CN117714546A

CN117714546A - Multi-cloud environment service grid application method

Info

Publication number: CN117714546A
Application number: CN202311736593.3A
Authority: CN
Inventors: 王鹏飞; 程昕云; 汤铭; 夏飞; 何金陵; 李亚乔; 王智慷; 刘喆; 宋浒
Original assignee: Information and Telecommunication Branch of State Grid Jiangsu Electric Power Co Ltd
Current assignee: Information and Telecommunication Branch of State Grid Jiangsu Electric Power Co Ltd
Priority date: 2023-12-15
Filing date: 2023-12-15
Publication date: 2024-03-15

Abstract

The invention discloses a multi-cloud environment service grid application method, which comprises the following steps: constructing an application service grid based on a multi-cloud environment; collecting user data in a front-end browser to obtain behavior labels of users related to various application services; generating and storing a user portrait through the behavior label of the user; and analyzing the power service demands, setting corresponding service routing rules for corresponding application services based on the user portraits, and carrying out power service application on the user dimension by adopting an application service grid. The invention realizes gridding management of the multi-cloud service, micro-service decoupling and asynchronous system management, the front-end data acquisition technology only needs to pay attention to the change of the front-end page, does not need to consider the deployment architecture and the cross-cloud problem of the rear end, and has lower maintenance cost and complexity.

Description

Multi-cloud environment service grid application method

Technical Field

The invention relates to the technical field of multi-cloud management, micro-service management and service grid management, in particular to a multi-cloud environment service grid application method.

Background

At present, in the power business industry, service management and operation and maintenance can only be carried out on multi-cloud environment data through apm, and then fusion is carried out based on a business data center, so that the coupling degree is high, the maintenance cost is extremely high, as shown in fig. 2, the framework capability does not have the gridding management of multi-cloud service, and therefore micro-service decoupling and asynchronous system management cannot be carried out, namely, the operation and maintenance service is high in availability, risk safety monitoring, business capability enhancement and the like.

The invention with the application number of CN202010690731.9 discloses a base station position determining method, a device, equipment and a storage medium, aiming at the power industry, a network base station-based micro grid and a user portrait technology are combined, and the micro grid is in the aspect of the network base station and does not relate to the cloud environment service grid management of the power industry although the user portrait technology is applied.

The invention with the application number of CN202311213983.2 discloses a unified application monitoring platform based on power grid business, as shown in figure 3, by constructing a function of a business center all-link monitoring tool, all-link monitoring of business center service and interface call is realized, but the management of a multi-cloud environment service grid in the power industry is not involved.

The invention with the application number of CN201810474841.4 discloses a method and a system for constructing an electric power user portrait based on big data, which have the beneficial effects of effectively extracting data characteristics from mass data, improving the utilization value of the data, comprehensively and comprehensively observing user characteristics in a life cycle, improving the electricity utilization experience of customers and preventing potential risk users; the invention also does not relate to meshing management of multi-cloud services.

Disclosure of Invention

The invention aims to provide a method for applying a service grid in a multi-cloud environment, which is based on multi-cloud fusion management of a power service system, marks the flow of the multi-cloud environment and generates a user portrait of power service characteristics, and the portrait is combined with the service grid capability to effectively realize the application of the following scenes, such as multi-cloud environment release and A/B verification of service operation and service high availability dimension; user dimension value user analysis, directional flow management and control, fusing, current limiting, authorization authentication and risk security monitoring; personalized popularization of service dimension and the like, thereby the electric power industry has grid management of multi-cloud service and micro-service decoupling and asynchronous system management are realized.

In order to achieve the technical purpose, the invention adopts the following technical scheme:

a multi-cloud environment service grid application method, the multi-cloud environment service grid application method comprising the steps of:

s1, a Kubernetes platform is installed in a multi-cloud cluster service of a power service system, an ission tool is installed in a cluster of the Kubernetes platform, a routing strategy and a flow control strategy of an ission control panel are initialized, application service discovery related to power service, communication routing and load balancing among application services are achieved through an Envoy component, and an application service grid based on a multi-cloud environment is constructed;

S2, collecting user data in a front-end browser, storing the collected data, optimizing a traditional hierarchical clustering algorithm by combining a condensation hierarchical clustering algorithm and K-Means, and specifically, performing hierarchical clustering on the collected data by using the condensation hierarchical clustering algorithm to obtain clustering results of different layers; taking clusters with deeper layers in the clustering result as initial K clustering centers; optimizing the initial K clustering centers through a K-Means algorithm, and classifying and marking the collected user traffic to obtain behavior labels of users related to each application service;

s3, performing hierarchical recursive search on the user traffic with the classification marks obtained in the step S2 by using an Apriori association rule algorithm, mining frequent item sets and association rules, discovering frequently and simultaneously occurring behaviors according to the association rules, generating user portraits by using the discovered relevant behavior labels of the frequently and simultaneously occurring behaviors, and storing the user portraits;

and S4, analyzing each power service requirement, setting a corresponding service routing rule for the corresponding application service based on the user portrait generated in the step S3, and carrying out power service application on the user dimension by adopting the application service grid based on the cloud environment constructed in the step S1.

Further, in step S2, the process of obtaining the behavior tags of the users related to the respective application services includes the following steps:

s21, acquiring flow information in a browser from a cloud environment, and tracking and recording power business behaviors of a user in a power website or a power application program; filtering the missing values which do not meet the condition and the data which do not meet the definition of the tag;

s22, integrating the preprocessed recorded data to generate a fact label; the fact label comprises purchase times, active days, purchase types, search times, purchase amount, browse commodity types and login times;

s23, based on the fact label, using a K-Means algorithm to divide data points with similar characteristics into the same category, and generating a model label; the model labels comprise high-value users, high-activity users, user satisfaction, user interest and hobbies, user risk scores, product purchase preferences and user association relations;

s24, classifying the recorded behavior data based on the model labels, and generating a prediction label according to classification results; the predictive labels include potential purchasing groups, potential losing groups, high repurchase intention users and high-value sensitive users;

S25, adding User, broswersource, dataLable to the URL parameter of the user flow, and marking the URL parameter; user, broswersource and DataLable represent, respectively, user information, browser information, and flow marks; the user information records the source and the user name of the user; the browser information records the browser type, the user source and the channel; traffic is marked as behavioral tags including facts tags, model tags, and predictive tags.

Further, in step S21, the power business actions include clicking button data, browsing data, form data, interaction data, and event data.

Further, in step S3, the process of generating the user portrait includes the following steps:

based on the power business application service requirement, performing hierarchical recursive search on the user traffic with the classification marks obtained in the step S2 by using an Apriori association rule algorithm, mining frequent item sets and association rules of the user traffic, and finding frequent and concurrent behaviors according to the association rules; and integrating and analyzing the basic information of the user and the frequently and simultaneously occurring behavior data, classifying and storing the flow marks, and generating the user portraits related to the power business application service.

Further, in step S3, the number of associations between user behaviors is counted using the degree of support, the degree of support= (number of simultaneous occurrence of condition and conclusion)/(total number of transactions); the probability of necessarily occurring between user behaviors is evaluated using a confidence level, confidence level= ((number of simultaneous occurrences of condition and conclusion)/(number of occurrences of condition)).

Further, in step S3, the user portrait includes a power user type, a user marketing tag, an operation and maintenance tag, a demand tag, and a power system tag;

the power consumer type comprises industrial enterprises, commercial enterprises, residents and government regulatory staff;

the user marketing label comprises general electricity users, high-quality users, risk users and blacklist users;

the operation and maintenance label comprises a normal flow user, a current limiting user and a high flow user;

the demand labels comprise service discovery, load balancing, flow management, fault recovery and security, service operation and maintenance, high availability of service, risk security monitoring, value user analysis and directional flow management and control;

the power system mark comprises power supply management, equipment operation and maintenance management, financial management, power demand management, power consumption metering settlement, market transaction, power consumption data inquiry, power consumption service application, electric charge payment, power market supervision, power enterprise operation audit, power policy formulation, power transaction, settlement and risk management.

Further, the power industry application service includes: service operation and service high-availability dimension multi-cloud environment release and A/B verification; user dimension value user analysis, directional flow management and control, fusing, current limiting, authorization authentication and risk security monitoring; personalized popularization of service dimension.

Further, in step S4, the process of performing the connection configuration on the user dimension by using the application service grid based on the cloud environment includes the following steps:

screening out relevant user portrait labels according to the requirement of the current-limiting fusing configuration of the relation, adding the screened user portrait labels in the fusing rule of the relation in a self-defining way, and defining parameter conditions corresponding to different levels of fusing rules in the DestinationRule, wherein the parameters comprise the maximum waiting request number, the maximum connection number, the connection error number, the interval time for triggering fusing, the interval time for removing the service and the more than how much request service can be removed;

associating the defined fusing rules into a VirtualService, and matching the fusing rules of the user portrait tag by using a VirtualService configuration; the route configuration of the virtual service is used for matching a subset rule of the DestinationRule;

When the power user performs service operation, the user portrait mark corresponding to the power user is called, and the corresponding fusing rule is triggered according to the called user portrait mark, so that fusing occurs in the designated grid range.

Further, in step S4, the process of performing a/B testing using the application service grid based on the cloud environment includes the following steps:

screening out user portrait labels related to the A/B test requirements, and adding the screened user portrait labels in the A/B test rules of the relation in a user-defined manner;

when the power user performs service operation, the user portrait mark corresponding to the power user is called, if the called user portrait is matched with the user portrait of the A/B test requirement, the Envoy component is adopted to route the traffic to the service-demo-v2, otherwise, the traffic is routed to the service-demo-v1, so that the traffic is distributed to different service versions, and the A/B test is performed.

Further, in step S4, each power service requirement is parsed, and a corresponding service routing rule is set for the corresponding application service based on the user service attribute and the user portrait generated in step S3.

Compared with the prior art, the invention has the following beneficial effects:

according to the application method of the multi-cloud environment service grid, based on the multi-cloud fusion management of the power service system, the multi-cloud environment flow is marked, the user portraits of the power service characteristics are generated, and the portraits are combined with the service grid capability to effectively realize the application of the following scenes, such as multi-cloud environment release and A/B verification of service operation and service high availability dimension; user dimension value user analysis, directional flow management and control, fusing, current limiting, authorization authentication and risk security monitoring; personalized popularization of service dimension and the like, thereby the electric power industry has grid management of multi-cloud service and micro-service decoupling and asynchronous system management are realized.

Drawings

FIG. 1 is a flowchart of a method for applying a multi-cloud environment service grid of the present invention;

FIG. 2 is a schematic diagram of the operation of a service grid relationship component;

FIG. 3 is a block diagram of a unified application monitoring platform based on grid services in the prior art;

FIG. 4 is a user portrait creation flowchart;

FIG. 5 is a schematic diagram of a clustering algorithm involving a label generation process;

FIG. 6 is a schematic diagram of an association rule algorithm involving tags;

FIG. 7 is a schematic diagram of a user representation, k8s, and service grid combination;

FIG. 8 is a schematic diagram of a service grid atio fusing and current limiting application;

fig. 9 is a diagram of a service grid application monitoring console.

Detailed Description

Embodiments of the present invention are described in further detail below with reference to the accompanying drawings.

Background/noun interpretation:

service grid technology: based on the micro-service architecture technology, services and resources are organized into a grid so that services can be dynamically discovered, interacted and collaborated. The service grid technology comprises functions of service registration and discovery, load balancing, fault recovery, security authentication, monitoring, log and the like, and uses the capabilities of proxy mode, service discovery load balancing and the like to improve the reliability, expandability and security among services.

Taking the example of the ission (mainstream service grid platform), see fig. 2, which includes service discovery, load balancing, traffic management, fault recovery and security, the ission uses Envoy as its proxy component, which is a high-performance, extensible proxy for handling and managing communications between micro-services. The atio is tightly integrated with a Kubernetes isovolumetric orchestration platform, so that the construction and management of micro services in a cloud environment is simpler and more efficient.

Kubernetes: kubernetes (commonly abbreviated as K8 s) is an open-source container orchestration platform for automated deployment, extension, and management of containerized applications, herein using K8s capabilities for multi-cloud service deployment, enabling multi-cloud management capabilities.

And (3) multi-cloud data fusion: the method is used for centralized management and monitoring of a plurality of cloud services. Through multi-cloud service management, a user can uniformly manage resources, applications and services of different cloud service providers, and the background technology is that user behaviors are collected through a front end, and business traffic is aggregated, stored and analyzed, so that cross-cloud management of business system data is achieved.

The embodiment of the invention discloses a method for applying a service grid in a multi-cloud environment, which comprises the following steps:

S1, a Kubernetes platform is installed in a multi-cloud cluster service of a power service system, an ission tool is installed in a cluster of the Kubernetes platform, a routing strategy and a flow control strategy of an ission control panel are initialized, application service discovery related to power service, communication routing and load balancing among application services are achieved through an Envoy component, and an application service grid based on a multi-cloud environment is constructed.

S2, collecting user data in a front-end browser, storing the collected data, optimizing a traditional hierarchical clustering algorithm by combining a condensation hierarchical clustering algorithm and K-Means, and specifically, performing hierarchical clustering on the collected data by using the condensation hierarchical clustering algorithm to obtain clustering results of different layers; taking clusters with deeper layers in the clustering result as initial K clustering centers; and optimizing the initial K clustering centers through a K-Means algorithm, and classifying and marking the collected user traffic to obtain the behavior labels of the users related to each application service.

S3, performing hierarchical recursive search on the user traffic with the classification marks obtained in the step S2 by using an Apriori association rule algorithm, mining frequent item sets and association rules, finding frequently and simultaneously occurring behaviors according to the association rules, and generating and storing user portraits through the found related behavior labels of the frequently and simultaneously occurring behaviors.

The multi-cloud environment service grid application method comprises the technologies of k8s multi-cloud management, grid service management, data flow labels, user portraits and the like, mainly comprises the steps of collecting data/flow (mainly related to user data) of a power service system from a multi-cloud environment, generating user portraits based on a big data algorithm, and combining service grid capability to apply, such as service operation and maintenance, high-availability of service, risk security monitoring, value user analysis and directional flow management and control.

The overall scheme of the multi-cloud environment service grid application method is shown in fig. 1, and specifically comprises the following steps:

constructing a service grid: firstly, k8s is installed in a multi-cloud cluster service of a power business system, an atio tool is installed in the k8s cluster, a routing strategy and flow control of an atio control panel are initialized, application service discovery is achieved through an Envoy component (when one service wants to communicate with another service, the service discovery request is sent to the Envoy of the atio, the atio control plane returns position information of a target service, so that the request can be correctly routed to an instance of the target service), namely, communication routing and load balancing among the application services form a grid of the application service, and the grid service and the load balancing information are displayed by using a kali component, see fig. 9. The text of FIG. 9 is for illustration of a grid schematic of an application service, and is not intended to be limiting

Generating a user traffic label: and secondly, user data in a front-end browser is collected and stored through a user flow marking and user portrait generating method, and a traditional hierarchical clustering algorithm is optimized by combining aggregation hierarchical clustering and K-Means, namely the collected data is subjected to hierarchical clustering by the aggregation hierarchical clustering algorithm, so that clustering results of different layers are obtained. Taking clusters with deeper layers in the result as initial K cluster centers; and optimizing the initial clustering center through a K-Means algorithm, so that the collected user flow is classified and marked, and behavior labels such as 'large charge "," arrearage overdue "," triggering wind control blackout' and the like are obtained.

Generating big data user portraits: thirdly, carrying out layer-by-layer recursive search on the user flow with the mark by using an Apriori association rule algorithm, mining frequent item sets and association rules of the user flow, finding frequently and simultaneously occurring behaviors according to the association rules, generating user images through relevant behavior labels and storing the user images, such as high-quality users with good credit, high-power consumption, "risk users with overdue arrears, blacklist users with arrears still high consumption, and the like.

Service grid and application: fourth, based on the portrait of the user and the established service grid, the user dimension can be fused, limited in current, A/BTest, recommended for user service, etc., for example, the service of the 'blacklist user' is fused, any consuming operation is not allowed to be continued, and the fused limit can be canceled only if the service is returned to the 'normal user' after recharging. In this embodiment, the power industry application service includes: service operation and service high-availability dimension multi-cloud environment release and A/B verification; user dimension value user analysis, directional flow management and control, fusing, current limiting, authorization authentication and risk security monitoring; personalized popularization of business dimension, and the like.

Compared with the technical scheme that business coding and customization are needed to be respectively realized when various applications in the power industry are realized in the prior art, the optimized scheme can be operated uniformly through the capability of the service grid, so that the coding cost is saved, the capability management is also facilitated, meanwhile, the cross-cloud application is realized, the problems of decoupling and cross-cloud management of the micro-service are finally solved, and the application range of the business is also enhanced.

The user traffic sign, user portrayal generation and service grid application of the present invention will be described in detail with reference to the accompanying drawings.

The user flow marking and user portrait generating method is shown in detail in fig. 4 to 6, and comprises the following steps:

s21, acquiring flow information in a browser from a cloud environment, and tracking and recording power business behaviors (such as clicking button data, browsing data, form data, interaction data, event data and the like) of a user in a power website or a power application program; and filtering the missing values which are not in accordance with the conditions and the data which are not in accordance with the label definition.

S22, integrating the preprocessed recorded data to generate a fact label; the fact tag includes the number of purchases, the number of days of activity, the type of purchase, the number of searches, the amount of purchases, the type of goods browsed and the number of logins.

S23, based on the fact label, using a K-Means algorithm to divide data points with similar characteristics into the same category, and generating a model label; the model tags include high value users, highly active users, user satisfaction, user interests, user risk scores, product purchase preferences, and user associations.

S24, classifying the recorded behavior data based on the model labels, and generating a prediction label according to classification results; predictive labels include potential purchasing populations, potential churning populations, high repurchase intent users, and high value sensitive users.

In the invention, the data traffic marking refers to marking and classifying the data traffic of the cloud fusion so as to facilitate the management, analysis and application of the data. By marking the data flow, different types of data can be distinguished and classified, and subsequent data processing and analysis work is facilitated. The method comprises the steps of collecting user information data through Cookie marking, URL parameter marking, user operation behavior, click button event analysis and the front-end browser collecting method, writing the collecting result into a big data process library, and carrying out flow marking through aggregation hierarchical clustering and K-Means: integrating and analyzing basic information, behavior data, interest labels and the like of a user, wherein the basic information comprises information such as age, gender, geographic position, electricity utilization habit and the like; and finally, classifying and storing the flow marks to generate the portrait of the user.

Firstly, data of a front-end browser are collected, and flow information user information, browser information and flow marks in the browser are obtained: { User, broswersource, dataLable }. The user information records unique information such as the source, the user name and the like of the user; browser information records browser type, user source, channel and the like; the flow mark records other key information in the user object and transmits the key information to big data for analysis and classification; URL parameter tag: along with Cookie tag logic, user, broswersource, dataLable is appended to url.

Secondly, analyzing and collecting the behavior of the user, tracking, recording and analyzing the behavior of the user in the website or the application program (clicking button data, browsing data, form data, interaction data and event data), storing the data into a large database for classification and recording, and also referring to User, broswersource, dataLable for simple classification.

And then preprocessing the data, wherein the data preprocessing can filter the missing values which are not in accordance with the conditions and the data which are not in accordance with the label definition by using the conventional data cleaning, and select the characteristics of the label characteristics which are in accordance with the enumerated clustering algorithm, so that the quality of the data is ensured to be complete. Assume that one of the acquisition instances is as follows: { "user_id": "12345", "timestamp": "2023-11-25T 12:00:00 "," action ": "click", "page_url": "https: "device_info"// www.idea.com/products ": { "device_type": "Mobile", "os": "iOS" }, "location": { "laytude": 40.7128, "longitude": -74.0060 }; for this piece of data, the data can be flushed to (big data traffic logic filters out unwanted fields): user= { id=123456 }, broswersource= { ios, dataLable { type=click, url= "https }: v/www.idea.com/products ",.

Finally, the data points with similar characteristics are divided into the same category by using a K-Means algorithm, and model tag content (such as high-value users, high-activity users and user risk scores/risk users) is generated.

Taking high value traffic label generation as an example, high value users are exemplified as follows: extracting the characteristics in the DataLable mark: and extracting the DataLable features according to the high-value user standard in the big data service. Including frequency of purchases, amount of purchases, number of times high value products are viewed, etc. In some examples, machine learning may also be used to identify high value users such as K-means clustering algorithms, random forests, and the like.

And (3) carrying out user traffic marking on the collected data of the cloud environment, establishing a specified user traffic label, and customizing and generating a user image based on the traffic label (the customized traffic label refers to the definition content of the user portrait). In this embodiment, the user portraits refer to the detailed descriptions and features of the users formed by analyzing and mining data such as behaviors, interests, preferences and the like of the users, so as to better understand the needs of the users and provide personalized services and recommendations, and the user portraits and data traffic markers can be used in combination, so that the interests and needs of the users can be better understood by marking and analyzing the behaviors and data traffic of the users, and thus the personalized recommendations and services can be provided. Meanwhile, the data flow is marked and classified, so that the data can be better managed and utilized, the value and the application effect of the data are improved, and the user portrait is generated mainly based on the business of the power industry. The user portrait comprises a power user type, a user marketing label, an operation and maintenance label, a demand label and a power system label;

In step S3, the process of generating a user portrait includes the steps of:

The Apriori association rule algorithm is used to find frequent item sets and association rules: the a priori nature indicates that if a certain set of terms is frequent, then all its subsets must also be frequent; the connection property indicates that if the first k-1 items of two item sets are identical, they can be connected into one larger item set.

The support and confidence formulas are used for capturing the high-frequency behavior obtained from the label DataLable in the user traffic, such as recharging, consumption, wind control, blackening and some user operation habits.

The support is used to count the number of associations between actions, which user actions occur frequently, such as multiple overdue, arrears, etc. of the user.

Support = (number of simultaneous occurrences of condition and conclusion)/(total number of transactions).

Confidence is used to evaluate the likelihood of necessarily occurring between behaviors, and to count the ratio of the number of simultaneous occurrences of a condition and a conclusion to the number of occurrences of the condition. For example, after each payment overdue by the user a, the probability that the next month will continue overdue is certain. This value must occur when approaching 1 and not when 0.

Confidence= ((number of simultaneous occurrences of condition and conclusion)/(number of occurrences of condition)).

The foregoing process is described below by way of example.

Suppose we have the following tag data result set:

transaction event = a user [ { (premium user ', (premium user) }, B user { (premium user ', (restrictive user ', (premium user), ' risk security monitor ',) C user { (risk user ', (restrictive user ', (blacklist user ', (service delivery and service high availability ',) D user { (risk user ', (business enterprise ', (blacklist user ', (service delivery and service high availability ',) E user { (premium user ', (resident ', (restrictive user ',) blacklist user ', } ]

1. Firstly, the support degree counts 5 pieces of total transaction data, the risk user happens 2 times, and the blacklist user happens 3 times.

2. Carrying out reliability statistics on the result set;

3. the confidence statistics shows that the probability that the risk user and the blacklisted user will occur simultaneously is 100%, users C and D.

4. User images are generated and stored (listing herein high-value clients, high-risk groups high-value, high-task two portrayal types): the user A is high-value, and service recommendation is preferably carried out on the user A; c and D are high-risk, global wind control measures such as current limiting and fusing are needed, corresponding user portraits can be obtained in service operation later, and service grid application is carried out in a request message header.

The code is as follows:

rule=apriori (transactions, min_support=2, min_confidence=0.5) prints the generated association rule.

Fig. 8 is a service grid atio fusing and current limiting application schematic diagram. Taking the case of the IstioLimit fusing as an example, in step S4, the process of performing the IstioLimit fusing configuration on the user dimension by adopting the application service grid based on the cloud environment comprises the following steps:

when the power user performs service operation, the user portrait mark corresponding to the power user is called, and the corresponding fusing rule is triggered according to the called user portrait mark, so that fusing occurs in the designated grid range. Fig. 7 is a schematic diagram of a user portrait, k8s and service grid combination principle, and the text part in fig. 7 is merely for illustrating the schematic diagram of the combination principle, and is not meant to be limiting.

The Istio current limiting fusing configuration:

adding custom labels (e.g., user portraits are high-risk users) to the istio fuse rules;

the fuse rules for matching user portrait tags using a VirtualService configuration:

the match field is thus used to match requests with a particular user-profile tag (high-task) and to configure the fuse rules for those requests. When having a user-profile: when the request of the high-risk tag arrives, a defined fusing rule is triggered, that is, after the high-value user is matched, fusing occurs in a designated grid range.

The flow restriction rules based on the user portrayal labels are configured using the VirtualService:

the current limiting method is similar to fusing:

1. when a high-value tag exists in the message header, triggering the delay of the high-value tag:

percent：50

fixedDelay：5s

current limiting rules:

percentage: 50: the percentage of delay faults applied to the request is specified. In this case, 50% of the requests will be restricted from transmission.

fixedDelay:5s: the fixed delay time was set to 5 seconds.

The above scenario means that when a request with a high-value header tag enters the service grid, 50% of the requests are restricted from transmission, while there will be a 5 second latency delay before the response returns.

In step S4, the process of performing a/B testing using the application service grid based on the cloud environment includes the following steps:

a/B test: the method is used for testing whether the functions of two or more versions meet expectations, belongs to a lightweight release test strategy and is configured as follows:

the label information of the user is "high-value", then the traffic will be routed to service-demo-v2, otherwise the traffic will be routed to service-demo-v1; and distributing the flow to different service versions by high-value users in the user portrait to realize A/B testing capability. By introducing a high-value user tag into a specified route, then the target server can be arbitrarily specified by proxy through the specified route.

The invention is based on user portraits, and the service grid capability can also realize the capabilities of load balancing, user security authorization, personalized recommendation, user value analysis (monitoring and tracking) and the like.

Routing and load balancing: its request is routed to the nearest service node according to the user tag or geographic location in the user profile or to a specific service instance according to the user's preferences.

Security function in combination with the relation: in combination with authentication and authorization information for the user profile, it is ensured that only authenticated users can access specific services or resources.

Traffic management function using istio: through the traffic management function of the ission, specific user requests may be routed to different service versions or micro-service instances based on the geographic location of the user, the device type, or other attributes.

Monitoring and tracking functions: by combining the user portrayal and the monitoring and tracking functions of the relation (performance index of the service, tracking/analyzing service call chain, dashboard), the interaction behavior between the user and the service can be monitored and analyzed.

In summary, the multi-cloud environment service grid can help the micro-service architecture to decouple and manage an asynchronous system, namely, high availability of operation and maintenance services, risk safety monitoring and service capability enhancement.

In step S4, analyzing each power service requirement, and setting a corresponding service routing rule for the corresponding application service based on the user service attribute and the user portrait generated in step S3; comprising the following steps: the method comprises the steps of directly using the user portrait to combine with the capability application of the service grid, directly using the user service attribute to combine with the capability application of the service grid, or combining the user service attribute and the user portrait as rule parameters and then combining with the capability application of the service grid. For example, if the balance of the user A of a certain portrait type is 0 (the field is read from a service library), the service grid capability is directly used for carrying out cross-cloud directional pushing of the user message, so that the user is reminded to recharge.

As described above, the management of the micro service architecture of the present invention is performed on the service grid, and the service grid appears to decouple the micro service, that is, the service grid capability is above the micro service discovery, and when a traffic goes to the service application, it is routed by the Envoy component of the ission first. Both the service grid and the K8S are container-level management; micro-services are application-level management, so input data does not require the relevant content of the micro-service architecture either. Compared with the multi-cloud data acquisition through apm, the front-end data acquisition technology only needs to pay attention to the change of the front-end page, does not need to consider the deployment architecture and the cross-cloud problem of the rear end, and is lower in maintenance cost and complexity.

It will be appreciated by those skilled in the art that embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein. The solutions in the embodiments of the present application may be implemented in various computer languages, for example, object-oriented programming language Java, and an transliterated scripting language JavaScript, etc.

The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

While preferred embodiments of the present application have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. It is therefore intended that the following claims be interpreted as including the preferred embodiments and all such alterations and modifications as fall within the scope of the application.

It will be apparent to those skilled in the art that various modifications and variations can be made in the present application without departing from the spirit or scope of the application. Thus, if such modifications and variations of the present application fall within the scope of the claims and the equivalents thereof, the present application is intended to cover such modifications and variations.

Claims

1. The application method of the multi-cloud environment service grid is characterized by comprising the following steps of:

2. The method for applying the cloud environment service grid according to claim 1, wherein the process of obtaining the behavior tags of the users related to the respective application services in step S2 comprises the steps of:

3. The method of claim 1, wherein in step S21, the power business actions include clicking button data, browsing data, form data, interaction data, and event data.

4. The method for applying the multi-cloud environment service grid according to claim 1, wherein in step S3, the process of generating the user portrait includes the steps of:

5. The method according to claim 4, wherein in step S3, the number of associations between user behaviors is counted using a degree of support, the degree of support= (the number of simultaneous occurrence of conditions and conclusions)/(the total number of transactions); the probability of necessarily occurring between user behaviors is evaluated using a confidence level, confidence level= ((number of simultaneous occurrences of condition and conclusion)/(number of occurrences of condition)).

6. The method according to claim 1, wherein in step S3, the user portraits include a power user type, a user marketing tag, an operation and maintenance tag, a demand tag, and a power system tag;

7. The method of claim 1, wherein the power industry application service comprises: service operation and service high-availability dimension multi-cloud environment release and A/B verification; user dimension value user analysis, directional flow management and control, fusing, current limiting, authorization authentication and risk security monitoring; personalized popularization of service dimension.

8. The application method of the service grid of the multi-cloud environment according to claim 1, wherein in step S4, the process of performing the istio current limiting fusing configuration on the user dimension by using the service grid of the multi-cloud environment comprises the following steps:

9. The application method of the service grid of the multi-cloud environment according to claim 1, wherein in step S4, the process of performing the a/B test using the service grid of the multi-cloud environment comprises the steps of:

10. The method according to claim 1, wherein in step S4, each power service requirement is parsed, and a corresponding service routing rule is set for the corresponding application service based on the user service attribute and the user portraits generated in step S3.