CN114143324A

CN114143324A - Load balancing method and device based on application market architecture

Info

Publication number: CN114143324A
Application number: CN202111255848.5A
Authority: CN
Inventors: 胡敏
Original assignee: Shanghai Zhuo You Network Technology Co ltd
Current assignee: Shanghai Zhuo You Network Technology Co ltd
Priority date: 2021-10-27
Filing date: 2021-10-27
Publication date: 2022-03-04

Abstract

The application realizes reasonable and balanced distribution of data flow and the like of externally-served services to the gateway service equipment through the balance loader and balanced distribution of flow data and the like of the same service to target service equipment under the service through the cluster registration equipment, thereby realizing load balancing and service flow distribution of the service, and further balancing the load of each gateway service equipment and each service equipment.

Description

Load balancing method and device based on application market architecture

Technical Field

The present application relates to the field of computer technologies, and in particular, to a load balancing method and device based on an application market framework.

Background

With the pace of the development of science and technology in the times, the aim of serving clients is to continuously update iteration to meet the practical functional requirements of users, and the progress is made. Through the continuous development of application markets, different functions are accumulated day by day, all the functions can be integrated in a single service, and the borne service is huge. With the increasing of the access flow of the service user, the maintenance cost of the server is also higher, the high concurrent user service access requirement cannot be supported, the service expansion is not flexible enough, the mobile phone end user experience of the application market is not good, the data flow loss is caused by the occurrence of single point failure, and the like.

In the prior art, the technology used in the current application market is a high-performance web server for external services: the NGINX server performs load balancing, but with the continuous increase of market demands, the cost of service expansion is higher and higher, the performance of a market system is poorer and poorer, more and more services need to be deployed and restarted each time the services are updated, and the risk of abnormal business accidents is higher and higher. Among them, the current services have the following disadvantages:

for example, the possibility of traffic loss is high: if single point of failure happens, the loss of data flow is easy to cause.

As another example, the business modules are unclear: all current services are mixed in an application program project, and each service module cannot be completely independently deployed.

As another example, deployment costs are large: when the services are updated, upgraded and deployed, all the service services in the whole market are available, each update may have a certain risk, and the risk value of service deployment is increased because multiple services are operated and restarted.

For another example, service scalability is poor: as code grows, the time to build and deploy also increases. Each change in functionality or repair of a defect results in the need to re-deploy the entire application service. The mode of full deployment is very time-consuming, large in influence range and high in risk, so that the frequency of online deployment of the time-length project is low.

As another example, the amount of concurrency currently supported is low: the current suite of services supports a high concurrency which is too low.

As another example, service risks are high: when the business service version updates and upgrades the service, if the service is abnormal, the whole market service is possibly affected to be paralyzed.

Disclosure of Invention

An object of the present application is to provide a load balancing method and device based on an application market architecture, which implement load balancing of services and distribution of service traffic, thereby balancing loads of each gateway service device and each service device.

According to an aspect of the present application, there is provided a load balancing method based on an application market framework, applied to a client, wherein the method includes:

sending a service access request of an external service to a load balancer, so that the load balancer responds to the service access request, screens an available gateway service device from at least one gateway service device in a load balancing manner, and forwards the service access request to the available gateway service device;

after analyzing the application data in the service access request by the available gateway service equipment to obtain a service identifier and a service routing code, acquiring at least one available service equipment corresponding to the service identifier and a network address of a routing equipment corresponding to the available service equipment from cluster registration equipment, and screening a target service equipment and a network address of a target routing equipment corresponding to the target service equipment from the at least one available service equipment by adopting a polling method;

and receiving the service data returned by the load balancer, wherein the service data is obtained by accessing the service of the target routing equipment service corresponding to the network address by the target service equipment.

Further, in the above method, the method further includes:

sending a service access request of an external service to a main domain name server so that the main domain name server carries out domain name resolution on the service access request to obtain a corresponding Internet Protocol (IP) of the external service;

if the resolution of the main domain name server fails, sending a service access request of an external service to the standby domain name server, so that the standby domain name server performs domain name resolution on the service access request to acquire a corresponding external service IP;

wherein, the sending the service access request of the external service to the load balancer includes:

and sending a service access request of the external service to the load balancer through the external service IP.

According to another aspect of the present application, there is also provided a load balancing method based on an application market architecture, applied to a load balancer, wherein the method includes:

receiving a service access request for external services sent by a client;

responding to the service access request, screening an available gateway service device from at least one gateway service device in a load balancing manner, and forwarding the service access request to the available gateway service device, so that after the available gateway service device analyzes application data in the service access request to obtain a service identifier and a service routing code, acquiring a network address of at least one available service device corresponding to the service identifier and a network address of a corresponding routing device from cluster registration equipment, and screening a network address of a target service device and a corresponding target routing device from the at least one available service device by adopting a polling method;

receiving service data returned by the available gateway service equipment, wherein the service data is obtained by accessing the service of the target routing equipment service corresponding to the network address by the target service equipment;

and returning the service data to the client.

Further, in the above method, the load balancing screening of an available gateway service device from at least one gateway service device in response to the service access request includes:

and responding to the service access request, and screening an available gateway service device from the at least one gateway service device in a load balancing manner by adopting a weighted least connection scheduling algorithm.

Further, in the above method, the method further includes:

adding new gateway service equipment, and configuring a corresponding Internet Protocol (IP), a service port and a weight value of a load participated by the service port for the new gateway service equipment, wherein an initial value of the weight value is not zero.

Further, in the above method, the method further includes:

and changing the weight value of one or more gateway service devices in all the gateway service devices.

According to another aspect of the present application, there is also provided a load balancing method based on an application market framework, applied to a gateway service device, where the method includes:

receiving a service access request for an external service sent by the load balancer;

analyzing the application data in the service access request to obtain a service identifier and a service routing code;

sending the service identifier and the service routing code to cluster registration equipment so that the cluster registration equipment determines and returns at least one available service equipment corresponding to the service identifier and a network address of routing equipment corresponding to the available service equipment according to the service routing code;

receiving the network address of at least one available service device corresponding to the service identifier and the corresponding routing device, which is returned by the cluster registration device;

screening out a target business service device and a network address of a target routing device corresponding to the target business service device from the at least one available business service device by adopting a polling method;

initiating a target access request to the target service equipment, wherein the target access request comprises a network address of the target routing equipment, so that the target service equipment accesses and returns service data of a service served by the target routing equipment corresponding to the network address;

and receiving the service data from the target service equipment and returning the service data to the load balancer so that the load balancer returns the service data to the client.

According to another aspect of the present application, there is also provided a load balancing method based on an application market architecture, applied to a cluster registration device, where the method includes:

recording network addresses of all business service equipment and corresponding routing equipment;

receiving a service identifier and a service routing code sent by available gateway service equipment; the service identifier and the service routing code are obtained by analyzing application data in a received service access request for an external service sent by the load balancer by the available gateway service equipment;

determining at least one available service device corresponding to the service identifier and a network address of a routing device corresponding to the available service device from all the service devices according to the service routing code;

sending the network address of at least one available service device corresponding to the service identifier and a routing device corresponding to the available service device to the available gateway service device, so that the available gateway service device screens out a target service device and a network address of a target routing device corresponding to the target service device from the at least one available service device by adopting a polling method, and initiates a target access request to the target service device, wherein the target access request comprises the network address of the target routing device; and the target business service equipment accesses and returns business data of the business served by the target routing equipment corresponding to the network address to the available gateway service equipment.

Further, in the above method, the method further includes:

receiving a logout request sent by a service device to be logout, wherein the logout request comprises network addresses of the service device to be logout and a routing device thereof;

and deleting the network addresses of the service equipment to be logged out and the routing equipment thereof from a local record.

Further, in the above method, the method further includes:

periodically sending a heartbeat packet to each business service device in all the business service devices;

and receiving heartbeat responses returned by one or more of all the business service devices within a preset time period, or deleting the network addresses of the business service devices which do not return the heartbeat responses and the corresponding routing devices from local records after the preset time period is exceeded.

According to another aspect of the present application, there is also provided a load balancing method based on an application market architecture, applied to a business service device, where the method includes:

receiving a target access request sent by available gateway service equipment, wherein the target access request comprises a network address of target routing equipment; the network address of the target routing equipment is determined by the target service equipment which is selected from at least one available service equipment corresponding to the service identification and received by the cluster registration equipment by the available gateway service equipment by adopting a polling method;

accessing and acquiring service data of the service of the target routing equipment corresponding to the network address;

and returning the service data to the available gateway service equipment so that the available gateway service equipment sends the service data to a balanced loader, and the balanced loader returns the service data to the client.

Further, in the above method, the method further includes:

and uploading the network addresses of the service equipment and the corresponding routing equipment to the cluster registration equipment.

Further, in the above method, the method further includes:

and sending a logout request to the cluster registration equipment, wherein the logout request comprises the network addresses of the service equipment to be logout and the routing equipment thereof, so that the cluster registration equipment deletes the network addresses of the service equipment to be logout and the routing equipment thereof.

Further, in the above method, the method further includes:

and receiving the heartbeat packet sent by the cluster registration equipment, and returning a heartbeat response to the cluster registration equipment within a preset time period.

According to another aspect of the present application, there is also provided a non-volatile storage medium having stored thereon computer-readable instructions, which, when executed by a processor, cause the processor to implement the method for load balancing based on an application market architecture as described above.

According to another aspect of the present application, there is also provided a client for load balancing based on an application market architecture, wherein the client includes:

one or more processors;

a computer-readable medium for storing one or more computer-readable instructions,

when executed by the one or more processors, cause the one or more processors to implement a method for application market architecture-based load balancing for clients as described above.

According to another aspect of the present application, there is also provided a load balancer for load balancing based on an application market architecture, wherein the load balancer includes:

one or more processors;

when executed by the one or more processors, cause the one or more processors to implement a load balancing method based on an application market architecture as described above at the load balancer end.

According to another aspect of the present application, there is also provided a gateway service device for load balancing based on an application market architecture, wherein the gateway service device includes:

one or more processors;

when executed by the one or more processors, the one or more computer-readable instructions cause the one or more processors to implement a method for load balancing based on an application market architecture, such as that described above for a gateway service device.

According to another aspect of the present application, there is also provided a cluster registration device for load balancing based on an application market architecture, where the cluster registration device includes:

one or more processors;

when executed by the one or more processors, cause the one or more processors to implement a method for application market architecture-based load balancing on a cluster-registered device side as described above.

According to another aspect of the present application, there is also provided a service device for load balancing based on an application market architecture, where the service device includes:

one or more processors;

when executed by the one or more processors, the one or more computer-readable instructions cause the one or more processors to implement the method for load balancing based on the application market architecture as described above for the business service equipment side.

Compared with the prior art, the method and the device have the advantages that an application market architecture of the client, the balancing loader, the gateway service equipment, the cluster registration equipment and the business service equipment is constructed, a business access request of external services is sent to the load balancer at the client, so that the load balancer responds to the business access request, one available gateway service equipment is screened from at least one gateway service equipment in a load balancing mode, and the business access request is forwarded to the available gateway service equipment; after analyzing the application data in the service access request by the available gateway service equipment to obtain a service identifier and a service routing code, acquiring at least one available service equipment corresponding to the service identifier and a network address of a routing equipment corresponding to the available service equipment from cluster registration equipment, and screening a target service equipment and a network address of a target routing equipment corresponding to the target service equipment from the at least one available service equipment by adopting a polling method; the available gateway service equipment initiates a target access request to the target service equipment, wherein the target access request comprises a network address of the target routing equipment, so that the target service equipment accesses and returns service data of a service served by the target routing equipment corresponding to the network address; and the available gateway service equipment receives the service data from the target service equipment and returns the service data to the load balancer, so that the load balancer returns the service data to the client. The method and the device realize reasonable and balanced distribution of data flow and the like of the externally-served service to the gateway service equipment through the balanced loader, and realize balanced distribution of the flow data and the like of the same service to the target service equipment under the service through the cluster registration equipment, thereby realizing load balance of the service and distribution of the service flow, and further balancing the load of each gateway service equipment and each service equipment.

Drawings

Other features, objects and advantages of the present application will become more apparent upon reading of the following detailed description of non-limiting embodiments thereof, made with reference to the accompanying drawings in which:

FIG. 1 is a schematic diagram illustrating an architecture of an application market framework constructed in a load balancing method based on the application market framework according to an aspect of the present application;

FIG. 2 is a schematic flow chart of a load balancing method based on an application market architecture in an actual application scenario according to an aspect of the present application;

FIG. 3 illustrates a device interaction diagram of a load balancing methodology based on an application marketplace architecture, in accordance with an aspect of the subject application;

FIG. 4 illustrates a schematic diagram of service registration in a load balancing methodology based on an application marketplace architecture, according to an aspect of the subject application;

FIG. 5 illustrates a schematic diagram of service deregistration in a load balancing methodology based on an application marketplace architecture, according to one aspect of the subject application;

FIG. 6 illustrates a schematic diagram of a heartbeat mechanism in a load balancing method based on an application market framework according to an aspect of the present application.

The same or similar reference numbers in the drawings identify the same or similar elements.

Detailed Description

The present application is described in further detail below with reference to the attached figures.

In a typical configuration of the present application, the terminal, the device serving the network, and the trusted party each include one or more processors (CPUs), input/output interfaces, network interfaces, and memory.

The memory may include forms of volatile memory in a computer readable medium, Random Access Memory (RAM) and/or non-volatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM). Memory is an example of a computer-readable medium.

Computer-readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), Read Only Memory (ROM), Electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), Digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include non-transitory computer readable media (transient media), such as modulated data signals and carrier waves.

As shown in fig. 1, the present application constructs a system schematic diagram of Load balancing of an application market architecture, where the application market architecture includes a Client (e.g., a Mobile Client), a Load balancer (Elastic Load Balance, ELB), a gateway service device (corresponding to gateway 1, gateway 2, … …, gateway N, etc. in fig. 1), a cluster registration device (corresponding to the registration center in fig. 1), and a business service device (corresponding to API a-1, API a-2, … …, API a-N, API ib-1, API B-2, … …, API B-N, … …, and API O-1, API O-2 … …, API O-N in fig. 1), where N is a positive integer greater than or equal to 1, and as can be seen from fig. 1, there may be one or more business service devices corresponding to different businesses, the various devices in the application marketplace architecture are further explained below:

the Client (Mobile Client) can be a Mobile phone version Client in an application market, an HD version Client in the application market, a music version Client in the application market, and the like.

The Internet is a collection of global information resources.

Elastic Load Balance (Elastic Load Balance), such as Huacheng cloud ELB, is used for automatically distributing access traffic to a plurality of cloud servers (such as gateway service equipment and the like), expanding the external service capability of an application system and realizing higher-level application fault tolerance.

The gateway service equipment belongs to a core API treatment service center reconstructed by the application market architecture, and the functional logic of the gateway service equipment is simple and clear. The method provides external services for the mobile phone client, filters invalid access, withstands illegal requests, and calls internal business services in a load balancing manner, and the internal services can access and acquire data only through gateway services, so that the safety of the internal business services is better protected and the internal business services are prevented from being invaded.

Cluster registry equipment (namely an Etcd registry), wherein the Etcd aims to construct a highly available distributed key value (key-value) database; the interior of the Etcd adopts a raft protocol as a consistency algorithm, and the Etcd is realized based on Go language; according to benchmark data provided by the official, the Etcd single instance supports 2k + fast reading operation per second; the Etcd adopts a raft algorithm to realize the availability and consistency of data of a distributed system and ensure the reliability of the data; SSL certificate verification is supported, and safety is guaranteed. Zookeeper and etcd can both be registration and discovery for services, but etcd provides HTTP API interactions and is much simpler to use than Zookeeper.

The service equipment (namely Rest API service) comprises an APP marker service, a Report data service and a low-old Java service, wherein the APP marker service is an application market APP service, and the two servers are deployed; report data service is data service such as exposure, click, download and the like applied to market, and is deployed by two servers; the low-old Java service is an application market low-version old service overall system, namely an application market service system before market architecture reconstruction, and is deployed by two servers.

The application market architecture constructed in the embodiment of the application is a distributed system, the load balance of local services is realized by skillfully using Etcd cluster service discovery and registration, the high-performance server is not purchased with high cost, the existing equipment resources are utilized, the loss of data flow caused by single-point failure of the server is effectively avoided, and the problems of overlarge data flow and overweight network load are solved. The specific work flow of the application market architecture constructed by the embodiment of the application is shown in figures 2 and 3, wherein, fig. 2 is a schematic diagram of an actual application scenario, fig. 3 is a schematic diagram of an interaction flow of a load balancing method based on an application market architecture according to an aspect of the present application, the method comprises a client, an elastic load balance (such as cloud ELB), a gateway service device (gateway API service), a cluster registration device, a business service device and the client, the client may be a mobile phone client or a mobile terminal, the following description describes an embodiment in which the client is preferably a mobile phone client, the method comprises a step S11, a step S12, a step S13, a step S14, a step S15, a step S21, a step S22, a step S23, a step S24, a step S31, a step S32 and a step S33, and specifically comprises the following steps:

in step S11, the client sends a service access request for an external service to the load balancer.

And step S21, the load balancer receives the service access request for the external service sent by the client.

Step S22, the load balancer responds to the service access request, and selects an available gateway service device from at least one gateway service device for load balancing, and forwards the service access request to the available gateway service device.

Here, when the client accesses the market, the load balancer ELB selects an available gateway service device in a balanced manner, so that the available gateway service device determines which different market version type the service access request comes from according to different routing information, and realizes the respective load balancing of the application markets of different versions, where the specific implementation codes are as follows:

g.POST("/",httpHand.DecodeReqM(),market.MarketApi)

g.POST("/ls",httpHand.DecodeReqMLs(),market.MarketApiLs)

g.POST("/hd",httpHand.DecodeReqMHd(),market.MarketApiHd)。

in step S31, the available gateway service device receives a service access request for an external service sent by the load balancer.

Step S32, the available gateway service device parses the application data in the service access request to obtain a service identifier and a service routing code.

Step S33, the available gateway service device sends the service identifier and the service routing code to a cluster registration device, so that the cluster registration device determines and returns at least one available service device corresponding to the service identifier and the network address of the routing device corresponding to the available service device according to the service routing code;

step S41, the cluster register device records all the service devices and the network addresses of the corresponding routing devices;

step S42, the cluster register device receives the service mark and the service route code sent by the available gateway service device; the service identifier and the service routing code are obtained by analyzing application data in a received service access request for an external service sent by the load balancer by the available gateway service equipment;

step S43, the cluster registration device determines at least one available service device corresponding to the service identifier and the network address of the corresponding routing device from all the service devices according to the service routing code; for example, the available gateway service connects to a cluster registration device: the etcd registration center obtains the service identifier and the corresponding service routing code through decryption analysis of the request data, and sends the service identifier and the service routing code to the cluster registration equipment; and the cluster registration equipment queries the cluster registration equipment according to the service routing code: and the etcd registration center stores an IP list of the intranet server and a network address of the routing equipment, wherein the intranet server corresponds to the available one or more API service services corresponding to the service identifier.

Step S44, the cluster registration device sends the network address of at least one available service device corresponding to the service identifier and its corresponding routing device to the available gateway service device, so that the available gateway service device screens out the network address of a target service device and its corresponding target routing device from the at least one available service device by using a polling method, and initiates a target access request to the target service device, where the target access request includes the network address of the target routing device; the target business service equipment accesses and returns business data of business served by the target routing equipment corresponding to the network address to the available gateway service equipment;

step S34, the available gateway service device receives the network address of at least one available service device corresponding to the service identifier and its corresponding routing device, which is returned by the cluster registration device;

step S35, the available gateway service device screens out a target service device and the network address of the corresponding target routing device from the at least one available service device by a polling method; for example, the available gateway service device screens out a target service device and a network address of a target routing device corresponding to the target service device from at least one available service device acquired from the cluster registration device by using a random polling method, where the specific implementation code is as follows:

step S36, the available gateway service device initiates a target access request to the target service device, where the target access request includes the network address of the target routing device, so that the target service device accesses and returns service data of a service served by the target routing device corresponding to the network address;

in step S51, the target service device receives a target access request sent by an available gateway service device, where the target access request includes a network address of a target routing device; the network address of the target routing equipment is determined by the target service equipment which is selected from at least one available service equipment corresponding to the service identification and received by the cluster registration equipment by the available gateway service equipment by adopting a polling method;

in step S52, the target service device accesses and obtains service data of a service served by the target routing device corresponding to the network address; here, the service device accesses and obtains the service data of the service of the target routing device corresponding to the network address through the hypertext transfer protocol HTTPS, and since only the services supporting the HTTPS access can be independent service services, the development language of the service services is not limited by the architectural service at all, the existing resources can be used by force according to the actual conditions of the service and the related personnel, the service is expanded quickly and efficiently, and the implementation code for specifically obtaining the service data is as follows:

in step S53, the target service device returns the service data to the available gateway service device, so that the available gateway service device sends the service data to the balancing loader, and the balancing loader returns the service data to the client.

Step S37, the available gateway service device receives the traffic data from the target traffic service device and returns the traffic data to the load balancer, so that the load balancer returns the traffic data to the client;

step S23, the load balancer receives service data returned by the available gateway service device, where the service data is obtained by the target service device accessing a service served by a target routing device corresponding to the network address;

step S24, the load balancer returns the service data to the client.

Step S12, the client receives the service data returned by the load balancer, where the service data is obtained by the target service device accessing the service served by the target routing device corresponding to the network address.

Through the steps S11 to S12, S21 to S24, S31 to S37, S41 to S44, and S51 to S53, it is realized that data traffic and the like of the service to the outside are reasonably and equally distributed to the gateway service device by the balancing loader, and traffic data and the like of the same service are equally distributed to the target service device under the service by the cluster registration device, so that load balancing and traffic distribution of the service are realized, and loads of each gateway service device and each service device are balanced.

In the load balancing method based on the application market architecture provided in the embodiment of the application, the service system architecture is constructed by changing the NGINX + single architecture into the load balancer ELB + Gin framework + etcd cluster + distributed service framework, so that the service system architecture can be split into a plurality of completely independent service projects according to the service module, and different types of services correspond to different service devices; the external service can balance the data traffic of the service by adopting a load balancer, and the access request corresponding to the same service is subjected to load balancing distribution and application traffic to different service devices by using a polling scheduling method in the gateway service device. In the embodiment of the application, the load balancing of the gateway service is realized through the load balancer, and as the access of the application flow increases, the existing network structure does not need to be changed, the existing service does not need to be stopped, and only a new gateway service device needs to be simply added into a service group of the gateway service device, so that the flexible service expansion can be realized, and the capacity of coping with the external service is improved. In the embodiment of the application, the cluster registration device is also used skillfully: the registration of the etcd registration center realizes the local load balance among the RESTAPI business services of a plurality of business service devices corresponding to each business, and the loss of data flow caused by single-point failure of a server can be effectively avoided only by utilizing the existing device resources, so that the problems of overlarge data flow and overlarge network load are generally solved; meanwhile, the system has a balance strategy with various forms to reasonably and uniformly distribute data flow to each server; if the existing server needs to be upgraded and expanded, the existing network structure is not required to be changed, the existing service is not required to be stopped, and only a new gateway service device needs to be simply added into the service group of the gateway service device.

Next to the foregoing embodiments of the present application, an application market architecture-based load balancing method for a client according to an aspect of the present application further includes:

sending a service access request of an external service to a main Domain Name server, so that the main Domain Name server performs Domain Name resolution on the service access request to acquire a corresponding Internet Protocol (IP) of the external service, wherein the main Domain Name server can perform Domain Name resolution on the service access request through a Domain Name System (DNS); if the resolution of the main domain name server fails, starting a standby domain name server, and sending a service access request of an external service to the standby domain name server, so that the standby domain name server performs domain name resolution on the service access request through a domain name system to acquire a corresponding external service IP; the sending of the service access request of the external service to the load balancer can specifically be realized by sending the service access request of the external service to the load balancer through the external service IP, so that the external service of the service is satisfied through the load balancer.

The domain name server adopts the main domain name server and the standby domain name server, so that in the embodiment of the application, if the main domain name server is not accessed, the standby domain name server can be switched to access, and the problem that the normal access service of a user is influenced due to the failure of domain name system DNS resolution is solved to a certain extent.

Next to the foregoing embodiment of the present application, the step S22, in response to the service access request, of screening an available gateway service device from at least one gateway service device in a load balancing manner, specifically includes: and the load balancer responds to the service access request, and selects an available gateway service device from the at least one gateway service device in a load balancing mode by adopting a weighted least connection scheduling algorithm. For example, the extranet service IP may send a service access request for an extranet service through the load balancer ELB, and through a weighted minimum connection scheduling algorithm of the load balancer, the load balancing selects one available gateway service device from one or more gateway service devices, where the weighted minimum connection scheduling algorithm is that each service node may use a corresponding weight value to represent the processing capability of the service node, and a system administrator may dynamically set the corresponding weight value, where the default weight value is 1, and the weighted minimum connection scheduling makes the number of established connections of the service node and the weight value thereof as proportional as possible when allocating a new connection request. Meanwhile, a domain name walking load balancer ELB can be configured in a background to be mapped to internet protocol IPs, service ports and weight values of loads participated by the service ports, if the weight value of a certain cloud service IP is configured to be 0, the cloud service IP does not participate in external operation services, and the reduction or the upgrade of the services can be realized by changing the weight value and directly setting the weight value to be 0 under the condition of not influencing the access and use of any existing client.

Following the foregoing embodiment of the present application, an application market architecture-based load balancing method at a load balancer end according to an aspect of the present application further includes:

For example, if the traffic flow is greatly increased and capacity expansion is required, new gateway service equipment is additionally deployed to run a new gateway service through the additional new gateway service equipment, and a corresponding internet protocol IP and a service port and a weight value of a load participating in the service port are configured for the new gateway service equipment in a load balancer ELB background.

and changing the weight value of one or more gateway service devices in all the gateway service devices. For example, if the traffic is busy, the weight values of multiple gateway service devices in all the gateway service devices may be increased, so that the gateway service devices with increased weight values can share the load; if the business is idle, reducing the weight values of one or more gateway service devices in all the gateway service devices, so that fewer gateway service devices participate in the business service; if there is a prohibition on running of a certain service (for example, upgrading the service), the weight value of the gateway service device corresponding to the service that is prohibited from running may be set to 0, so that the weight value of one or more gateway service devices in all gateway service devices may be changed according to the requirement of load balancing.

Next to the foregoing embodiment of the present application, an application market architecture-based load balancing method for a service device in an aspect of the present application further includes:

and uploading the network addresses of the service equipment and the corresponding routing equipment to the cluster registration equipment. As shown in fig. 4, the service device starts service, and establishes a link with the cluster registration device Etcd, that is, the service device is linked to the cluster registration device Etcd, that is, a registration center, and traverses the service device and its corresponding route, and uploads the network addresses of the service device (corresponding to the server IP) and its corresponding route device to the cluster registration device Etcd together, so that the cluster registration device can record the network addresses of all the service devices and their corresponding route devices, and registration and upload of the service devices corresponding to all the service services are realized.

Following the foregoing embodiment of the present application, an application market architecture-based load balancing method according to an aspect of the present application further includes:

the service equipment sends a logout request to the cluster registration equipment, wherein the logout request comprises network addresses of the service equipment to be logout and the routing equipment thereof, so that the cluster registration equipment deletes the network addresses of the service equipment to be logout and the routing equipment thereof;

the cluster registration equipment receives a logout request sent by service equipment to be logout, wherein the logout request comprises network addresses of the service equipment to be logout and routing equipment thereof, and deletes the network addresses of the service equipment to be logout and the routing equipment thereof from a local record.

As shown in fig. 5, when a service needs to be restarted, an internal suspension service interface provided by the service is called, and a logout request is sent to a cluster registration device Etcd through the internal suspension service, where the logout request includes network addresses of a service device to be logout and a routing device thereof, so that the cluster registration device Etcd suspends deletion of the network addresses of the service device to be logout and the routing device thereof. The gateway service is changed by the business service registration information, and the information of the suspended business server can be deleted in time. The gateway service equipment can not provide the suspended service to the external request, and the gateway service equipment is seamlessly integrated with the elastic telescopic service, so that the load distribution is automatically expanded according to the service flow, and the flexibility and the availability of the service can be ensured. Until the gateway service device corresponding to the suspended service is restarted and the routing information such as the network address of the routing device is registered in the cluster registration device Etcd cluster, the corresponding service device will not start the external service provision again, and the specific implementation code is as follows:

the cluster registration equipment periodically sends a heartbeat packet to each of all the business service equipment, and receives heartbeat responses returned by one or more of the business service equipment in all the business service equipment in a preset time period, or deletes the network addresses of the business service equipment which does not return the heartbeat responses and the corresponding routing equipment in all the business service equipment from a local record after the preset time period is exceeded;

and the service equipment receives the heartbeat packet sent by the cluster registration equipment and returns a heartbeat response to the cluster registration equipment within a preset time period.

As shown in fig. 6, the cluster registration device (corresponding to the ETCD registration center in fig. 6) periodically sends a heartbeat packet to each of the all service devices, so that each service device receives the heartbeat packet sent by the cluster registration device, and returns a heartbeat response to the cluster registration device within a preset time period, so that the cluster registration device receives the heartbeat response returned by one or more service devices of the all service devices within the preset time period, and the cluster registration device regularly sends the heartbeat packet, so as to timely ensure that all the service devices in the cluster registration device are available service devices, and simultaneously, monitor the health status of the service devices. Or, after the preset time period is exceeded, the cluster registration device deletes the network addresses of the service devices which do not return the heartbeat response and the corresponding routing devices from the local records, so as to delete the service devices which do not perform the service, and save the storage resources.

For example, each gateway service equipment API service uses the monitoring thread invoke of the cluster registration equipment Etcd (to find the configuration information of the service equipment RESTAPI of any changed node in time. matching with the node information configuration stored in the current service system, the service monitoring state is monitored in real time, so as to ensure the availability of the service equipment recorded in the cluster registration equipment Etcd, elastically stretch the seamless integration of local internal services, and achieve the purpose of automatic expansion, if there is a new node configuration, the new node configuration is inserted into the system memory node information configuration, so as to achieve automatic load expansion, if the information configuration of the service equipment corresponding to the detected node is not consistent, the node information configuration of the service equipment recorded in the system is updated, if the node of the service equipment needs to be deleted, the node information configuration is directly removed from the service system memory, the specific implementation code is as follows:

through the embodiment of the application, the API request speed of the business service equipment accessing the market is well optimized, the average response time is within 60ms, and compared with the old server in the prior art, the API request speed is improved by 10-40 times; the new application market framework adopted by the application has strong performance and the concurrency is improved by more than thousand times, so that the massive business access requirements of users are met; the business modules corresponding to all the business service equipment are independent, and the influence range is minimized when abnormal faults occur; the method is easy to flexibly expand new business service equipment to provide new business services, thereby realizing the capacity expansion of mass business; the service is highly available, double-activity disaster tolerance is supported, and seamless real-time switching is realized.

one or more processors;

Here, for details of each embodiment in the client based on the load balancing of the application market architecture, reference may be specifically made to corresponding portions of the embodiments of the load balancing method based on the application market architecture of the client, and details are not described here again.

one or more processors;

when executed by the one or more processors, cause the one or more processors to implement an application market architecture-based load balancing method such as the load balancer described above.

Here, for details of each embodiment of the load balancer based on the application market architecture, reference may be made to corresponding parts of the embodiment of the load balancer based on the application market architecture, and details are not described herein again.

one or more processors;

Here, for details of each embodiment in the gateway service device based on load balancing of the application market architecture, reference may be specifically made to corresponding parts of the embodiment of the load balancing method based on the application market architecture at the gateway service device side, and details are not described here again.

one or more processors;

Here, for details of each embodiment in the cluster registration device based on load balancing of the application market architecture, reference may be specifically made to a corresponding part of the embodiment of the load balancing method based on the application market architecture at the cluster registration device side, and details are not described here.

one or more processors;

Here, for details of each embodiment in the service device based on load balancing of the application market architecture, reference may be specifically made to a corresponding part of the embodiment of the load balancing method based on the application market architecture at the service device side, and details are not described here again.

To sum up, the application market architecture of the client, the balanced load device, the gateway service device, the cluster registration device and the service device is constructed, a service access request of external service is sent to the load balancer at the client, so that the load balancer responds to the service access request, selects one available gateway service device from at least one gateway service device in a load balanced manner, and forwards the service access request to the available gateway service device; after analyzing the application data in the service access request by the available gateway service equipment to obtain a service identifier and a service routing code, acquiring at least one available service equipment corresponding to the service identifier and a network address of a routing equipment corresponding to the available service equipment from cluster registration equipment, and screening a target service equipment and a network address of a target routing equipment corresponding to the target service equipment from the at least one available service equipment by adopting a polling method; the available gateway service equipment initiates a target access request to the target service equipment, wherein the target access request comprises a network address of the target routing equipment, so that the target service equipment accesses and returns service data of a service served by the target routing equipment corresponding to the network address; and the available gateway service equipment receives the service data from the target service equipment and returns the service data to the load balancer, so that the load balancer returns the service data to the client. The method and the device realize reasonable and balanced distribution of data flow and the like of the externally-served service to the gateway service equipment through the balanced loader, and realize balanced distribution of the flow data and the like of the same service to the target service equipment under the service through the cluster registration equipment, thereby realizing load balance of the service and distribution of the service flow, and further balancing the load of each gateway service equipment and each service equipment.

It should be noted that the present application may be implemented in software and/or a combination of software and hardware, for example, implemented using Application Specific Integrated Circuits (ASICs), general purpose computers or any other similar hardware devices. In one embodiment, the software programs of the present application may be executed by a processor to implement the steps or functions described above. Likewise, the software programs (including associated data structures) of the present application may be stored in a computer readable recording medium, such as RAM memory, magnetic or optical drive or diskette and the like. Additionally, some of the steps or functions of the present application may be implemented in hardware, for example, as circuitry that cooperates with the processor to perform various steps or functions.

In addition, some of the present application may be implemented as a computer program product, such as computer program instructions, which when executed by a computer, may invoke or provide methods and/or techniques in accordance with the present application through the operation of the computer. Program instructions which invoke the methods of the present application may be stored on a fixed or removable recording medium and/or transmitted via a data stream on a broadcast or other signal-bearing medium and/or stored within a working memory of a computer device operating in accordance with the program instructions. An embodiment according to the present application comprises an apparatus comprising a memory for storing computer program instructions and a processor for executing the program instructions, wherein the computer program instructions, when executed by the processor, trigger the apparatus to perform a method and/or a solution according to the aforementioned embodiments of the present application.

It will be evident to those skilled in the art that the present application is not limited to the details of the foregoing illustrative embodiments, and that the present application may be embodied in other specific forms without departing from the spirit or essential attributes thereof. The present embodiments are therefore to be considered in all respects as illustrative and not restrictive, the scope of the application being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any reference sign in a claim should not be construed as limiting the claim concerned. Furthermore, it is obvious that the word "comprising" does not exclude other elements or steps, and the singular does not exclude the plural. A plurality of units or means recited in the apparatus claims may also be implemented by one unit or means in software or hardware. The terms first, second, etc. are used to denote names, but not any particular order.

Claims

1. A load balancing method based on an application market architecture is applied to a client, wherein the method comprises the following steps:

2. The method of claim 1, wherein the method further comprises:

3. A load balancing method based on an application market architecture is applied to a load balancer, wherein the method comprises the following steps:

receiving a service access request for external services sent by a client;

and returning the service data to the client.

4. The method of claim 3, wherein the load balancing screening of an available gateway service device from at least one gateway service device in response to the service access request comprises:

5. The method of claim 3, wherein the method further comprises:

6. The method of claim 3 or 5, wherein the method further comprises:

7. A load balancing method based on an application market architecture is applied to gateway service equipment, wherein the method comprises the following steps:

8. A load balancing method based on an application market architecture is applied to cluster registration equipment, wherein the method comprises the following steps:

9. The method of claim 8, wherein the method further comprises:

10. The method of claim 8, wherein the method further comprises:

11. A load balancing method based on an application market architecture is applied to business service equipment, wherein the method comprises the following steps:

12. The method of claim 11, wherein the method further comprises:

13. The method of claim 11, wherein the method further comprises:

14. The method of claim 11, wherein the method further comprises:

15. A non-transitory storage medium having stored thereon computer readable instructions which, when executed by a processor, cause the processor to implement the method of any one of claims 1 to 14.

16. A client for load balancing based on an application market architecture, wherein the client comprises:

one or more processors;

when executed by the one or more processors, cause the one or more processors to implement the method of claim 1 or 2.

17. A load balancer for load balancing based on an application market architecture, wherein the load balancer comprises:

one or more processors;

the one or more computer-readable instructions, when executed by the one or more processors, cause the one or more processors to implement the method of any one of claims 3-6.

18. A gateway service device for load balancing based on an application market architecture, wherein the gateway service device comprises:

one or more processors;

when executed by the one or more processors, cause the one or more processors to implement the method of claim 7.

19. A cluster registration device for load balancing based on an application market architecture, wherein the cluster registration device comprises:

one or more processors;

when executed by the one or more processors, cause the one or more processors to implement the method of any one of claims 8-10.

20. A business service device based on load balancing of an application market architecture, wherein the business service device comprises:

one or more processors;

when executed by the one or more processors, cause the one or more processors to implement the method of any one of claims 11-14.