WO2023050901A1

WO2023050901A1 - Load balancing method and apparatus, device, computer storage medium and program

Info

Publication number: WO2023050901A1
Application number: PCT/CN2022/099587
Authority: WO
Inventors: 杨利国; 欧阳宇腾; 邱路达
Original assignee: 深圳前海微众银行股份有限公司
Priority date: 2021-09-30
Filing date: 2022-06-17
Publication date: 2023-04-06
Also published as: CN113810304A

Abstract

Embodiments of the present application provide a load balancing method and apparatus, an electronic device, a computer storage medium and a computer program product. The method comprises: after receiving a connection request sent by each client among a plurality of clients, acquiring a load value of each server in a current network; determining a server list of each client according to the load value of each server, and issuing a corresponding server list to each client, the server list comprising at least two servers; after determining that each client establishes a TCP long connection with each server in the corresponding server list, reacquiring a load value of each server in the current network; according to the reacquired load value of each server, determining a target server that does not meet a set load balancing condition; and finding a target client connected to the target server, and adjusting the server list of the target client, the target client being one client among the plurality of clients. The method may effectively solve the problem of server overload.

Description

A load balancing method, device, equipment, computer storage medium and program

Cross References to Related Applications

This application is based on a Chinese patent application with application number 202111165088.9 and an application date of September 30, 2021. The applicant is Shenzhen Qianhai WeBank Co., Ltd., and the application name is "a load balancing method, device, equipment and computer storage medium” and claim the priority of this Chinese patent application. The entire content of this Chinese patent application is hereby incorporated into this application as a reference.

technical field

The present application relates to the technical field of cloud computing of financial technology (Fintech), involving but not limited to a load balancing method, device, electronic equipment, computer storage medium and computer program product.

Background technique

With the development of computer technology, more and more technologies are applied in the financial field, and the traditional financial industry is gradually transforming into financial technology. However, due to the security and real-time requirements of the financial industry, higher requirements are also placed on technology.

In related technologies, for load balancing, there are several typical load balancing schemes such as hardware-based F5 equipment, software-based Nginx (engine x), and Linux Virtual Server (Linux Virtual Server, LVS); In the case of Transmission Control Protocol (TCP) long connections, it is difficult to ensure that the amount of messages processed by each server node behind the load balancer is relatively balanced; Protocol (Internet Protocol, IP) is sent to the same processing server node at the same time, causing the problem of server overload, which in turn affects the processing capacity of the entire server cluster.

Contents of the invention

The present application provides a load balancing method, device, electronic equipment, computer storage medium, and computer program product, which can solve the problem of server overload in the related art when transmitting messages based on TCP long connections.

The technical scheme of the present application is realized like this:

An embodiment of the present application provides a load balancing method, the method comprising:

After receiving the connection request sent by each client in the multiple clients, obtain the load value of each server in the current network; determine the server list of each client according to the load value of each server, and send to the Each client sends a corresponding server list; the server list includes at least two servers;

After determining that each client establishes a TCP long connection with each server in the corresponding server list, reacquire the load value of each server in the current network; according to the reacquired load value of each server, determine whether The target server that satisfies the set load balancing conditions;

Searching for a target client connected to the target server, and adjusting the server list of the target client; the target client is one of the multiple clients.

In some embodiments, the searching for a target client connected to the target server includes:

Obtain at least two clients establishing TCP long connections with the target server;

According to the amount of messages sent by each of the at least two clients to the target server, from among the at least two clients, determine the client that sends the smallest amount of messages as the target client, or , determining any client whose message volume is less than or equal to the first set threshold as the target client.

Understandably, since the target server is a server that does not meet the set load balancing conditions in the current network, it is necessary to adjust the load value of the target server to achieve a balanced state; The amount of messages sent by each client is determined, therefore, based on the amount of messages sent by each client to the target server, the target client can be determined in a targeted manner; in addition, from each client connected to the target server , select the client with the smallest amount of messages sent or less than a certain threshold as the target client, which can effectively prevent the jitter problem during the subsequent load adjustment of the target server and ensure the stability of the system.

In some embodiments, the determining the target server that does not meet the set load balancing condition according to the load value of each server obtained again includes:

Sorting the load values of each server obtained again to obtain a sorting result;

From the sorting results, the server with the highest load value and the server with the lowest load value are obtained; when it is determined that the load difference between the server with the highest load value and the server with the lowest load value is greater than a second set threshold, The server with the highest load value is determined as the target server that does not meet the set load balancing condition; or,

Determining any server with a load value greater than a third set threshold in the sorting result as a target server that does not satisfy the set load balancing condition.

It can be seen that in the embodiment of the present application, after obtaining the load value of each server, by sorting the load value of each server, it is possible to more intuitively determine the target that does not meet the set load balancing conditions from the sorting results. server. In addition, by determining the server with the highest load value or greater than a certain threshold as the target server for subsequent load balancing adjustments, it is possible to avoid affecting the processing capacity of the entire server cluster due to the low processing capacity of the target server.

In some embodiments, the adjusting the server list of the target client includes:

The server with the highest load value in the server list of the target client is removed, and a server with the lowest current load value is added to the server list of the target client.

It can be seen that in the embodiment of the present application, by removing the server with the highest load value from the server list of the target client, the risk of overload due to the server continuing to process the business data of the target client can be reduced.

In some embodiments, the method also includes:

Before receiving the connection request sent by each client in the multiple clients, receive the load determination parameters reported by each server in the current network; the load determination parameters include the number of connections of the server, the server's last statistics At least one of the periodic processing message volume and the last time the server was assigned to the client;

The acquisition of the load value of each server in the current network includes:

The load value of each server in the current network is determined according to the load determination parameters reported by each server in the current network.

It can be seen that in this embodiment of the present application, based on the load determination parameters reported by each server in the current network, the load value of each server can be accurately obtained.

In some embodiments, when the load determination parameters include the number of connections of the server, the amount of messages processed by the server in the last statistical period, and the time when the server was allocated to the client last time, the load determination parameters further include connection The weight value of the number, the weight value of the processing message volume, and the weight value of the time parameter;

The determining the load value of each server in the current network according to the load determination parameters reported by each server in the current network includes:

According to the number of server connections reported by each server in the current network and the weight value of the number of connections, the amount of messages processed by the server in the last statistical period and the weight value of the amount of messages processed, the time when the server was assigned to the client last time and The time parameter weight value determines the load value of each server in the current network.

It can be seen that in the embodiment of the present application, by further limiting the parameters included in the load determination parameters, the load value of each server in the current network can be obtained more accurately.

In some embodiments, the method also includes:

In the current statistical period, every time the adjustment of the server list is completed, the number of adjustments will be increased by 1;

Calculating the ratio of the total number of times of adjusting the server list in the current statistical period to the total number of connections of each server in the current network to obtain the proportion of adjusted sessions in the current statistical period;

Whether to continue to adjust the server list in the current statistical period is determined according to the relationship between the adjusted session ratio and the fourth set threshold.

It can be seen that in this embodiment of the application, the number of adjustments to the server list in the current statistical cycle is controlled by adjusting the session ratio of the current statistical cycle, which can prevent the problem of poor system stability caused by large-scale server connection adjustments.

In some embodiments, the determining whether to continue adjusting the server list in the current statistical period according to the size relationship between the adjusted session ratio and the fourth set threshold includes:

When it is determined that the adjusted session proportion is greater than or equal to the fourth set threshold, waiting for the end of the current statistics period;

When it is determined that the adjusted session proportion is less than the fourth set threshold and it is determined that there is a target server that does not meet the set load balancing condition, continue to adjust the server list in the current statistical period.

It can be seen that in the embodiment of the present application, by adjusting the relationship between the session ratio and the fourth threshold, the number of adjustments to the server list in the current statistical cycle is controlled, so as to achieve load balancing in the current statistical cycle and ensure that the system stability.

The embodiment of the present application also proposes a load balancing device, the device includes a first determination module, a second determination module and an adjustment module, wherein,

The first determining module is configured to obtain the load value of each server in the current network after receiving the connection request sent by each client in the plurality of clients; determine the load value of each server according to the load value of each server A server list of the client, sending a corresponding server list to each client; the server list includes at least two servers;

The second determination module is configured to reacquire the load value of each server in the current network after determining that each client establishes a TCP long connection with each server in the corresponding server list; The load value of the server determines the target server that does not meet the set load balancing conditions;

An adjustment module, configured to search for a target client connected to the target server, and adjust the server list of the target client; the target client is one of the plurality of clients.

An embodiment of the present application provides an electronic device, the device includes a memory, a processor, and a computer program stored on the memory and operable on the processor, and the processor implements one or more of the aforementioned technologies when executing the program The load balancing method provided by the solution.

An embodiment of the present application provides a computer storage medium, where the computer storage medium stores a computer program; after the computer program is executed, the load balancing method provided by the foregoing one or more technical solutions can be implemented.

The embodiment of the present application also provides a computer program product, including computer readable code, when the computer readable code is run in the electronic device, the processor in the electronic device executes to implement the aforementioned one or more The load balancing method provided by the technical solution.

The embodiment of the present application proposes a load balancing method, device, electronic equipment, computer storage medium and computer program product. The method includes: after receiving a connection request sent by each client among multiple clients, obtaining the current network The load value of each server; according to the load value of each server, determine the server list of each client, and send the corresponding server list to each client; the server list includes at least two server; after determining that each client establishes a TCP long connection with each server in the corresponding server list, reacquire the load value of each server in the current network; according to the reacquired load value of each server , determine the target server that does not meet the set load balancing conditions; find the target client connected to the target server, adjust the server list of the target client; the target client is one of the multiple clients end.

It can be seen that in the embodiment of the present application, before the client establishes a connection with the server, by obtaining the load value of each server in the current network, the list of servers that each client is allowed to connect can be determined; that is, the client can be connected to To multiple servers with lower load values, so that the client's business volume can be sent to different servers for processing in a more balanced manner. Further, after the client establishes a connection with the server, if it is found that the target server does not meet the set load balancing conditions, then adjust the server list of the target client connected to the target server to achieve the purpose of controlling the overall load of the entire server cluster, Maintain the dynamic balance of the load value of each server to effectively solve the problem of server overload.

Description of drawings

FIG. 1a is a schematic structural diagram of a load balancing in the related art;

FIG. 1b is a schematic structural diagram of another load balancing in the related art;

FIG. 2a is a schematic flow diagram of a load balancing method in an embodiment of the present application;

FIG. 2b is a schematic structural diagram of a load balancing adjustment in an embodiment of the present application;

FIG. 2c is a schematic flow diagram of another load balancing method according to an embodiment of the present application;

FIG. 2d is a schematic structural diagram of another load balancing adjustment in the embodiment of the present application;

FIG. 3 is a schematic diagram of the composition and structure of a load balancing device according to an embodiment of the present application;

FIG. 4 is a schematic structural diagram of an electronic device provided by an embodiment of the present application.

Detailed ways

The present application will be described in further detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the embodiments provided here are only used to explain the present application, and are not intended to limit the present application. In addition, the embodiments provided below are some embodiments for implementing the application, rather than providing all the embodiments for implementing the application. In the case of no conflict, the technical solutions described in the embodiments of the application can be combined in any manner implement.

It should be noted that, in the embodiments of the present application, the term "comprising", "comprising" or any other variation thereof is intended to cover a non-exclusive inclusion, so that a method or device comprising a series of elements not only includes the explicitly stated elements, but also include other elements not explicitly listed, or also include elements inherent in implementing the method or apparatus. Without further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of additional related elements (such as steps in the method or A unit in an apparatus, for example, a unit may be part of a circuit, part of a processor, part of a program or software, etc.).

The term "and/or" in this article is just an association relationship describing associated objects, which means that there can be three kinds of relationships, for example, I and/or J, which can mean: I exists alone, I and J exist at the same time, and I exist alone J these three situations. In addition, the term "at least one" herein means any one of a variety or any combination of at least two of the more, for example, including at least one of I, J, R, may mean including from I, Any one or more elements selected from the set formed by J and R.

For example, the load balancing method provided in the embodiment of the present application includes a series of steps, but the load balancing method provided in the embodiment of the present application is not limited to the steps described. Similarly, the load balancing device provided in the embodiment of the present application includes a A series of modules, but the load balancing device provided by the embodiment of the present application is not limited to include the explicitly recorded modules, and may also include modules that need to be set up to obtain relevant task data or process based on task data.

The embodiments of the present application can be applied to a computer system composed of servers, and can operate together with many other general-purpose or special-purpose computing system environments or configurations. Here, the server may be a distributed cloud computing technology environment including a small computer system, a large computer system, and so on.

Electronic devices such as the server can realize corresponding functions through the execution of program modules. Generally, program modules may include routines, programs, objects, components, logic, data structures, etc., that perform particular tasks or implement particular abstract data types. The computer system/server can be practiced in distributed cloud computing environments where tasks are performed by remote processing devices that are linked through a communications network. In a distributed cloud computing environment, program modules may be located in both local and remote computing system storage media including storage devices.

At present, TCP-based connections include long connections and short connections. For scenarios with a large amount of messages to be processed and high real-time requirements, TCP long connections will be used to ensure low system throughput and delay.

For most application scenarios, hardware-based F5 devices, or software-based Nginx, LVS and other typical load balancing solutions can be well satisfied, but there are exceptions. If you want to use TCP long-term connections, It is difficult to ensure that the amount of messages processed by each server node behind the load balancer is relatively balanced by using the above scheme.

This is because the magnitude of data transmission for different types of businesses often varies greatly. Some business IP instances may only handle a few transactions, while some business IP instances may have hundreds of thousands, or even over 100,000 transactions. Millions of business volume, so for the IP instances of these two types of business, no matter what kind of load balancing scheduling algorithm is collected, once the load balancing scheduler makes a decision and the business IP is successfully connected to the server assigned by the load balancing scheduler, The load balancing scheduler will no longer intervene in the data exchange between the business IP and the server. Therefore, it will inevitably lead to a large difference in the amount of messages processed by the server node. This may cause many IPs with high business volume to be sent to the same processing server at the same time. The situation on the server node caused the server to be overloaded; as shown in Figure 1a, the data traffic transmitted by client 1 and client 2 is 3Kb/s, but the data traffic transmitted by client 3 and client 4 is The data traffic is 1Mb/s; the load balancing scheduler allocates according to the number of connections. Both server 1 and server 2 are allocated two connections, but the amount of messages processed by these two servers is not at the same level at all.

In related technologies, there are also mainstream solutions for load balancing based on TCP long connections, and its core idea can be referred to in Figure 1b; by introducing a reverse proxy server and establishing a connection between the reverse proxy server and each client, Keep the TCP long connection session between the client and the server, and then use a certain load balancing algorithm to forward the message to the back-end server cluster (server 1 to server n); here, the load balancing algorithms used are: average load, round-robin Inquiry load, priority group, etc.

It can be seen that although the above solution can solve the load balancing problem of the server in the case of TCP long connection, the solution has the following disadvantages:

1) When one of the servers is abnormal, the client needs to reconnect to the reverse proxy server, which will affect the business during the disconnection between the client and the server.

2) The reverse proxy server may also have problems with the capacity of the network card or the upper limit of the number of connections.

3) When the processing capacity of the reverse proxy server is not enough, it is difficult to complete the expansion of the server.

It can be seen that although the introduction of a reverse proxy server can solve the problem of server load balancing in the case of TCP long connections, the above-mentioned shortcomings make this solution difficult.

In view of the above technical problems, the following embodiments are proposed.

In some embodiments of the present application, the load balancing method can be realized by using a processor in the load balancing device, and the processor can be an Application Specific Integrated Circuit (ASIC), a Digital Signal Processor (Digital Signal Processor, DSP), digital signal processing device (Digital Signal Processing Device, DSPD), programmable logic device (Programmable Logic Device, PLD), field programmable logic gate array (Field Programmable Gate Array, FPGA), central processing unit (Central Processing Unit , CPU), controller, microcontroller, microprocessor at least one.

Figure 2a is a schematic flow diagram of a load balancing method in the embodiment of the present application, as shown in Figure 2a, the method includes the following steps:

Step 200: After receiving the connection request sent by each of the multiple clients, obtain the load value of each server in the current network; determine the server list of each client according to the load value of each server, and send each The client sends a corresponding server list; the server list includes at least two servers.

Exemplarily, the client is used to access specific business data and forward the business data to the back-end server for subsequent further processing; the server belongs to the back-end processing module and is used to process the business data forwarded by the client .

In the embodiment of the present application, the load balancing method is a method of balancing the load value of each server in the current network when the client establishes a TCP long connection with the server; Before each client establishes a TCP long connection with the server, the client does not know which servers are available in the current network; therefore, when each client needs to send business data to the server, it needs to first send a connection request to the scheduling module. In this way, The scheduling module can then receive the connection request sent by each client in the multiple clients. Here, by introducing the scheduling module, the complexity of connection management between each client and the server can be reduced; the scheduling module will be described in detail in conjunction with FIG. 2b below.

Figure 2b is a schematic structural diagram of a load balancing adjustment in the embodiment of the present application, as shown in Figure 2b, the structural diagram includes: client 1, client 2, server 1 to server n, and a scheduling module; wherein, the scheduling module One end of the scheduling module is connected to client 1 and client 2 respectively, and is responsible for receiving the connection requests sent by these two clients; the other end of the scheduling module is respectively connected to n servers in the current network to collect The data related to its own load reported by the server; client 1 is connected to server 1 and server 2, and client 2 is connected to server 2 and server 3; that is, the client can connect to multiple servers at the same time.

It can be seen that, compared with the need to introduce a reverse proxy server between the client and the server in the related art, the embodiment of the present application does not need to introduce a reverse proxy server, but each client directly connects to the corresponding server, That is, there is no need to connect to each server through the reverse proxy server, so it can solve the problem of high difficulty in implementing load balancing due to the upper limit of the network card capacity or connection number of the reverse proxy server in the related technology; in addition, the embodiment of the present application allows The client connects to multiple server nodes at the same time; in this way, when one of the servers is abnormal, the client can continue to process business through other server nodes, that is, it will not affect the business during the disconnection between the client and the abnormal server, ensuring the server's High availability; further, when a server is offline, upgraded, restarted, etc. changes or abnormalities occur, the expansion of the server can also be quickly completed in the embodiment of the application, which can solve the problem of related technologies due to the limited processing capacity of the reverse proxy server. The problem of high difficulty in server expansion.

In some embodiments, the above method may further include: before the scheduling module receives the connection request sent by each client among the plurality of clients, receiving load determination parameters reported by each server in the current network.

Exemplarily, the load determination parameter may include at least one of the number of connections of the server, the amount of messages processed by the server in the last statistical period, and the last time the server was assigned to the client; The number of connections between the server and the client during the statistical period.

Exemplarily, the statistical cycle and timing time can be preset in the server, so that the server can count the load determination parameters in each statistical cycle, and when the timing time arrives, the load determination parameters in the last statistical cycle closest to the timing time The load determination parameters are reported to the scheduling module.

Exemplarily, obtaining the load value of each server in the current network may include: the scheduling module determines the load value of each server in the current network in each statistical period according to the load determination parameters reported by each server in the current network.

In some embodiments, when the load determination parameters include the number of connections of the server, the amount of messages processed by the server in the last statistical period, and the time when the server was allocated to the client last time, the load determination parameters also include the weight value of the number of connections, Process message volume weight value and time parameter weight value.

Here, the weight value of the number of connections, the weight value of the processed message volume, and the weight value of the time parameter may be set according to actual scenarios, which are not limited in this embodiment of the present application.

Exemplarily, according to the load determination parameters reported by each server in the current network, determining the load value of each server in the current network may include: according to the number of connections of the server reported by each server in the current network and the weight value of the connection number, The amount of messages processed by the server in the last statistical cycle and the weight value of the amount of processed messages, the time when the server was assigned to the client last time and the weight value of time parameters determine the load value of each server in the current network.

Exemplarily, after the scheduling module receives the load determination parameters regularly reported by each server in the current network, it can obtain the load value Lr of each server in the last statistical period according to the server load evaluation formula (1).

Lr＝Abs(C-Min(C)-ΔC)*WC+Abs(M-Min(M)-ΔM)*WM+WT/(T-T1)%LMR (1)

Here, Abs represents the calculated absolute value; C represents the number of server connections; Min(C) represents the connection number with the smallest value among the connections of all servers in the current network; ΔC represents the tolerance value of the number of server connections. If the difference in the number of connections is less than this value, it can be considered that the number of connections on the server is balanced; WC indicates the weight of the number of connections in the server load evaluation formula, corresponding to the weight of the number of connections above. M represents the amount of messages processed by the server in the last statistical cycle; Min(M) represents the amount of messages processed with the least amount of messages among all servers in the last statistical cycle; ΔM represents the tolerance value of the amount of messages processed. If the message volume difference is less than this value, it can be considered that the messages processed by the server are balanced; WM indicates the weight value of the processed message volume in the server load evaluation formula, which corresponds to the above weight value of the processed message volume. T indicates the current system time; T1 indicates the last time the server was assigned to the client; WT indicates the weight value of the time parameter in the server load evaluation formula, corresponding to the weight value of the above time parameter; LMR indicates the time parameter in the server load evaluation The maximum value allowed in the formula.

It should be noted that the embodiment of the present application does not limit the values of the above parameters included in the formula (1), and may be set according to actual scenarios. It can be seen that when the server has more connections, the server load value is higher; when the server processes more messages, the server load value is higher; when the server was assigned to the client last time, the closer to the current time , the higher the server load value.

Exemplarily, after the scheduling module receives the connection request sent by each of the multiple clients, it can determine the number of servers that each client is allowed to connect to according to the importance of each client's business or the business volume , denoted as cn; here, when the importance of a certain client’s business or business volume is high, the client can be allowed to connect to a larger number of servers; when the importance of a certain client’s business or business volume is low , the client can be allowed to access a smaller number of servers.

In the embodiment of the present application, before the scheduling module receives the connection request sent by each of the multiple clients, it can obtain the load value of each server in the last statistical period according to the formula (1); for example, According to the load value of each server, the implementation manner of determining the server list of each client may be: the scheduling module selects cn servers with lower load values as the server list of each client according to the load value of each server; Here, the server list of each client includes at least two servers, so that when the connection between the client and one of the servers is abnormal, the high availability of the service can still be guaranteed.

Exemplarily, the server list may include multiple servers, but not all servers in the current network. In this way, the problem of too high connection number of each server caused by connecting all servers can be effectively avoided.

Exemplarily, after the scheduling module obtains the server list of each client, it sends the corresponding server list to each client; in this way, each client can establish a TCP long connection with each server in the corresponding server list.

Step 201: After determining that each client establishes a TCP long connection with each server in the corresponding server list, reacquire the load value of each server in the current network; according to the reacquired load value of each server, determine that the Set the target server for load balancing conditions.

Exemplarily, after determining that each client establishes a TCP long connection with each server in the corresponding server list, each client will poll and send business data to each server that has established a TCP long connection; the following uses a The client and two servers are taken as an example for description.

For example, when client 1 polls to send business data to server 1 and server 2 that have established a TCP long connection, once server 1 receives the business data sent by the client, it will count the load of the latest statistical period Determine the parameters, the load determination parameters include the amount of processing messages corresponding to the server connection session, the amount of processing messages corresponding to this server, the number of connections to this server, and the time when this server is assigned to client 1; then when the timing time arrives, Server 1 reports the above-mentioned load determination data to the scheduling module; in this way, the scheduling module can reacquire the load value of server 1 according to the formula (1); here, the acquisition process of server 2 load value is similar to the above-mentioned acquisition process, which is not mentioned here. Let me repeat. In this way, the scheduling module can reacquire the load value of each server in the current network.

It can be seen that during the data transmission process between the client and each server in the corresponding server list, the load value of each server changes dynamically; each server will regularly report the load determination parameters of the latest statistical cycle; thus, the scheduling The module can re-determine the current load value of each server. And this method of dynamically obtaining load determination parameters to realize load balancing can effectively avoid the processing capacity of the entire server cluster being affected due to the low processing capacity of a certain server.

In some embodiments, according to the load value of each server obtained again, determining the target server that does not meet the set load balancing condition may include: sorting the load value of each server obtained again to obtain the sorting result; Sort the results to determine the target servers that do not meet the set load balancing conditions.

Exemplarily, for the reacquired load value of each server, it can be sorted according to the order of the load value from high to low to obtain the sorting result; then, according to the sorting result, determine the target server that does not meet the set load balancing condition .

It can be seen that in the embodiment of the present application, after obtaining the load value of each server, by sorting the load value of each server, it is possible to more intuitively determine the target that does not meet the set load balancing conditions from the sorting results. server.

Exemplarily, an implementation manner of determining a target server that does not meet the set load balancing condition according to the sorting result may be: from the sorting result, obtain the server with the highest load value and the server with the lowest load value; When the load difference between the server with the highest load value and the server with the lowest load value is greater than the second set threshold, the server with the highest load value is determined as the target server that does not satisfy the set load balancing condition.

Exemplarily, after obtaining the server with the highest load value and the server with the lowest load value from the sorting results, first calculate the load difference between the server with the highest load value and the server with the lowest load value; If it is determined according to the comparison result that the load difference is greater than the second set threshold, it means that there are target servers that do not meet the set load balancing conditions, and the server with the highest load value is determined to be unsatisfied. Set the target server for load balancing conditions; on the contrary, if it is determined according to the comparison result that the load difference is less than or equal to the second set threshold, it means that there is no target server that does not meet the set load balancing conditions at this time, that is, the currently established All servers with TCP persistent connections meet the load balancing conditions.

Exemplarily, another implementation manner of determining a target server that does not meet the set load balancing condition according to the sorting result may be: determining any server whose load value is greater than the third set threshold in the sorting result as not meeting Set the target server for load balancing conditions.

Exemplarily, assuming that servers greater than the third set threshold in the sorting result include server 1 and server 2, and the load value relationship between them is server 1>server 2, then server 1 can be determined as not meeting the set load The target server of the load balancing condition may also determine the server 2 as the target server that does not meet the set load balancing condition.

Exemplarily, if the load values in the sorting results are all less than or equal to the third set threshold, it means that there is no target server that does not meet the set load balancing conditions at this time, that is, all servers currently establishing TCP long connections meet the load Equilibrium conditions.

Here, the values of the second set threshold and the third set threshold may be determined according to actual conditions, which is not limited in this embodiment of the present application.

It can be seen that in this embodiment of the present application, by determining the server with the highest load value or greater than a certain threshold as the target server for subsequent load balancing adjustments, it is possible to avoid affecting the entire server due to the low processing capacity of the target server The processing power of the cluster.

Step 202: Search for a target client connected to the target server, and adjust the server list of the target client; the target client is one of the multiple clients.

In the embodiment of this application, after the target server that does not meet the set load balancing conditions is obtained according to the above step 201, all clients that have established TCP long connections with the target server are obtained, and the server list that needs to be adjusted is searched from these clients. The target client; here, adjusting the server list represents a load balancing adjustment.

In some embodiments, searching for a target client connected to the target server may include: acquiring at least two clients that establish a TCP long connection with the target server; The message volume, from at least two clients, determines the target client.

Exemplarily, since the target server will regularly report the latest statistical period to the scheduling module when processing the message volume (business data) sent by different clients, the respective processing message volumes of all connections of the target server; The amount of messages sent by each client establishing a TCP connection. Furthermore, the scheduling module may determine the target client from multiple clients establishing TCP connections with the target server according to the amount of messages sent by each client to the target server.

Understandably, since the target server is a server that does not meet the set load balancing conditions in the current network, it is necessary to adjust the load value of the target server to achieve a balanced state; The amount of messages sent by each client is determined. Therefore, based on the amount of messages sent by each client to the target server, the target client can be determined in a targeted manner.

In some embodiments, determining the target client from the at least two clients according to the amount of messages sent from each of the at least two clients to the target server may include: according to each of the at least two clients The amount of messages sent by the client to the target server. From at least two clients, the client with the smallest amount of messages sent is determined as the target client, or any client whose amount of messages sent is less than or equal to the first set threshold determined as the target client.

It can be seen that in the embodiment of the present application, from each client connected to the target server, the client that sends the smallest amount of messages or less than a certain threshold is selected as the target client, which can effectively prevent the subsequent load on the target server The jitter problem during adjustment ensures the stability of the system.

Exemplarily, the implementation manner of determining the target client may also be: from all clients establishing connection sessions with the target server, find out the client with the smallest connection session load, and determine the client as the target client; here , the connection session load Las of each client that establishes a connection session with the target server can be determined by formula (2).

Las＝WAT/(t-t1)+MC (2)

Here, WAT means the weight value of the time parameter in the client connection session load evaluation formula; t means the time when the client connects to the target server; t1 means the time of the client's last session adjustment, that is, the last time the client connected The time of the target server; MC indicates the amount of messages that need to be processed by this connection session in the latest statistical period, that is, the amount of messages sent by the client to the target server that needs to be processed by the target server.

It should be noted that the embodiment of the present application does not limit the values of the above parameters included in the formula (2), and may be set according to actual scenarios. It can be seen from formula (2) that if the number of messages processed by the target server is more, the connection session load of the corresponding client is higher, and if the time from the latest session adjustment is closer, the connection session load of the corresponding client is higher.

Exemplarily, after the connection session load of each client that establishes a connection session with the target server is obtained according to formula (2), the client with the smallest connection session load is determined as the target client, or the connection session load is less than or equal to the set The client with the given value is identified as the target client.

In the embodiment of the present application, after the target client is obtained according to the above steps, the server list of the target client is obtained, and the server list is adjusted to realize load balancing among the servers.

In some embodiments, adjusting the server list of the target client may include: removing the server with the highest load value from the server list of the target client, and adding a server with the lowest load value to the server list of the target client.

Exemplarily, after obtaining the server list of the target client, the server with the highest load value in the server list is removed, that is, the target server in the server list is removed, and a new one with the lowest current load value is added. server.

This is because the target server is a server with a high load value in the current network, and its data processing capability is low. If it is not removed, but continues to process the business data sent by the target client, it is very likely that the target server will crash due to overload. In this case, the processing capacity of the entire server cluster is affected; in the embodiment of this application, by adjusting the connection between the target client and the target server to the connection between the target client and the server with the lowest current load value, the target server no longer processes the target client In this way, the overload problem of the target server can be effectively solved, and the possibility of the target server crashing due to overload can be reduced; in addition, since the connection session load between the target client and the target server is minimal, this adjustment method can also Effectively prevent the vibration problem during the adjustment process and ensure the stability of the system.

In some embodiments, the above method may further include: adding 1 to the number of adjustments each time the server list is adjusted within the current statistical period; calculating the total number of adjustments to the server list within the current statistical period According to the ratio of the total number of connections of each server, the adjusted session proportion of the current statistical period is obtained; according to the relationship between the adjusted session proportion and the fourth threshold value, it is determined whether to continue to adjust the server list in the current statistical period.

In this embodiment of the present application, the number of adjustments to the server list in the current statistical period is controlled by adjusting the session ratio of the current statistical period, which can prevent the problem of poor system stability caused by large-scale server connection adjustments.

Exemplarily, every time the adjustment of the server list is completed means that every time the load balancing adjustment is completed; here, in the current statistical period, every time the adjustment of the server list is completed, the adjustment number is incremented by 1; Exemplarily, the current The total number of adjustments to the server list in the statistical period is recorded as AT, and the total number of connections to each server in the current network is recorded as CT, so the proportion of adjusted sessions in the current statistical period is AT/CT. Next, it is determined whether to continue to adjust the server list in the current statistical period according to the relationship between the adjusted session proportion and the fourth set threshold.

In some embodiments, according to the size relationship between the adjusted session ratio and the fourth set threshold, determining whether to continue adjusting the server list in the current statistical period may include: determining that the adjusted session ratio is greater than or equal to the fourth set threshold , wait for the end of the current statistical period; when it is determined that the proportion of adjusted sessions is less than the fourth set threshold and it is determined that there are target servers that do not meet the set load balancing conditions, continue to adjust the server list in the current statistical period.

Exemplarily, when it is determined that the value of the adjusted session ratio AT/CT is greater than or equal to the fourth set threshold, wait for the end of the current statistical period; that is, no load balancing adjustment is performed in this message statistical period to prevent the system from appearing in a large area. Connection flapping. Conversely, when it is determined that the value of the adjusted session ratio AT/CT is less than the fourth set threshold, then evaluate again whether the load among the servers in the current network is balanced, that is, determine whether there is a server that does not meet the set load balancing conditions. If the target server is unbalanced, if it is, continue to the next round of adjustment until the entire system achieves load balance again; otherwise, it means that the current statistical cycle has achieved load balance. Here, the value of the fourth set threshold is not limited, and can be set according to the actual scene.

The embodiment of the present application proposes a load balancing method, device, electronic equipment, computer storage medium and computer program product. The method includes: after receiving a connection request sent by each client among multiple clients, obtaining The load value of each server; according to the load value of each server, determine the server list of each client, and send the corresponding server list to each client; the server list includes at least two servers; after determining each client After establishing a TCP long connection with each server in the corresponding server list, reacquire the load value of each server in the current network; according to the reacquired load value of each server, determine the target server that does not meet the set load balancing conditions; Find the target client connected to the target server, and adjust the server list of the target client; the target client is one of the multiple clients. It can be seen that in the embodiment of the present application, before the client establishes a connection with the server, by obtaining the load value of each server in the current network, the list of servers that each client is allowed to connect can be determined; that is, the client can be connected to To multiple servers with lower load values, so that the client's business volume can be sent to different servers for processing in a more balanced manner. Further, after the client establishes a connection with the server, if it is found that the target server does not meet the set load balancing conditions, then adjust the server list of the target client connected to the target server to achieve the purpose of controlling the overall load of the entire server cluster, Maintain the dynamic balance of the load value of each server to effectively solve the problem of server overload.

In order to better reflect the purpose of the present application, further description will be made on the basis of the above-mentioned embodiments of the present application.

Fig. 2c is a schematic flow diagram of another load balancing method according to the embodiment of the present application. As shown in Fig. 2c, the flow may include:

Step A1: The server reports load determination parameters.

Exemplarily, server 1 to server n respectively report to the scheduling module the load determination parameters in the latest statistical period; the load determination parameters may include the number of server connections, the amount of messages processed by the server in the last statistical period, the last time the server was The time allotted to the client and the amount of messages processed for each connection session of the server.

Step A2: The client sends a connection request to the scheduling module.

Exemplarily, before the client needs to send service data to the server, it needs to send a connection request to the scheduling module.

Step A3: The scheduling module distributes the server list to the client.

Exemplarily, the scheduling module determines the load value of each server according to the load determination parameters reported by each server, and then assigns corresponding server list, and send the corresponding server list to the client.

Step A4: The client establishes a connection with each server in the server list at the same time.

Exemplarily, the server list of the client includes the server 1 and the server 2, that is, the client establishes a TCP long connection with the server 1 and the server 2 at the same time.

Step A5: The client polls to send business data.

Exemplarily, after determining that the client establishes the TCP long connection with the server 1 and the server 2, the client sends service data to the server 1 and the server 2 that have established the TCP long connection in a polling manner.

It should be noted that when any server receives the business data sent by the client, it will count the load determination parameters of the latest statistical period, and report the above load determination data to the scheduling module at regular intervals; in this way, the scheduling module can determine the load based on the load. Data, reacquire the load value of each server in the current network; and according to the reacquired load value, determine whether the server connected to the client meets the load balancing conditions, if not, adjust the server list of the client, so that the server with the client Each server connected to the end is in a balanced state.

For example, considering the large number of clients and servers in a large-scale system, when the server receives the business data sent by the client, the server needs to count the load determination parameters of each statistical period; that is, the server must process The business data sent by the client also needs to be counted to determine the parameters of the load. It can be seen that the server will undertake relatively heavy tasks in the load balancing process; in order to improve the processing capacity of the server, a monitoring module can be added on the basis of Figure 2b; Figure 2d is a schematic structural diagram of another load balancing adjustment in the embodiment of the present application. As shown in Figure 2d, the newly added monitoring module is connected between the scheduling module and each server, so that each The statistics and monitoring of the load determination parameters of each server, and report the load determination parameters of each server obtained from the statistics to the scheduling module; then control the list of servers connected to the client through the scheduling module; that is, complete the server part through the monitoring module tasks, which can improve the processing capacity of the entire server cluster.

In the embodiment of this application, by controlling the number of servers that the client is allowed to connect, messages with a large amount of business can be sent to the server in a more balanced manner, which can reduce the occurrence of an explosion in business volume caused by certain types of business promotion, causing server Crash situation. In addition, after the client connects to multiple servers at the same time, changes to a certain server, such as going offline, upgrading, restarting, or when the network between the client and a certain server is abnormal, will not affect the normal reporting of data from the client, improving system reliability.

Fig. 3 is a schematic diagram of the composition and structure of a load balancing device according to an embodiment of the present application. As shown in Fig. 3 , the device includes a first determination module 300, a second determination module 301 and an adjustment module 302, wherein,

The first determination module 300 is configured to obtain the load value of each server in the current network after receiving the connection request sent by each client in the plurality of clients; determine the load value of each server according to the load value of each server A server list of a client, and a corresponding server list is issued to each client; the server list includes at least two servers;

The second determination module 301 is configured to reacquire the load value of each server in the current network after determining that each client establishes a transmission control protocol TCP long connection with each server in the corresponding server list; Get the load value of each server to determine the target server that does not meet the set load balancing conditions;

The adjustment module 302 is configured to search for a target client connected to the target server, and adjust the server list of the target client; the target client is one of the multiple clients.

In some embodiments, the adjustment module 302 is configured to find the target client connected to the target server, including:

In some embodiments, the second determination module 301 is configured to determine the target server that does not meet the set load balancing condition according to the re-acquired load value of each server, including:

In some embodiments, the adjustment module 302 is configured to adjust the server list of the target client, including:

In some embodiments, the device further includes a receiving module configured to:

The first determining module 300 is configured to obtain the load value of each server in the current network, including:

The first determination module 300 is configured to determine the load value of each server in the current network according to the load determination parameters reported by each server in the current network, including:

In some embodiments, the adjustment module 302 is further configured to:

In some embodiments, the adjustment module 302 is configured to determine whether to continue adjusting the server list in the current statistical period according to the relationship between the adjustment session ratio and the fourth set threshold, including:

In practical applications, the above-mentioned first determination module 300, second determination module 301, adjustment module 302 and receiving module can all be realized by a processor located in an electronic device, and the processor can be ASIC, DSP, DSPD, PLD, FPGA , CPU, controller, microcontroller, microprocessor at least one.

In addition, each functional module in this embodiment may be integrated into one processing unit, each unit may exist separately physically, or two or more units may be integrated into one unit. The above-mentioned integrated units can be implemented in the form of hardware or in the form of software function modules.

If the integrated unit is implemented in the form of a software function module and is not sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this embodiment is essentially or The contribution made by related technologies or all or part of the technical solution can be embodied in the form of software products, the computer software products are stored in a storage medium, and include several instructions to make a computer device (which can be a personal computer) , server, or network device, etc.) or a processor (processor) executes all or part of the steps of the method in this embodiment. The aforementioned storage medium includes: various media that can store program codes such as U disk, mobile hard disk, read-only memory (Read Only Memory, ROM), RAM, magnetic disk or optical disk.

Specifically, the computer program instructions corresponding to a load balancing method in this embodiment can be stored on a storage medium such as an optical disk, a hard disk, or a USB flash drive. When the computer program instructions corresponding to a load balancing method in the storage medium When read or executed by an electronic device, any one of the load balancing methods in the foregoing embodiments is implemented.

Based on the same technical concept as the foregoing embodiments, see FIG. 4 , which shows an electronic device 400 provided by an embodiment of the present application, which may include: a memory 401 and a processor 402; wherein,

memory 401 configured to store computer programs and data;

The processor 402 is configured to execute a computer program stored in a memory, so as to implement any one of the load balancing methods in the foregoing embodiments.

In practical applications, the above-mentioned memory 401 can be a volatile memory (volatile memory), such as RAM; or a non-volatile memory (non-volatile memory), such as ROM, flash memory (flash memory), hard disk (Hard Disk Drive, HDD) or solid-state drive (Solid-State Drive, SSD); or a combination of the above-mentioned types of memory, and provide instructions and data to the processor 402.

The aforementioned processor 402 may be at least one of ASIC, DSP, DSPD, PLD, FPGA, CPU, controller, microcontroller, and microprocessor. It can be understood that, for different load balancing devices, the electronic device used to implement the above processor function may also be other, which is not specifically limited in this embodiment of the present application.

The embodiment of the present application also provides a computer program product, the computer program product carries a program code, and the instructions included in the program code can be used to execute the load balancing method described in the above method embodiment, for details, please refer to the above method embodiment , which will not be repeated here.

Wherein, the above-mentioned computer program product may be specifically implemented by means of hardware, software or a combination thereof. In an optional embodiment, the computer program product is embodied as a computer storage medium, and in another optional embodiment, the computer program product is embodied as a software product, such as a software development kit (Software Development Kit, SDK) etc. wait.

In some embodiments, the functions or modules included in the device provided by the embodiments of the present application can be used to execute the methods described in the above method embodiments, and its specific implementation can refer to the descriptions of the above method embodiments. For brevity, here No longer.

The above descriptions of the various embodiments tend to emphasize the differences between the various embodiments, the same or similar points can be referred to each other, and for the sake of brevity, details are not repeated herein.

The methods disclosed in the various method embodiments provided in this application can be combined arbitrarily without conflict to obtain new method embodiments.

The features disclosed in the various product embodiments provided in this application can be combined arbitrarily without conflict to obtain new product embodiments.

The features disclosed in each method or device embodiment provided in this application can be combined arbitrarily without conflict to obtain a new method embodiment or device embodiment.

Those skilled in the art should understand that the embodiments of the present application may be provided as methods, systems, or computer program products. Accordingly, the present application may take the form of a hardware embodiment, a software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage and optical storage, etc.) having computer-usable program code embodied therein.

The present application is described with reference to flowcharts and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the present application. It should be understood that each procedure and/or block in the flowchart and/or block diagram, and a combination of procedures and/or blocks in the flowchart and/or block diagram can be realized by computer program instructions. These computer program instructions may be provided to a general purpose computer, special purpose computer, embedded processor, or processor of other programmable data processing equipment to produce a machine such that the instructions executed by the processor of the computer or other programmable data processing equipment produce a An apparatus for realizing the functions specified in one or more procedures of the flowchart and/or one or more blocks of the block diagram.

These computer program instructions can also be loaded onto a computer or other programmable data processing device, causing a series of operational steps to be performed on the computer or other programmable device to produce a computer-implemented process, thereby The instructions provide steps for implementing the functions specified in the flow chart or blocks of the flowchart and/or the block or blocks of the block diagrams.

The above are only preferred embodiments of the present application, and are not intended to limit the protection scope of the present application.

Claims

A load balancing method, the method comprising:

After receiving the connection request sent by each client in the multiple clients, obtain the load value of each server in the current network; determine the server list of each client according to the load value of each server, and send to the Each client sends a corresponding server list; the server list includes at least two servers;

After determining that each client establishes a transmission control protocol TCP long connection with each server in the corresponding server list, reacquire the load value of each server in the current network; according to the reacquired load value of each server , to determine the target server that does not meet the set load balancing conditions;

Searching for a target client connected to the target server, and adjusting the server list of the target client; the target client is one of the multiple clients.
The method according to claim 1, wherein said searching for a target client connected to said target server comprises:

Obtain at least two clients establishing TCP long connections with the target server;

According to the amount of messages sent by each of the at least two clients to the target server, from among the at least two clients, determine the client that sends the smallest amount of messages as the target client, or , determining any client whose message volume is less than or equal to the first set threshold as the target client.
The method according to claim 1, wherein, according to the load value of each server obtained again, determining the target server that does not meet the set load balancing conditions includes:

Sorting the load values of each server obtained again to obtain a sorting result;

From the sorting results, the server with the highest load value and the server with the lowest load value are obtained; when it is determined that the load difference between the server with the highest load value and the server with the lowest load value is greater than a second set threshold, The server with the highest load value is determined as the target server that does not meet the set load balancing condition; or,

Determining any server with a load value greater than a third set threshold in the sorting result as a target server that does not satisfy the set load balancing condition.
The method according to claim 1, wherein said adjusting the server list of said target client comprises:

The server with the highest load value in the server list of the target client is removed, and a server with the lowest current load value is added to the server list of the target client.
The method according to claim 1, wherein the method further comprises:

Before receiving the connection request sent by each client in the multiple clients, receive the load determination parameters reported by each server in the current network; the load determination parameters include the number of connections of the server, the server's last statistics At least one of the periodic processing message volume and the last time the server was assigned to the client;

The acquisition of the load value of each server in the current network includes:

The load value of each server in the current network is determined according to the load determination parameters reported by each server in the current network.
The method according to claim 5, wherein, when the load determination parameters include the number of connections of the server, the amount of messages processed by the server in the last statistical period, and the time when the server was assigned to the client last time, the load The determination parameters also include the weight value of the connection number, the weight value of the processing message volume, and the weight value of the time parameter;

The determining the load value of each server in the current network according to the load determination parameters reported by each server in the current network includes:

According to the number of server connections reported by each server in the current network and the weight value of the number of connections, the amount of messages processed by the server in the last statistical period and the weight value of the amount of messages processed, the time when the server was assigned to the client last time and The time parameter weight value determines the load value of each server in the current network.
The method according to any one of claims 1-6, wherein the method further comprises:

In the current statistical period, every time the adjustment of the server list is completed, the number of adjustments will be increased by 1;

Calculating the ratio of the total number of times of adjusting the server list in the current statistical period to the total number of connections of each server in the current network to obtain the proportion of adjusted sessions in the current statistical period;

Whether to continue to adjust the server list in the current statistical period is determined according to the relationship between the adjusted session ratio and the fourth set threshold.
The method according to claim 7, wherein the determining whether to continue adjusting the server list in the current statistical period according to the size relationship between the adjusted session proportion and the fourth set threshold includes:

When it is determined that the adjusted session proportion is greater than or equal to the fourth set threshold, waiting for the end of the current statistics period;

When it is determined that the adjusted session proportion is less than the fourth set threshold and it is determined that there is a target server that does not meet the set load balancing condition, continue to adjust the server list in the current statistical period.
A load balancing device, the device comprising:

The first determining module is configured to obtain the load value of each server in the current network after receiving the connection request sent by each client in the plurality of clients; determine the load value of each server according to the load value of each server A server list of the client, sending a corresponding server list to each client; the server list includes at least two servers;

The second determination module is configured to reacquire the load value of each server in the current network after determining that each client establishes a TCP long connection with each server in the corresponding server list; The load value of the server determines the target server that does not meet the set load balancing conditions;

An adjustment module, configured to search for a target client connected to the target server, and adjust the server list of the target client; the target client is one of the plurality of clients.
An electronic device, the device comprising a memory, a processor, and a computer program stored on the memory and operable on the processor, when the processor executes the program, the method described in any one of claims 1 to 8 is realized method.
A computer storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the method according to any one of claims 1 to 8 is implemented.
A computer program product, comprising computer readable codes, when the computer readable codes are run in an electronic device, a processor in the electronic device executes the method according to any one of claims 1 to 8.