WO2012000455A1

WO2012000455A1 - Client terminal and load balancing method

Info

Publication number: WO2012000455A1
Application number: PCT/CN2011/076770
Authority: WO
Inventors: 黄志辉
Original assignee: 腾讯科技（深圳）有限公司
Priority date: 2010-07-02
Filing date: 2011-07-01
Publication date: 2012-01-05
Also published as: CN102316138A

Abstract

Disclosed are a client terminal and a load balancing method relating to the field of clustering. The method comprises: the client terminal acquiring a load profile, the load profile comprising at least a service node access method (101); the client terminal sending a call request to a corresponding service node in a service cluster on the basis of the load profile, allowing the service node to receive and process the call request, and then return the processing result (102); the client terminal receiving the processing result returned by the service node (103). The client terminal comprises: an acquisition module, a request module, and a receiving module. In the present invention, the client terminal acquires the load profile, sends the call request to the corresponding service node in the service cluster on the basis of the load profile, and receives the processing result returned by the service node, that is, the service node load relies completely on the requesting client terminal, thereby indirectly achieving the load balancing of the service cluster and increasing the service capability of said cluster.

Description

Method for client and load balancing

The present invention relates to the field of clustering, and in particular, to a method for client and load balancing. Background technique

With the rapid development of large-scale Internet applications, large-scale distributed cluster services have begun to be applied to enterprises with the popularity of the network. A cluster is a collection of loosely coupled service nodes consisting of two or more servers that provide users with network services or applications. A very important issue in clusters is load balancing. Load balancing means that the service nodes in the cluster bear the corresponding load according to their computing power.

Currently, when load balancing is implemented, a load balancer is usually set on the cluster side, and load balancing is performed by the load balancer. The specific process is as follows:

The client sends a call request to the cluster providing the service. After receiving the call request from the client, the load balancer in the cluster forwards the call request to the corresponding service node according to the load condition of each service node in the cluster. The service node processes the client's call request and returns the corresponding processing result.

In the process of implementing the present invention, the inventors have found that the prior art has at least the following disadvantages:

In the prior art, load balancing is performed on the cluster side, which consumes processing resources on the cluster side and reduces the service capability of the cluster. Summary of the invention

In order to improve the service capability of the cluster, the embodiment of the present invention provides a method for client and load balancing. The technical solution is as follows:

A method of load balancing, the method comprising:

The client obtains a load configuration, where the load configuration includes at least a manner of accessing the service node;

Sending, by the client, a call request to a corresponding service node in the service cluster according to the load configuration, so that the service node receives and processes the call request, and returns a processing result;

The client receives the processing result returned by the service node.

Further, before the sending, by the client, the calling request to the corresponding service node in the service cluster according to the load configuration, the client includes:

The client loads all the service nodes in the service cluster, and establishes communication communication with each service node one by one. Road.

Further, the sending, by the client, the calling request to the corresponding service node in the service cluster according to the load configuration includes:

Determining, by the client, whether a communication channel between itself and a corresponding service node in the service cluster is valid; if valid, the client sends a call request to a corresponding service node in the service cluster according to the load configuration. .

Further, when the manner of accessing the service node included in the load configuration is a sequential mode, the sending, by the client according to the load configuration, a call request to a corresponding service node in the service cluster includes:

The client sequentially selects one service node from the service clusters in a preset order, and sends a call request to the selected service node.

Further, when the manner of accessing the service node included in the load configuration is a random mode, the sending, by the client according to the load configuration, a call request to a corresponding service node in the service cluster includes:

The client randomly selects a service node from the service cluster and sends a call request to the selected service node.

A client, the client includes:

An obtaining module, configured to obtain a load configuration, where the load configuration includes at least a manner of accessing a service node, and a requesting module, configured to send, according to the load configuration, a call request to a corresponding service node in the service cluster, to enable the service node Receiving and processing the call request, and returning the processing result;

And a receiving module, configured to receive the processing result returned by the service node.

Further, the client further includes:

The establishing module is configured to load all the service nodes in the service cluster before establishing the request module, and establish communication channels with all the service nodes one by one.

Further, the requesting module is further configured to determine whether a communication channel between itself and a corresponding service node in the service cluster is valid;

If valid, a call request is sent to the corresponding service node in the service cluster based on the load configuration. Further, when the manner of accessing the service node in the load configuration acquired by the acquiring module is a sequential mode, the requesting module is specifically configured to sequentially select a service node from the service cluster according to a preset order. The selected service node sends a call request.

Further, when the manner of accessing the service node in the load configuration acquired by the obtaining module is a random mode, the requesting module is specifically configured to randomly select a service node from the service cluster to the selected service. The node sends a call request. The beneficial effects brought by the technical solutions provided by the embodiments of the present invention are:

The client obtains the load configuration, sends a call request to the corresponding service node in the service cluster according to the load configuration, and receives the processing result returned by the service node, that is, the load of the service node is completely dependent on the requested client, and the service cluster is indirectly implemented. The load balancing improves the service capability of the cluster; and establishes a communication channel with all the service nodes in advance, and determines that the communication channel of a service node is valid, and then sends a call request to the service node to prevent load balancing on the cluster side. The exception causes the client's call request to be abnormal, effectively enhancing the robustness of the entire service cluster, and implementing the adaptive call client. DRAWINGS

In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly described. It is obvious that the drawings in the following description are only some embodiments of the present invention. Other drawings may also be obtained from those of ordinary skill in the art in view of the drawings.

1 is a flowchart of a method for load balancing provided in Embodiment 1 of the present invention;

2 is a schematic diagram of a network architecture of load balancing provided in Embodiment 1 of the present invention;

3 is an information interaction diagram of load balancing provided in Embodiment 1 of the present invention;

4 is a schematic structural diagram of a client provided in Embodiment 2 of the present invention;

FIG. 5 is a schematic structural diagram of another client provided in Embodiment 2 of the present invention. detailed description

The embodiments of the present invention will be further described in detail below with reference to the accompanying drawings.

Example 1

Referring to FIG. 1, this embodiment provides a method for load balancing, including:

101: The client obtains a load configuration, where the load configuration includes at least a manner of accessing the service node;

102: The client sends a call request to a corresponding service node in the service cluster according to the load configuration, so that the service node receives and processes the call request, and returns a processing result;

103: The client receives the processing result returned by the service node.

The method provided in this embodiment obtains a load configuration by using a client, sends a call request to a corresponding service node in the service cluster according to the load configuration, and receives a processing result returned by the service node, that is, the load of the service node is completely dependent on the requested client. Indirectly, the load balancing of the service cluster is realized indirectly, and the service capability of the cluster is improved. See the load balancing network architecture shown in Figure 2, including the client and the service cluster. The client includes several specific clients, which are indicated by the client C1, the client C2, and the client Cn. The service cluster is divided into several actual processes. The client invokes the requested service node, and indicates the service node N1, the service node N2, the service node N3, and the service node Nn.

Based on the above network architecture, see Figure 3, which details the implementation of load balancing.

201: The client loads all the service nodes in the service cluster, and establishes communication channels with all the service nodes one by one; further, for security reasons, the establishment of the communication channel generally requires authentication between the client and the service cluster to ensure the legality of the channel. In this regard, this embodiment does not limit the specific authentication mode adopted.

202: The client obtains a load configuration, where the load configuration includes at least a manner of accessing the service node.

Specifically, the load configuration may be preset on the client, or may be obtained by the client from a certain configuration server. The load configuration is used to implement the load balancing of the service cluster. It specifies the information about the access mode and frequency of the service nodes of the service cluster. The embodiment does not limit the specific content of the load configuration.

203: Optionally, the client determines whether the communication channel between the user and the corresponding service node in the service cluster is valid; the client determines that the corresponding service node is abnormal from the established communication channel failure, the service is unavailable, and the client calls instead. The service of the available service node prevents the service node from being abnormal during the load balancing on the cluster side, causing the client's call request to be abnormal. This effectively enhances the robustness of the entire service cluster and implements the adaptive call client. Correspondingly, if it is valid, go to step 204.

Further, the client may also directly perform step 204 without determining whether the communication channel is valid.

204: The client sends a call request to a corresponding service node in the service cluster according to the load configuration.

Specifically, if the manner of accessing the service node included in the load configuration is a sequential mode, the client sequentially selects one service node from the service cluster according to a preset order, and sends a call request to the selected service node. If the way to access the service node included in the load configuration is random, the client randomly selects a service node from the service cluster and sends a call request to the selected service node.

The following is an example of distributing the load amount in a sequential manner by accessing the service node as an example to illustrate the process of load balancing. For each client, the following load configuration is set, taking client C1 as an example, client C1 and service node Nl,

N2 and N3 respectively establish communication channels, according to the preset order Nl, N2, N3, the first call request of the client CI is sent to the service node Nl, and the second call request is sent to the service node N2, and the third call is made. The request is sent to the service node N3, and then the fourth call request is sent to the service node N1, the fifth call request is sent to the service node N2, and the sixth call request is sent to the service node N3, and the service cluster is implemented again and again. Load balancing requirements for each service node.

The following is an example of randomly assigning the load amount in a random manner by accessing the service node, and illustrates the process of load balancing. For each client, the following load configuration is set. Taking client C1 as an example, client C1 and service nodes N1, N2, and N3 respectively establish communication channels, and client C1 randomly selects one service node N2 from the service cluster, and Send the first call request to it, then randomly select a service node Nl from the service cluster, and send a second call request to it, and so on, and implement the load on each service node in the service cluster by random selection. Equilibrium requirements.

In addition, the request threshold can also be set in the load configuration to limit the number of requests sent by the user within a certain period of time to prevent the server from crashing due to excessive requests.

205: The service node receives and processes the call request, and returns a processing result.

Specifically, after the client loads all the service nodes in the service cluster and establishes communication channels with all the service nodes one by one, the call request sent by the client and the processing result returned by the service node can be performed through the established communication channel. Transfer.

206: The client receives the processing result returned by the service node.

The method provided in this embodiment obtains a load configuration by using a client, sends a call request to a corresponding service node in the service cluster according to the load configuration, and receives a processing result returned by the service node, that is, the load of the service node is completely dependent on the requested client. Indirectly, the load balancing of the service cluster is realized indirectly, and the service capability of the cluster is improved; and a communication channel is established in advance with all service nodes, and after determining that the communication channel of a service node is valid, a call request is sent to the cluster to prevent the cluster from being activated. When the load balancing is performed on the side, the service node is abnormal, causing the client's call request to be abnormal. This effectively enhances the robustness of the entire service cluster and implements the adaptive call client. Example 2

Referring to FIG. 4, this embodiment provides a client, including:

The obtaining module 301 is configured to obtain a load configuration, where the load configuration includes at least a manner of accessing the service node, and the requesting module 302 is configured to send a call request to the corresponding service node in the service cluster according to the load configuration, so that the service node receives and processes Call the request, return the processing result;

The receiving module 303 is configured to receive a processing result returned by the service node.

Further, referring to FIG. 5, the client further includes:

The establishing module 304 is configured to load all service nodes in the service cluster before performing the requesting module 302, and establish communication channels with each of the service nodes one by one.

Further, the requesting module 302 is further configured to determine whether the communication channel between the user and the corresponding service node in the service cluster is valid;

If valid, a call request is sent to the corresponding service node in the service cluster based on the load configuration.

When the method of accessing the service node included in the load configuration acquired by the obtaining module 301 is in a sequential manner, the request is The module 302 is specifically configured to sequentially select a service node from the service cluster according to a preset sequence, and send a call request to the selected service node.

When the manner of accessing the service node included in the load configuration acquired by the obtaining module 301 is a random mode, the requesting module 302 is specifically configured to randomly select a service node from the service cluster and send a call request to the selected service node.

The client provided in this embodiment is the same as the method embodiment, and the specific implementation process is described in the method embodiment, and details are not described herein again.

The client provided by this embodiment sends a call request to the corresponding service node in the service cluster according to the load configuration, and receives the processing result returned by the service node, that is, the load of the service node is completely dependent on the requested client. Indirectly realizes the load balancing of the service cluster and improves the service capability of the cluster; and establishes a communication channel with all the service nodes in advance, and determines that the communication channel of a service node is valid, and then sends a call request to the cluster side. When the load balancing is performed, the service node is abnormal, causing the client's call request to be abnormal. This effectively enhances the robustness of the entire service cluster and implements the adaptive call client. It should be noted that: the client provided by the foregoing embodiment is only illustrated by dividing the foregoing functional modules when performing load balancing. In actual applications, the foregoing function allocation may be completed by different functional modules as needed. The internal structure of the client is divided into different functional modules to perform all or part of the functions described above.

All or part of the steps in the embodiments of the present invention may be implemented by using hardware, or may be instructed by a program to execute related hardware, and the program may be stored in a readable storage medium, such as an optical disk or a hard disk.

The above is only the preferred embodiment of the present invention, and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., which are within the spirit and scope of the present invention, should be included in the protection of the present invention. Within the scope.

Claims

Claim

A load balancing method, the method comprising:

The client receives the processing result returned by the service node.

2. The method according to claim 1, wherein the client sends a call request to a corresponding service node in the service cluster according to the load configuration:

The client loads all service nodes in the service cluster, and establishes communication channels with all service nodes one by one.

3. The method according to claim 2, wherein the sending, by the client, a call request to a corresponding service node in the service cluster according to the load configuration comprises:

The method according to claim 1, wherein, when the manner of accessing the service node included in the load configuration is a sequential mode, the client sends a corresponding service node to the service cluster according to the load configuration. Sending a call request includes:

The method according to claim 1, wherein, when the manner of accessing the service node included in the load configuration is a random mode, the client sends a corresponding service node to the service cluster according to the load configuration. Sending a call request includes:

6. A client, wherein the client includes: An obtaining module, configured to obtain a load configuration, where the load configuration includes at least a manner of accessing a service node, and a requesting module, configured to send, according to the load configuration, a call request to a corresponding service node in the service cluster, to enable the service node Receiving and processing the call request, and returning the processing result;

The client according to claim 6, wherein the client further comprises:

The establishing module is configured to load all the service nodes in the service cluster before performing the requesting module, and establish communication channels with each of the service nodes one by one.

The client according to claim 7, wherein the requesting module is further configured to determine whether a communication channel between itself and a corresponding service node in the service cluster is valid;

The client according to claim 6, wherein the requesting module is specifically configured according to a preset order when the manner of accessing the service node in the load configuration acquired by the obtaining module is a sequential mode. A service node is selected from the service cluster, and a call request is sent to the selected service node.

The client according to claim 6, wherein, when the manner of accessing the service node in the load configuration acquired by the obtaining module is a random mode, the requesting module is specifically used to randomly obtain the service. A service node is selected in the cluster, and a call request is sent to the selected service node.