system according to the static element tags;

[0030] The second static data acquisition module is used to utilize the active reporting mechanism of the host agent, the virtual machine agent and the container micro-isolation POD to collect the actively uploaded static data of the host agent, the virtual machine agent and the container micro-isolation POD during operation through the static data collection component;

[0031] A first labeling result generating module is used to label the workload by using a set label matching rule through a rule matching component, in combination with the passively uploaded static data and the actively uploaded static data collected by the static data collection component, to obtain a first labeling result;

[0032] The dynamic label element acquisition module is used to utilize the active reporting mechanism of the host agent, the virtual machine agent and the container micro-isolation POD to collect the dynamic label elements of the host agent, the virtual machine agent and the container micro-isolation POD during operation through the dynamic data collection component;

[0033] A second labeling result generating module is used to label the workload by combining the dynamic label elements collected by the dynamic data collection component through the rule clustering component to obtain a second labeling result;

[0034] A workload policy configuration module is used to perform policy configuration on the workload according to the first labeling result and the second labeling result obtained.

[0035] As a preferred solution of the cloud workload classification processing device, in the first static data acquisition module:

[0036] The static data collection component uses Kubernetes' RESTful API or kubectl command line tool to pull static data from the container orchestration system;

[0037] The static data collection component uses the cloud vendor SDK or RESTful API to pull static data from the cloud platform management system.

[0038] As a preferred solution of the cloud workload classification processing device, in the first static data acquisition module:

[0039] The static element tag includes configuration information, resource allocation status and environment variables; the storage format of the static element tag is JSON or YAML;

[0040] The passively uploaded static data includes hardware resource usage and software level information.

[0041] As a preferred solution of the cloud workload classification processing device, it also includes:

[0042] A data union module is used to model the static element tag as a data union of multiple data sources:

[0043] D={Dk,Dc}

[0044] Where D is the uniformly stored dataset; Dk is the dataset from the container orchestration system; Dc

[0045] It is a data set from the cloud platform management system;

[0046] The normalization processing module is used to normalize the tag values ​​of the collected static element tags. The normalization processing formula is:

[0047] X norm =(XX min ) / (X max -X min )

[0048] Where, X norm represents the normalized label value, X represents the original label value, and X min , X max Respectively represent the minimum and maximum values ​​in the data set;

[0049] The correlation analysis module is used to analyze the correlation of the static element labels by establishing a correlation matrix. The formula of the correlation matrix is:

[0050]

[0051] In the formula, r ij is the Pearson correlation coefficient between label i and label j; X i is the value of label i, X j is the value of label j; Cov(X i , X j ) is the variable X i and variable X j The covariance of Var(X i ) is the variable X i The variance of Var(X j ) is the variable X j The variance of .

[0052] As a preferred solution of the cloud workload classification processing device, in the first labeling result generation module, the label matching rule is defined by a logical expression and a regular expression;

[0053] In the second labeling result generating module, the rule clustering component labels the workload using a set clustering algorithm.

[0054] The beneficial effects of the present invention are as follows: passively uploaded static data in the container orchestration system and the cloud platform management system are obtained according to static element tags; actively uploaded static data of the host agent, virtual machine agent and container micro-isolation POD during operation is collected through the static data collection component; the workload is labeled by combining the passively uploaded static data and the actively uploaded static data collected by the static data collection component using the set tag matching rules through the rule matching component to obtain a first labeling result; the dynamic label elements of the host agent, virtual machine agent and container micro-isolation POD during operation are collected through the dynamic data collection component by utilizing the active reporting mechanism of the host agent, virtual machine agent and container micro-isolation POD; the workload is labeled by combining the dynamic label elements collected by the dynamic data collection component through the rule clustering component to obtain a second labeling result; and the workload is policy configured according to the first labeling result and the second labeling result. The present invention adopts a combination of static partitioning, rule-based dynamic partitioning, and clustering comprehensive partitioning; it can ensure the acquisition of the latest and accurate static data, thereby providing a solid foundation for subsequent management and monitoring work; it can better understand the operating status of the system, discover potential problems in time, and take corresponding optimization measures, thereby improving the stability and performance of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] In order to more clearly illustrate the implementation methods of the present invention or the technical solutions in the prior art, the drawings required for the implementation methods or the prior art descriptions are briefly introduced below. Obviously, the drawings in the following description are only exemplary, and for ordinary technicians in this field, other implementation drawings can be derived from the provided drawings without creative work.

[0056] The structures, proportions, sizes, etc. illustrated in this specification are only used to match the contents disclosed in the specification so as to facilitate understanding and reading by persons familiar with the technology. They are not used to limit the conditions under which the present invention can be implemented, and therefore have no substantial technical significance. Any structural modification, change in proportion or adjustment of size shall still fall within the scope of the technical contents disclosed in the present invention without affecting the effects and purposes that can be achieved by the present invention.

[0057] Figure 1 A schematic diagram of a cloud workload classification processing method flow chart provided by an embodiment of the present invention;

[0058] Figure 2 A schematic diagram of the technical architecture of the cloud workload classification processing method provided by an embodiment of the present invention;

[0059] Figure 3 A schematic diagram of the architecture of a cloud workload classification processing device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0060] The following is a description of the implementation of the present invention by specific embodiments. People familiar with the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. Obviously, the described embodiments are 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.

[0061] At present, the implementation method of micro-isolation is to divide all the services within the data center into several tiny network nodes according to specific principles, perform access control on these nodes based on dynamic policy analysis, logically isolate these nodes, and restrict users from moving horizontally. In the micro-isolation architecture, there is no longer the concept of internal and external networks. Instead, the data center network is isolated into many tiny computing units, which are referred to as nodes. Each node needs to be authenticated by the micro-isolation client to access the resources of other nodes. If the node identity authentication fails or does not have access rights, it will be intercepted by the client.

[0062] Among them, a node can be a portal website, a database, an audit device, or even a file server. As long as it has a certain data processing capability, it can become a node. No longer considered "trusted" because it is in the intranet, all nodes are logically isolated, and access between nodes is controlled. The more detailed the node division, the clearer the control center's visualization of the traffic of the entire data center network. Micro-isolation grouping often needs to consider multiple factors, such as the correlation between workloads, the resource utilization of computing nodes, and so on. Quickly determine the grouping based on the communication relationship between the captured workloads to adapt to the rapidly changing cloud business system.

[0063] In a hybrid architecture of bare metal, virtual machines, and containers, since the number of workloads that need to be managed is huge and changes rapidly, it is necessary to use automation technology to automatically label the workloads in preparation for subsequent label-based policy configuration. In a hybrid architecture with a large number of workloads that change frequently, neither a simple static element-based division method nor a rule-based dynamic element division method can achieve accurate division. In view of this, an embodiment of the present invention provides a cloud workload classification processing method and device, and the following is the specific content of an embodiment of the present invention.

[0064] Example 1

[0065] See also Figure 1 and Figure 2 , an embodiment of the present invention provides a cloud workload classification processing method, comprising the following steps:

[0066] S1. Calling the API interface of the container orchestration system and the cloud platform management system through the static data acquisition component to obtain the static element tags in the container orchestration system and the cloud platform management system, and obtaining the passively uploaded static data in the container orchestration system and the cloud platform management system according to the static element tags;

[0067] S2. Using the active reporting mechanism of the host agent, virtual machine agent and container micro-isolation POD, the static data actively uploaded by the host agent, the virtual machine agent and the container micro-isolation POD during operation is collected through the static data collection component;

[0068] S3, labeling the workload by using a set label matching rule through a rule matching component, combining the passively uploaded static data and the actively uploaded static data collected by the static data collection component, and obtaining a first labeling result;

[0069] S4. Using the active reporting mechanism of the host agent, virtual machine agent and container micro-isolation POD, the dynamic label elements of the host agent, the virtual machine agent and the container micro-isolation POD during operation are collected through the dynamic data collection component;

[0070] S5. Labeling the workload by combining the rule clustering component with the dynamic label elements collected by the dynamic data collection component to obtain a second labeling result;

[0071] S6. Perform policy configuration on the workload according to the first labeling result and the second labeling result.

[0072] In this embodiment, static element labels are a means of describing workloads in business logic. In a hybrid architecture, the following elements can be used to define labels for workloads:

[0073] (1) Bare metal:

[0074] Static elements: server model, host name, host operating system, host IP address, host subnet, host location, host security domain;

[0075] Dynamic elements: application software running on the host, ports open on the host, application systems, network traffic, etc.

[0076] (2) Virtual Machine

[0077] Static elements: VM name, VM operating system, VM IP address, VM subnet, cloud geographic region, cloud security group, cloud host image name, cloud host image type, cloud resource group, and various tags defined by the cloud management platform;

[0078] Dynamic elements: application software running on the virtual machine, ports opened by the virtual machine, application systems, network traffic, etc.

[0079] (3) Container

[0080] Taking the Docker+K8S architecture as an example, the labels that can identify containers include the following elements: static elements: container name, container image type, container image name, container image source, labels defined by the container orchestration system, etc.; dynamic elements: container mapping IP address, container mapping subnet, application system, network traffic, etc.

[0081] Specifically, in step S1, by calling the API of the container orchestration system and the API of the cloud platform management system, relevant static element tags can be actively pulled from the container orchestration system and the cloud platform management system. These static element tags include but are not limited to various configuration information, resource allocation conditions, and environment variables. In this way, it is possible to ensure that the latest and accurate static data is obtained, thereby providing a solid foundation for subsequent management and monitoring work.

[0082] In this embodiment, in step S2, the active reporting mechanism of the host agent, virtual machine agent and container micro-isolation POD is relied upon to collect static data during runtime. The host / virtual machine agent is a lightweight agent program running on the host / virtual machine, which can monitor the running status of the host in real time and report relevant information to the central management system. The container micro-isolation POD is a technology for implementing micro-isolation in a container environment, which can ensure the security of communication between containers and can actively report the running status and related static data of the container.

[0083] Thus, by combining steps S1 and S2, the runtime static data of the host, virtual machine, and container can be fully collected. This data not only includes the usage of hardware resources, such as CPU, memory, and storage, but also includes software-level information, such as operating system version, running applications and their configurations. Through the collection and analysis of this data, we can better understand the operating status of the system, discover potential problems in a timely manner, and take corresponding optimization measures to improve the stability and performance of the system.

[0084] In this embodiment, in step S3, the administrator can define a series of label matching rules through logical expressions and regular expressions, and the rule matching component executes the rules for the collected static data to label the workload. Then in step S4, the active reporting mechanism of the host agent, virtual machine agent and container micro-isolation POD is used to collect the dynamic label elements of the host agent, virtual machine agent and container micro-isolation POD during operation through the dynamic data collection component.

[0085] In this embodiment, in step S5, the administrator can define rules or not define rules for the collected dynamic label elements, and the rule clustering component executes the clustering algorithm to dynamically label the workload. The labeling results have also been fed back to the rule library in the form of new rules.

[0086] The clustering algorithm itself is an unsupervised learning method used to group objects in a data set so that objects in the same group (i.e., cluster) are as similar as possible, while objects in different groups are as different as possible. There are many types of clustering algorithms, among which the most commonly used is the K-means algorithm, which will not be described here.

[0087] In a possible embodiment, the static data collection component uses Kubernetes' RESTful API or kubectl command line tool to pull static data from the container orchestration system; wherein the pulled static data includes node and Pod configuration information (such as CPU, memory allocation; service discovery and load balancing rules; environment variables and labels (labels, annotations).

[0088] Among them, the Kubernetes data collection code is as follows:

[0089] "from kubernetes import client,config

[0090] defget_kubernetes_static_data():

[0091] #Load Kubernetes configuration

[0092] config.load_kube_config()

[0093] #Create an API client

[0094] v1 = client.CoreV1Api()

[0095] #Get node information

[0096] nodes = v1.list_node().items

[0097] node_data=[{"name":node.metadata.name,

[0098] "cpu":node.status.capacity["cpu"],

[0099] "memory":node.status.capacity["memory"]}

[0100] fornode in nodes]

[0101] #Get Pod Information

[0102] pods=v1.list_pod_for_all_namespaces().items

[0103] pod_data=[{"name":pod.metadata.name,

[0104] "namespace":pod.metadata.namespace,

[0105] "labels": pod.metadata.labels}

[0106] forpod inpods]

[0107] return{"nodes":node_data,"pods":pod_data}".

[0108] In a possible embodiment, the static data collection component uses the cloud vendor SDK or RESTful API to pull static data from the cloud platform management system. The pulled static data includes virtual machine instance configuration information (such as instance type, storage size); resource allocation (such as used storage, bandwidth); security group rules and IAM permissions.

[0109] In a possible embodiment, the storage format of the static element tag is JSON or YAML. The data storage structure is as follows:

[0110]

[0111] In a possible embodiment, the static element tag is further modeled as a data union of multiple data sources:

[0112] D={Dk,Dc}

[0113] In the formula, D is a dataset stored in a unified manner; Dk is a dataset from a container orchestration system; Dc is a dataset from a cloud platform management system;

[0114] The method further includes normalizing the collected tag values ​​of the static element tags for subsequent processing, wherein the normalization formula is:

[0115] X norm =(XX min ) / (X max -X min )

[0116] Where, X norm represents the normalized label value, X represents the original label value, and X min , X max Represents the minimum and maximum values ​​in the data set respectively.

[0117] In a possible embodiment, the method further includes: analyzing the relevance of the static element labels by establishing a relevance matrix, wherein the formula of the relevance matrix is:

[0118]

[0119] In the formula, r ij is the Pearson correlation coefficient between label i and label j; X i is the value of label i, X j is the value of label j; Cov(X i , X j ) is the variable X i and variable X j The covariance of Var(X i ) is the variable X i The variance of Var(X j ) is the variable X j The variance of .

[0120] In summary, the embodiment of the present invention calls the API interface of the container orchestration system and the cloud platform management system through the static data collection component to obtain the static element tags in the container orchestration system and the cloud platform management system, and obtains the passively uploaded static data in the container orchestration system and the cloud platform management system according to the static element tags; utilizes the active reporting mechanism of the host agent, virtual machine agent and container micro-isolation POD, and collects the actively uploaded static data of the host agent, the virtual machine agent and the container micro-isolation POD during operation through the static data collection component; utilizes the set tag matching rules through the rule matching component, combined with The passively uploaded static data and actively uploaded static data collected by the static data collection component are used to label the workload to obtain a first labeling result; the active reporting mechanism of the host agent, virtual machine agent and container micro-isolation POD is utilized to collect the dynamic label elements of the host agent, virtual machine agent and container micro-isolation POD during operation through the dynamic data collection component; the workload is labeled through the rule clustering component combined with the dynamic label elements collected by the dynamic data collection component to obtain a second labeling result; the workload is strategically configured according to the first labeling result and the second labeling result. The present invention adopts a combination of static partitioning, rule-based dynamic partitioning and clustering comprehensive partitioning; it can ensure the acquisition of the latest and accurate static data, thereby providing a solid foundation for subsequent management and monitoring work; it can better understand the operating status of the system, discover potential problems in a timely manner, and take corresponding optimization measures, thereby improving the stability and performance of the system.

[0121] It should be noted that the method of the embodiment of the present disclosure can be performed by a single device, such as a computer or a server. The method of the present embodiment can also be applied in a distributed scenario and completed by multiple devices cooperating with each other. In the case of such a distributed scenario, one of the multiple devices can only perform one or more steps in the method of the embodiment of the present disclosure, and the multiple devices will interact with each other to complete the described method.

[0122] It should be noted that the above describes some embodiments of the present disclosure. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the above embodiments and still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0123] Example 2

[0124] See also Figure 3 Embodiment 2 of the present invention provides a cloud workload classification processing device, including:

[0125] The first static data acquisition module 100 is used to call the API interface of the container orchestration system and the cloud platform management system through the static data acquisition component to obtain the static element tags in the container orchestration system and the cloud platform management system, and obtain the passively uploaded static data in the container orchestration system and the cloud platform management system according to the static element tags;

[0126] The second static data acquisition module 200 is used to utilize the active reporting mechanism of the host agent, the virtual machine agent and the container micro-isolation POD to collect the actively uploaded static data of the host agent, the virtual machine agent and the container micro-isolation POD during operation through the static data collection component;

[0127] The first labeling result generating module 300 is used to label the workload by using the set label matching rule through the rule matching component, in combination with the passively uploaded static data and the actively uploaded static data collected by the static data collection component, to obtain a first labeling result;

[0128] The dynamic label element acquisition module 400 is used to utilize the active reporting mechanism of the host agent, the virtual machine agent and the container micro-isolation POD to collect the dynamic label elements of the host agent, the virtual machine agent and the container micro-isolation POD during operation through the dynamic data collection component;

[0129] A second labeling result generating module 500 is used to label the workload by combining the dynamic label elements collected by the dynamic data collection component through the rule clustering component to obtain a second labeling result;

[0130] The workload policy configuration module 600 is used to perform policy configuration on the workload according to the first labeling result and the second labeling result obtained.

[0131] In this embodiment, in the first static data acquisition module 100:

[0132] The static data collection component uses Kubernetes' RESTful API or kubectl command line tool to pull static data from the container orchestration system;

[0133] The static data collection component uses the cloud vendor SDK or RESTful API to pull static data from the cloud platform management system.

[0134] In this embodiment, in the first static data acquisition module 100:

[0135] The static element tag includes configuration information, resource allocation status and environment variables; the storage format of the static element tag is JSON or YAML;

[0136] The passively uploaded static data includes hardware resource usage and software level information.

[0137] In this embodiment, it also includes:

[0138] The data union module 700 is used to model the static element tag as a data union of multiple data sources:

[0139] D={Dk,Dc}

[0140] In the formula, D is a dataset stored in a unified manner; Dk is a dataset from a container orchestration system; Dc is a dataset from a cloud platform management system;

[0141] The normalization processing module 800 is used to perform normalization processing on the collected tag values ​​of the static element tags. The normalization processing formula is:

[0142] X norm =(XX min ) / (X max -X min )

[0143] Where, X norm represents the normalized label value, X represents the original label value, and X min , X max Respectively represent the minimum and maximum values ​​in the data set;

[0144] The correlation analysis module 900 is used to analyze the correlation of the static element labels by establishing a correlation matrix. The formula of the correlation matrix is:

[0145]

[0146] In the formula, r ij is the Pearson correlation coefficient between label i and label j; X i is the value of label i, X j is the value of label j; Cov(X i , X j ) is the variable X i and variable X j The covariance of Var(X i ) is the variable X i The variance of Var(X j ) is the variable X j The variance of .

[0147] In this embodiment, in the first labeling result generating module 300, the label matching rule is defined by a logical expression or a regular expression;

[0148] In the second labeling result generating module 500, the rule clustering component labels the workload using a set clustering algorithm.

[0149] It should be noted that the information interaction, execution process and other contents between the modules of the above-mentioned device are based on the same concept as the method embodiment in Example 1 of the present application, and the technical effects they bring are the same as those of the method embodiment of the present application. For specific contents, please refer to the description in the method embodiment shown in the previous part of the present application, and will not be repeated here.

[0150] Example 3

[0151] Embodiment 3 of the present invention provides a non-transitory computer-readable storage medium, in which a program code for a cloud workload classification processing method is stored. The program code includes instructions for executing a cloud workload classification processing method of embodiment 1 or any possible implementation thereof.

[0152] The computer-readable storage medium may be any available medium that can be accessed by a computer or a data storage device such as a server or a data center that includes one or more available media. The available medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state drive (SSD)).

[0153] Example 4

[0154] Embodiment 4 of the present invention provides an electronic device, including: a memory and a processor;

[0155] The processor and the memory communicate with each other via a bus; the memory stores program instructions that can be executed by the processor, and the processor calls the program instructions to execute a cloud workload classification processing method of embodiment 1 or any possible implementation thereof.

[0156] Specifically, the processor can be implemented by hardware or by software. When implemented by hardware, the processor can be a logic circuit, an integrated circuit, etc.; when implemented by software, the processor can be a general-purpose processor implemented by reading software codes stored in a memory. The memory can be integrated in the processor or can be located outside the processor and exist independently.

[0157] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium, for example, the computer instructions can be transmitted from a website site, computer, server or data center by wired (e.g., coaxial cable, optical fiber, digital consumer customer line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) mode to another website site, computer, server or data center.

[0158] Obviously, those skilled in the art should understand that the above modules or steps of the present invention can be implemented by a general computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, and optionally, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, and in some cases, the steps shown or described can be executed in a different order than here, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. Thus, the present invention is not limited to any specific combination of hardware and software.

[0159] Although the present invention has been described in detail above by general description and specific embodiments, it is obvious to those skilled in the art that some modifications or improvements can be made to the present invention. Therefore, these modifications or improvements made without departing from the spirit of the present invention all belong to the scope of protection claimed by the present invention.

Claims

1. A cloud workload classification processing method, characterized in that: include: Calling the API interface of the container orchestration system and the cloud platform management system through the static data acquisition component to obtain the static element tags in the container orchestration system and the cloud platform management system, and obtaining the passively uploaded static data in the container orchestration system and the cloud platform management system according to the static element tags; Utilizing the active reporting mechanism of the host agent, virtual machine agent and container micro-isolation POD, the static data actively uploaded by the host agent, the virtual machine agent and the container micro-isolation POD during operation is collected through the static data collection component; The workload is labeled by using a set label matching rule through a rule matching component in combination with the passively uploaded static data and the actively uploaded static data collected by the static data collection component to obtain a first labeling result; Utilize the active reporting mechanism of the host agent, virtual machine agent and container micro-isolation POD, and collect the dynamic label elements of the host agent, virtual machine agent and container micro-isolation POD during operation through the dynamic data collection component; Labeling the workload by combining the rule clustering component with the dynamic label elements collected by the dynamic data collection component to obtain a second labeling result; The workload is policy-configured according to the first labeling result and the second labeling result.

2. A cloud workload classification processing method according to claim 1, characterized in that: The static data collection component uses Kubernetes' RESTful API or kubectl command line tool to pull static data from the container orchestration system; The static data collection component uses the cloud vendor SDK or RESTful API to pull static data from the cloud platform management system.

3. The cloud workload classification processing method according to claim 1, characterized in that: The static element tag includes configuration information, resource allocation status and environment variables; the storage format of the static element tag is JSON or YAML; The passively uploaded static data includes hardware resource usage and software level information.

4. The cloud workload classification processing method according to claim 3, characterized in that: Also included is modeling the static element tag as a data union of multiple data sources: D={Dk,Dc} In the formula, D is a dataset stored in a unified manner; Dk is a dataset from a container orchestration system; Dc is a dataset from a cloud platform management system; The method further includes normalizing the collected tag values ​​of the static element tags, wherein the normalization formula is: X norm =(X-X min ) / (X max -X min ) Where, X norm represents the normalized label value, X represents the original label value, and X min , X max Represents the minimum and maximum values ​​in the data set respectively.

5. A cloud workload classification processing method according to claim 4, characterized in that: Also includes: The correlation of the static element labels is analyzed by establishing a correlation matrix, and the formula of the correlation matrix is: In the formula, r ij is the Pearson correlation coefficient between label i and label j; X i is the value of label i, X j is the value of label j; Cov(X i , X j ) is the variable X i and variable X j The covariance of Var(X i ) is the variable X i The variance of Var(X j ) is the variable X j The variance of .

6. The cloud workload classification processing method according to claim 1, characterized in that: The tag matching rules are defined by logical expressions and regular expressions; The rule clustering component labels the workload using a set clustering algorithm.

7. A cloud workload classification processing device, characterized in that: include: A first static data acquisition module is used to call the API interface of the container orchestration system and the cloud platform management system through the static data acquisition component to obtain the static element tags in the container orchestration system and the cloud platform management system, and obtain the passively uploaded static data in the container orchestration system and the cloud platform management system according to the static element tags; The second static data acquisition module is used to utilize the active reporting mechanism of the host agent, the virtual machine agent and the container micro-isolation POD to collect the actively uploaded static data of the host agent, the virtual machine agent and the container micro-isolation POD during operation through the static data collection component; A first labeling result generating module is used to label the workload by using a set label matching rule through a rule matching component, in combination with the passively uploaded static data and the actively uploaded static data collected by the static data collection component, to obtain a first labeling result; The dynamic label element acquisition module is used to utilize the active reporting mechanism of the host agent, the virtual machine agent and the container micro-isolation POD to collect the dynamic label elements of the host agent, the virtual machine agent and the container micro-isolation POD during operation through the dynamic data collection component; A second labeling result generating module is used to label the workload by combining the dynamic label elements collected by the dynamic data collection component through the rule clustering component to obtain a second labeling result; A workload policy configuration module is used to perform policy configuration on the workload according to the first labeling result and the second labeling result obtained.

8. The cloud workload classification processing device according to claim 7, characterized in that: In the first static data acquisition module: The static data collection component uses Kubernetes' RESTful API or kubectl command line tool to pull static data from the container orchestration system; The static data collection component uses the cloud vendor SDK or RESTful API to pull static data from the cloud platform management system.

9. The cloud workload classification processing device according to claim 7, characterized in that: In the first static data acquisition module: The static element tag includes configuration information, resource allocation status and environment variables; the storage format of the static element tag is JSON or YAML; The passively uploaded static data includes hardware resource usage and software level information.

10. The cloud workload classification processing device according to claim 9, characterized in that: Also includes: A data union module is used to model the static element tag as a data union of multiple data sources: D={Dk,Dc} In the formula, D is a dataset stored in a unified manner; Dk is a dataset from a container orchestration system; Dc is a dataset from a cloud platform management system; The normalization processing module is used to normalize the tag values ​​of the collected static element tags. The normalization processing formula is: X norm =(X-X min ) / (X max -X min ) Where, X norm represents the normalized label value, X represents the original label value, and X min , X max Respectively represent the minimum and maximum values ​​in the data set; The correlation analysis module is used to analyze the correlation of the static element labels by establishing a correlation matrix. The formula of the correlation matrix is: In the formula, r ij is the Pearson correlation coefficient between label i and label j; X i is the value of label i, X j is the value of label j; Cov(X i , X j ) is the variable X i and variable X j The covariance of Var(X i ) is the variable X i The variance of Var(X j ) is the variable X j The variance of In the first labeling result generating module, the label matching rule is defined by a logical expression and a regular expression; In the second labeling result generating module, the rule clustering component labels the workload using a set clustering algorithm.

Images