CN111614570A - A flow control system and method for service mesh - Google Patents

Abstract

Embodiments of the present invention provide a flow control system and method for a service mesh. The system includes: a flow-limiting subsystem, configured to perform flow-limit control on the flow of the flow-limited service data through a token bucket flow-limit algorithm, and forward the key service data to the service degradation subsystem; the service degradation subsystem is used to perform service level control on the network bandwidth of the key service data forwarded by the current limiting subsystem according to the service quality level, and obtain the key after the flow control. Serve. The embodiments of the present invention improve the service quality of the service grid in the narrow bandwidth network environment by performing current limiting and service degradation on different types of service data.

Description

A flow control system and method for service mesh

technical field

The present invention relates to the technical field of flow control, and in particular, to a flow control system and method for a service grid.

Background technique

With the introduction of the micro-service architecture, it has brought a huge impact on the field of service design architecture. More and more traditional service architectures are transforming towards the micro-service architecture, and the Service Mesh service grid architecture is derived, which combines basic communication with Let the service mesh do it. Based on technologies such as current limiting and service degradation, it is indispensable in microservice grids to ensure service quality in service grids in a narrow-bandwidth network environment.

Although service meshes can ensure service stability by providing circuit breakers, current limiting, and downgrades, and use rapid communication to ensure the quality of service of microservice clusters in general, for example, Sentinel components provide current limiting, downgrade, and load protection and other functions to solve problems such as high concurrency, surge in business traffic, high load and network delay; csemesher can simplify the development and maintenance of microservices for users, and provide an integrated microservice governance method to improve the living environment of microservices . However, in the ever-changing network environment, the service quality of the existing service mesh tools needs to be improved.

Therefore, there is an urgent need for a flow control system and method for a service mesh to solve the above problems.

SUMMARY OF THE INVENTION

In view of the problems existing in the prior art, the embodiments of the present invention provide a system and method for traffic control for a service mesh.

In a first aspect, an embodiment of the present invention provides a flow control system for a service mesh, including:

The current limiting subsystem is used to control the flow of the current limiting service data through the token bucket current limiting algorithm, and forward the key service data to the service degradation subsystem;

The service degradation subsystem is used for performing service level control on the network bandwidth of the key service data forwarded by the current limiting subsystem according to the service quality level, so as to obtain the key service after the flow control.

Further, the system further includes: a microservice resource management subsystem, which is used for continuous monitoring and status diagnosis of the microservice.

Further, the current limiting subsystem includes:

A filter module, configured to filter the received task request, and parse according to the task request to obtain the server name and port corresponding to the task request;

The flow control policy execution module is used to optimize the token bucket current limiting algorithm, and generate policy-triggered dynamic tokens according to the optimized token bucket current limiting algorithm.

Further, the service degradation subsystem includes:

The forwarding control module is used to store the task request in the request thread pool, and forward the task request to the target service according to the configuration information and the task request, so as to obtain the service instance corresponding to the task request, or, According to the configuration information and the task request, directly obtain the service instance corresponding to the task request;

A data compression and downgrade module is configured to perform data compression and downgrade processing on the key service data according to the network bandwidth, service instance CPU and memory information, and obtain the key service data after data compression.

Further, the microservice resource management subsystem includes:

The service configuration management module is used to provide an operation interface for users to deploy business services and edit service configuration information;

The service monitoring module is used to monitor the resource status of the host node and microservices;

The service status management module is used to monitor the services existing in the service mesh and diagnose the corresponding service and pod running status.

In a second aspect, an embodiment of the present invention provides a traffic control method for a service mesh, including:

Obtain the service data corresponding to the task request, and analyze and judge the service data. If the service data is the current-limiting service data, the flow of the current-limiting service data is limited by the token bucket current-limiting algorithm. control;

If the service data is key service data, then according to the service quality level, the service level control is performed on the network bandwidth of the key service data, and the key service after flow control is obtained.

Further, if the service data is key service data, then according to the service quality level, perform service level control on the network bandwidth of the key service data, and obtain the key service after flow control, including:

According to the service instance compression ratio formula, the service quality compression ratio is obtained, and the service quality level generated according to the service quality compression ratio is used to perform service level control on the network bandwidth of the key service data. The service instance compression ratio formula is:

Among them, Ratio(S _i ) represents the compression ratio of the i-th service to be downgraded, B _i represents the network bandwidth between the user task requesting end and the server end at the current moment; a ₁ , a ₂ and a ₃ respectively represent the weight parameters , a ₁ +a ₂ +a ₃ =1; U(C _i ) represents the CPU usage rate at the current moment, U(M _i ) represents the memory usage rate at the current moment, and B ₀ represents the connection between the user task requester and the server side The default value of network bandwidth, U(C ₀ ) represents the default value of CPU usage, and U(M ₀ ) represents the default value of memory usage.

Further, before the flow of the current-limiting service data is controlled by the token bucket current-limiting algorithm if the service data is the current-limiting service data, the method further includes:

Optimize the token bucket current limiting algorithm to obtain the optimized token bucket current limiting algorithm;

According to the optimized token bucket current limiting algorithm, a policy-triggered dynamic token is generated.

Further, generating a policy-triggered dynamic token according to the optimized token bucket current limiting algorithm includes:

Get the time of each token application, after the next token application request arrives, get the token application time difference between the current moment and the previous moment;

Obtain the number of tokens that can be added at the current moment according to the token application time difference and the token adding rate;

Obtain the final number of tokens added to the token bucket according to the number of tokens that can be added, the maximum number of tokens in the token bucket, and the current number of remaining tokens.

In a third aspect, an embodiment of the present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, the processor implementing the program as described in the second aspect when the processor executes the program Steps of the provided method.

A flow control system and method for a service grid provided by the embodiments of the present invention improves the service quality of the service grid in a narrow bandwidth network environment by performing current limiting and service degradation on different types of service data.

Description of drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the accompanying drawings that need to be used in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description These are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can also be obtained according to these drawings without creative efforts.

FIG. 1 is a schematic structural diagram of a flow control system for a service mesh provided by an embodiment of the present invention;

FIG. 2 is a schematic structural diagram of a current limiting subsystem provided by an embodiment of the present invention;

3 is a schematic structural diagram of a service degradation subsystem provided by an embodiment of the present invention;

FIG. 4 is a schematic structural diagram of a microservice resource management subsystem provided by an embodiment of the present invention;

5 is an overall framework diagram of a flow control system for a service mesh provided by an embodiment of the present invention;

6 is a schematic flowchart of a flow control method for a service mesh provided by an embodiment of the present invention;

7 is a schematic diagram of a current limiting execution step provided by an embodiment of the present invention;

8 is a schematic flowchart of a service degradation control provided by an embodiment of the present invention;

9 is a schematic diagram of a policy-triggered dynamic token provided by an embodiment of the present invention;

FIG. 10 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.

Detailed ways

In order to make the purposes, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments These are some embodiments of the present invention, but not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

FIG. 1 is a schematic structural diagram of a flow control system for a service mesh provided by an embodiment of the present invention. As shown in FIG. 1 , an embodiment of the present invention provides a flow control system for a service mesh, including:

The current limiting subsystem 101 is configured to perform current limiting control on the flow of the current limiting service data through the token bucket current limiting algorithm, and forward the key service data to the service degradation subsystem.

In this embodiment of the present invention, the current limiting subsystem 101 mainly performs current limiting on data traffic of non-critical services (ie, current limiting service data). Specifically, in this embodiment of the present invention, the current-limiting subsystem 101 needs to implement interception of a service request, firstly judging whether the request is current-limited, and forwarding the non-current-limiting service (that is, the key service) to the service degradation subsystem 102; For the service that needs to be limited, based on the token bucket current limiting algorithm, by obtaining the network bandwidth configuration data, the token placement rate and the token bucket size in the token bucket algorithm are controlled, so as to control the flow of the current limiting service data. control. In the embodiment of the present invention, the token bucket algorithm is applied to the service grid, so as to control the flow control of each micro-service, and then realize that according to the network conditions, the network bandwidth of the corresponding level is allocated to the service with the corresponding higher priority, and By setting a whitelist, the response of critical services is not restricted by flow control.

The service degradation subsystem 102 is configured to perform service level control on the network bandwidth of the key service data forwarded by the current limiting subsystem according to the service quality level, so as to obtain the flow-controlled key service.

In the embodiment of the present invention, the service degradation subsystem 102 is configured to use a corresponding service degradation policy in a network environment with narrow bandwidth and weak connection, and realize the service level selection of the service instance by monitoring the network conditions, and then realize the service level selection of the service instance in the special network. In this case, the data is compressed according to the degradation ratio calculated by the degradation algorithm. By sacrificing the cost of the service response data quality, the response probability of the service to the user is improved, and the switching of service quality is automatically realized as much as possible, so that the data is as good as possible. transmission. It should be noted that, in this embodiment of the present invention, the service degradation subsystem 102 may also manually set the preset service quality level.

A flow control system for a service grid provided by an embodiment of the present invention improves the service quality of the service grid in a narrow bandwidth network environment by performing current limiting and service degradation on different types of service data.

On the basis of the above embodiment, the system further includes: a microservice resource management subsystem, which is used for continuous monitoring and status diagnosis of the microservice.

In the embodiment of the present invention, the micro-service resource management subsystem is used to provide an operation interface for the client to deploy business services, edit service configuration information, and monitor the service grid and the service usage of the service grid, so as to realize the user-end pairing Microservices perform continuous monitoring and service status diagnosis.

FIG. 2 is a schematic structural diagram of a current limiting subsystem provided by an embodiment of the present invention. Referring to FIG. 2, on the basis of the foregoing embodiment, the current limiting subsystem includes:

The filter module is used for filtering the received task request, and analyzing the task request to obtain the server name and port corresponding to the task request.

In this embodiment of the present invention, the filter module is configured to filter external requests from the service grid, and parse out the requested service name and port through the user's task request. Specifically, in the embodiment of the present invention, the filter module first parses and judges the request, encapsulates the request parameters into an entity, and returns the request response corresponding to the forwarding of the request to the user. In the embodiment of the present invention, according to the functional requirements of the filter module, a uniform format url is used to access the business service, the filter module reads the configuration switch parameters related to flow control from Redis, and chooses to forward the request directly or needs to pass the flow Control the strategy, and set the Servlet as the filter entry to receive the request from the client. When the request from the client arrives, use the initialization method to put the request and response into the context object, parse the URI in the user request, and take out the service name and access of the real request. Port number to obtain configuration item information. It should be noted that, in this embodiment of the present invention, if the current limiting mode is enabled, the current limiting whitelist is matched, and if the task request exists in the whitelist, the request preprocessor package is called according to the service name and port number. The request entity is directly forwarded to the corresponding service, and then the request response is forwarded to the client, otherwise the task request is flow-limited; if the pass token can be obtained from the token bucket, the task request is forwarded to The corresponding service, otherwise intercept the request.

The flow control policy execution module is used to optimize the token bucket current limiting algorithm, and generate policy-triggered dynamic tokens according to the optimized token bucket current limiting algorithm.

In the embodiment of the present invention, the flow control policy execution module is used to optimize the token bucket current limiting algorithm, and calculates the current task request through the optimized token bucket current limiting algorithm according to the network environment and the flow control algorithm. Respond or intercept. Referring to Figure 2, the flow control policy execution module is mainly used to construct a token bucket and respond to the filter module's application for a token. Specifically, the flow control policy execution module optimizes the token bucket algorithm according to the established The token generation strategy triggers the dynamic generation of tokens. When there is a token application, the allocation strategy is selected according to the size of the token bucket and the current number of tokens. In the embodiment of the present invention, the token bucket is implemented based on Redis, which ensures the atomicity of data read and write so that the token bucket maintains data consistency. After the token bucket is calculated, the flow control policy execution module returns the result of the application to the filter module, if the token is obtained, it is released, otherwise the user request is intercepted.

FIG. 3 is a schematic structural diagram of a service degradation subsystem provided by an embodiment of the present invention. Referring to FIG. 3, on the basis of the foregoing embodiment, the service degradation subsystem includes:

The forwarding control module is used to store the task request in the request thread pool, and forward the task request to the target service according to the configuration information and the task request, so as to obtain the service instance corresponding to the task request, or, According to the configuration information and the task request, the service instance corresponding to the task request is directly obtained.

In the embodiment of the present invention, the forwarding control module is used to store the task request in the request thread pool. When the system is initialized, the forwarding control module reads the configuration information and the task request, and determines whether to obtain a service instance or directly forward the task. request to the target service. When the user configuration downgrade type is the downgrade between service instances, the forwarding control module will monitor the network bandwidth and obtain configuration information, query the corresponding service in the cluster, and select the service instance corresponding to the service according to the current network bandwidth and the user-configured threshold , and return the name and virtual IP of the real service instance to the request forwarding control module, so that the forwarding control module can dynamically configure the traffic rules in istio and migrate the traffic to the service instance of the corresponding version. Further, the forwarding control module forwards the request stored in the request thread pool to the real service, and the real service sends the request response to the forwarding control module, and responds to the user through the filter module.

A data compression and downgrade module is configured to perform data compression and downgrade processing on the key service data according to the network bandwidth, service instance CPU and memory information, and obtain the key service data after data compression.

In the embodiment of the present invention, as shown in FIG. 3 , the data compression and degradation module is used to compress key service data, so as to achieve the purpose of data compression and degradation. According to the proportion of a downgraded service response data in the cluster that needs to be downgraded, considering the three dimensions of network bandwidth, service instance CPU and memory, the adaptive service downgrade algorithm is used to automatically realize the function of data compression according to the downgrade threshold and the current network bandwidth. .

FIG. 4 is a schematic structural diagram of a microservice resource management subsystem provided by an embodiment of the present invention. Referring to FIG. 4 , on the basis of the foregoing embodiment, the microservice resource management subsystem includes:

The service configuration management module is used to provide an operation interface for users to deploy business services and edit service configuration information.

In the embodiment of the present invention, an operation interface is provided through a service configuration management module to realize foreground management of micro-service resources, so that users can deploy business services and edit service configuration information through this module. Specifically, the client accesses the foreground through the browser to perform related operations, and sends an http request to the background server. The request is processed through the logic function processing unit of the background server. The logic function processing unit sends data requests to the grid cluster and the database. The grid cluster And the database sends the response request to the logic function processing unit, and the logic function processing unit processes the data and returns it to the front-end browser.

The service monitoring module is used to monitor the resource status of the host node and microservices.

In the embodiment of the present invention, the service monitoring function of the service monitoring module is used for monitoring at two levels. One is overall monitoring, which mainly refers to the unified monitoring of the distribution of services in the cluster and the running status of each host node; the other is the overall monitoring. One refers to the monitoring of specific microservice operation indicators, including indicators such as throughput and resource usage.

The service status management module is used to monitor the services existing in the service mesh and diagnose the corresponding service and pod running status.

In the embodiment of the present invention, the service state management module checks the service details of the service existing in the service grid, and diagnoses the running state of the service instance corresponding to the service grid. Specifically, in this embodiment of the present invention, a service whose details are to be viewed is set, and then the request is sent to the background SvcController controller. The background logic queries the data of the service from the database according to the service name, and queries the kubernetes cluster according to the service name. The service and pod information corresponding to the service in .

FIG. 5 is an overall frame diagram of a flow control system for a service mesh provided by an embodiment of the present invention. Referring to FIG. 5, in an embodiment of the present invention, the flow control system includes a The center, the current limiting subsystem, the service degradation subsystem and the business cluster, among which, the limited current whitelist, the flow control configuration center (used to store the flow control configuration information), and the service degradation configuration center (used to store the service) are set in the configuration storage center. downgrade configuration information) and the service management database.

FIG. 6 is a schematic flowchart of a flow control method for a service mesh provided by an embodiment of the present invention. As shown in FIG. 6 , an embodiment of the present invention provides a flow control method for a service mesh, including:

Step 601: Acquire the service data corresponding to the task request, and analyze and judge the service data. If the service data is the current-limiting service data, use the token bucket current-limiting algorithm to control the traffic of the current-limiting service data. Perform current limiting control.

Step 602 , if the service data is key service data, perform service level control on the network bandwidth of the key service data according to the service quality level to obtain a flow-controlled key service.

In an embodiment of the present invention, FIG. 7 is a schematic diagram of a current limiting execution step provided by an embodiment of the present invention. Referring to FIG. 7, the specific steps are:

Step 701 , determine whether the traffic class key exists, if so, go to step 703 , if not, go to step 702 . It should be noted that the embodiment of the present invention is based on distributed cache-based cluster current limiting, and traffic allocation needs to be performed according to different traffic categories. Therefore, when accessing for the first time, the traffic category token bucket needs to be initialized;

Step 702, if the traffic class key does not exist, initialize the traffic class token bucket;

Step 703, if it exists, calculate the number of tokens to be placed;

Step 704, compare the current request time with the last token placement time of the current key, define delta as the time difference, the number of tokens placed=delta/r, and r represents the rate at which tokens are added, wherein, when placing tokens, Ensure that the number of tokens does not exceed the capacity of the token bucket, and at the same time, reset the time to put tokens;

Step 705, put the token within the capacity range of the token bucket;

Step 706, if the token acquisition is successful, that is, the capacity of the token bucket can be put into the token, then go to step 710; if it cannot be put into the current token bucket, go to step 707;

Step 707, determine whether there is an idle token bucket, if so, go to step 709; otherwise, the token acquisition fails, and the request is intercepted;

Step 708, intercept the request;

Step 709, placing the token in the free token bucket;

Step 710, allow the request.

FIG. 8 is a schematic flowchart of a service degradation control provided by an embodiment of the present invention. Referring to FIG. 8, the specific steps are:

Step 801, determine whether there is a custom downgrade policy to perform service downgrade selection for the currently requested service, if yes, go to step 803, if not, go to step 802;

Step 802, through the data compression downgrade module, automatically execute the budget downgrade policy on the business service instance;

Step 803, query service instance information from Etcd;

Step 804, obtain the instance version configuration from the downgrade configuration information stored in the downgrade configuration center;

Step 805, formulate a traffic allocation strategy, and perform service level control on the data;

Step 806: Accept the user's request for accessing the service, and return the data of the corresponding service quality.

The embodiment of the present invention provides a flow control method for a service grid, which improves the service quality of a service grid in a narrow bandwidth network environment by performing current limiting and service degradation on different types of service data.

On the basis of the above embodiment, if the service data is key service data, then according to the service quality level, perform service level control on the network bandwidth of the key service data, and obtain the key service after flow control, including:

According to the service instance compression ratio formula, the service quality compression ratio is obtained, and the service quality level generated according to the service quality compression ratio is used to perform service level control on the network bandwidth of the key service data. The service instance compression ratio formula is:

Among them, Ratio(S _i ) represents the compression ratio of the i-th service to be downgraded, B _i represents the network bandwidth between the user task requesting end and the server end at the current moment; a ₁ , a ₂ and a ₃ respectively represent the weight parameters , a ₁ +a ₂ +a ₃ =1; U(C _i ) represents the CPU usage rate at the current moment, U(M _i ) represents the memory usage rate at the current moment, and B ₀ represents the connection between the user task requester and the server side The default value of network bandwidth, U(C ₀ ) represents the default value of CPU usage, and U(M ₀ ) represents the default value of memory usage.

In the embodiment of the present invention, for a certain downgraded service S _i , when a request comes, obtain the network bandwidth B _i of the previous network collection period, and use the bandwidth B _i as the network bandwidth at the current moment. Since the quality of the service response data needs to be dynamically adjusted according to the current network bandwidth, it is necessary to obtain the network bandwidth data between the server and the client. In this embodiment of the present invention, the iperf tool is used to call the dmc command to detect the client and server-side network speed B _i ; monitor the resource usage information of service instances in the cluster through kubernetes, and collect data on the current CPU usage U(C _i ) and the current memory usage U(M _i ). Further, according to the actual situation of the system, the embodiment of the present invention sets default values B ₀ , U(C ₀ ) and U(M ₀ ) under normal environment and resource conditions, and sets weight parameters {α ₁ , α ₂ , α ₃ } identifies the impact weight of each index, so as to limit the degradation ratio through the above data.

Further, in the embodiment of the present invention, the ratio to which each service instance needs to be compressed is calculated by the service instance compression ratio formula, and the service instance with the largest Ratio value is selected from the service instance, and the corresponding service instance ip is returned, in order to minimize the service instance. The degree of data quality reduction, select the service instance with the smallest compression ratio, that is, select the service instance compressed to the largest ratio to respond to service requests.

On the basis of the above embodiment, before the flow of the current-limiting service data is controlled by the token bucket current-limiting algorithm if the service data is the current-limiting service data, the method further includes: include:

Optimize the token bucket current limiting algorithm to obtain the optimized token bucket current limiting algorithm;

According to the optimized token bucket current limiting algorithm, a policy-triggered dynamic token is generated.

On the basis of the above embodiment, generating a policy-triggered dynamic token according to the optimized token bucket current limiting algorithm includes:

Get the time of each token application, after the next token application request arrives, get the token application time difference between the current moment and the previous moment;

Obtain the number of tokens that can be added at the current moment according to the token application time difference and the token adding rate;

Obtain the final number of tokens added to the token bucket according to the number of tokens that can be added, the maximum number of tokens in the token bucket, and the current number of remaining tokens.

FIG. 9 is a schematic diagram of a policy-triggered dynamic token provided by an embodiment of the present invention. Referring to FIG. 9 , in the embodiment of the present invention, the token bucket algorithm is optimized and improved to provide a policy-triggered dynamic token. Bucket algorithm. The traditional token bucket algorithm needs to configure a dedicated thread to run repeatedly. In order to reduce the number of Redis operations and improve the performance of the flow control module, the embodiment of the present invention designs the token generation as a trigger generation method.

Specifically, in the embodiment of the present invention, the time of each token application is recorded by setting a variable, and after the next request for token application from the filter module arrives, the time difference between the current moment and the last token application time is calculated. , calculate the number of tokens that can be added during this period according to the rate of adding tokens, and the calculation method is expressed as:

Among them, currMillis represents the moment of the current application, lastMillis represents the moment of the last token application, and r represents the rate of adding tokens, in units/second. The number of tokens finally added to the token bucket is expressed as:

resultPermits=min(reservePermits,maxPermits-storedPermits);

Among them, maxPermits represents the size of the token bucket, and storedPermits represents the number of tokens currently existing.

Further, in the embodiment of the present invention, the service users deployed in the service grid need to be divided into different levels according to the business, that is, divided into various types of traffic. In general, for multi-classified traffic, a separate token bucket is defined for each class of traffic to achieve isolation from other classes. However, using such a solution will bring about performance problems such as wasted network resources and low bandwidth usage. When The tokens in a certain type of token bucket have been used up, and the corresponding type of requests are blocked or in a waiting state; at the same time, the request frequency of another type is not so high due to business needs, so that the There are always a large number of tokens in the bucket. In this case, network resources are in a wasted state; and, for requests for categories that are in short supply, the response will be delayed because no tokens are intercepted or queued. The quality of service suffers as a result. In view of the above problems, the embodiments of the present invention use semi-isolated token buckets to implement control and shaping of multi-category traffic, so that tokens of a certain type of traffic are allowed to use the remaining tokens for other categories under the condition that the tokens of a certain type of traffic meet the minimum requirements.

FIG. 10 is a schematic structural diagram of an electronic device provided by an embodiment of the present invention. Referring to FIG. 10 , the electronic device may include: a processor (processor) 1001, a communications interface (Communications Interface) 1002, a memory (memory) 1003, and a communication bus 1004, wherein , the processor 1001 , the communication interface 1002 , and the memory 1003 communicate with each other through the communication bus 1004 . The processor 1001 can call the logic instruction in the memory 1003 to execute the following method: obtain the service data corresponding to the task request, and analyze and judge the service data, if the service data is the current-limiting service data, pass the token Bucket current limiting algorithm, to perform current limiting control on the flow of the current limiting service data; if the service data is key service data, perform service level control on the network bandwidth of the key service data according to the service quality level, and obtain Critical services after flow control.

In addition, the above-mentioned logic instructions in the memory 1003 can be implemented in the form of software functional units and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on this understanding, the technical solution of the present invention can be embodied in the form of a software product in essence, or the part that contributes to the prior art or the part of the technical solution. The computer software product is stored in a storage medium, including Several instructions are used to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, Read-Only Memory (ROM, Read-Only Memory), Random Access Memory (RAM, Random Access Memory), magnetic disk or optical disk and other media that can store program codes .

On the other hand, an embodiment of the present invention further provides a non-transitory computer-readable storage medium on which a computer program is stored, and the computer program is implemented when executed by a processor to execute the service mesh provided by the above embodiments The traffic control method, for example, includes: acquiring service data corresponding to the task request, and analyzing and judging the service data, if the service data is the current-limiting service data, using the token bucket current-limiting algorithm to limit the The flow of the streaming service data is flow-limited; if the service data is key service data, the network bandwidth of the key service data is subject to service level control according to the service quality level to obtain a flow-controlled key service.

The device embodiments described above are only illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in One place, or it can be distributed over multiple network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution in this embodiment. Those of ordinary skill in the art can understand and implement it without creative effort.

From the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and certainly can also be implemented by hardware. Based on this understanding, the above-mentioned technical solutions can be embodied in the form of software products in essence or the parts that make contributions to the prior art, and the computer software products can be stored in computer-readable storage media, such as ROM/RAM, magnetic A disc, an optical disc, etc., includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform the methods described in various embodiments or some parts of the embodiments.

Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, but not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that it can still be The technical solutions described in the foregoing embodiments are modified, or some technical features thereof are equivalently replaced; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims (10)

1. A flow control system for a services grid, comprising:

the flow limiting subsystem is used for carrying out flow limiting control on the flow of the flow limiting service data through a token bucket flow limiting algorithm and forwarding the key service data to the service degradation subsystem;

and the service degradation subsystem is used for controlling the service level of the network bandwidth of the key service data forwarded by the flow limiting subsystem according to the service quality level to obtain the key service after flow control.

2. The flow control system for a services grid according to claim 1, wherein said system further comprises: and the micro-service resource management subsystem is used for carrying out continuous monitoring and state diagnosis on the micro-service.

3. The flow control system for a services grid according to claim 1, wherein the flow limiting subsystem comprises:

the filter module is used for filtering the received task request and analyzing the received task request according to the task request to obtain a server name and a port corresponding to the task request;

and the flow control strategy execution module is used for optimizing the token bucket flow limiting algorithm and generating a strategy triggered dynamic token according to the optimized token bucket flow limiting algorithm.

4. The flow control system for a services grid according to claim 1, wherein said service downgrading subsystem comprises:

the forwarding control module is used for storing a task request into a request thread pool, and forwarding the task request to a target service according to configuration information and the task request so as to obtain a service instance corresponding to the task request, or directly obtaining the service instance corresponding to the task request according to the configuration information and the task request;

and the data compression degradation module is used for performing data compression degradation processing on the key service data according to the network bandwidth, the service instance CPU and the memory information to obtain the key service data after data compression.

5. The flow control system for a services grid according to claim 2, wherein said micro-service resource management subsystem comprises:

the service configuration management module is used for providing an operation interface for a user to deploy business service and edit service configuration information;

the service monitoring module is used for monitoring the resource condition of the host node and the micro-service;

and the service state management module is used for monitoring the service existing in the service grid and diagnosing the corresponding service and pod running states.

6. A method of flow control for a services grid, comprising:

acquiring service data corresponding to a task request, analyzing and judging the service data, and if the service data is current-limiting service data, performing current-limiting control on the flow of the current-limiting service data through a token bucket current-limiting algorithm;

and if the service data is key service data, performing service level control on the network bandwidth of the key service data according to the service quality level to obtain the key service after flow control.

7. The method of claim 6, wherein if the service data is key service data, performing service level control on the network bandwidth of the key service data according to a service quality level to obtain a flow-controlled key service, comprises:

obtaining a service quality compression ratio according to a service instance compression ratio formula, and performing service level control on the network bandwidth of the key service data according to a service quality level generated according to the service quality compression ratio, wherein the service instance compression ratio formula is as follows:

wherein Ratio (S)_i) Representing the compression ratio of the ith service to be degraded, B_iThe network bandwidth between the user task request end and the server end at the current moment is represented; a is₁、a₂And a₃Respectively represent weight parameters, a₁+a₂+a₃＝1；U(C_i) Indicates the CPU utilization at the present time, U (M)_i) Indicating the occupancy rate of the memory at the current time, B₀Indicating a default value of network bandwidth, U (C), between the user task request side and the server side₀) Indicates the CPU utilization default, U (M)₀) Indicating the occupancy rate default of the memory.

8. The method of claim 6, wherein before performing flow restriction control on the flow of the limited service data through a token bucket flow restriction algorithm if the service data is the limited service data, the method further comprises:

optimizing the token bucket current limiting algorithm to obtain an optimized token bucket current limiting algorithm;

and generating a strategy triggering type dynamic token according to the optimized token bucket current limiting algorithm.

9. The method of traffic control for a service grid according to claim 8, wherein said generating policy-triggered dynamic tokens according to said optimized token bucket throttling algorithm comprises:

acquiring the time of each token application, and acquiring the time difference between the current time and the previous time of the token application after the request of the next token application arrives;

obtaining the token adding number at the current moment according to the token application time difference and the token adding rate;

and obtaining the final added token number of the token bucket according to the token addable number, the maximum token bucket number and the current remaining token number.

10. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the steps of the flow control method according to any one of claims 6 to 9 are implemented when the program is executed by the processor.

Images