Load balancing method and device based on health examination, equipment and medium

Abstract

The invention discloses a load balancing method and device based on health examination, equipment and a medium, and relates to the technical field of cloud computing, and the method comprises the steps: a control end issues corresponding configuration instructions to a health examination end and a load balancing end based on received configuration information, enabling the load balancing end to configure at the load balancing node according to the configuration instruction, and creating a back-end server; after the health examination terminal receives a configuration instruction, determining a target health examination mode and a corresponding examination period from a plurality of self-defined health examination modes according to the configuration instruction, and performing health examination on the back-end server based on the target health examination mode and the examination period so as to send an obtained health examination result to the control terminal; and the control end adjusts the weight of the back-end server based on the health examination result and a preset adjustment strategy so as to carry out load balancing processing. Therefore, the working pressure of the load balancing system can be relieved, and the flexibility and the operation and maintenance efficiency of the system are improved.

Description

Technical Field

[0001] The present invention relates to the field of cloud computing technology, and in particular to a load balancing method, device, equipment and medium based on health check. Background Art

[0002] Due to the rapid development of technologies such as computer communication networks, network overload and overload problems occur frequently. To solve the above problems, it is necessary to use load balancing technology to provide more efficient cloud computing service capabilities. Load balancing is built on the existing network structure, which can expand the bandwidth of network devices and servers, enhance network data processing capabilities, and improve network flexibility and availability. Existing load balancing technology is usually accompanied by a health check mechanism to regularly monitor the later servers and dynamically adjust the server weights, thereby significantly improving the reliability and performance of the system. However, since the load balancer not only needs to distribute the load, but also needs to perform health checks, the load balancing node has a large workload. When replacing the load balancer, it is necessary to reconfigure or develop its health check logic, which is difficult to maintain and the replacement process is complicated. In addition, since each load balancer may perform a health check on the backend server, it may receive multiple health check requests, which has a large workload, low operation and maintenance efficiency, and a relatively single health check method, and flexibility needs to be improved.

[0003] In summary, how to alleviate the workload of the load balancing system and improve the flexibility and operation and maintenance efficiency of the system is an urgent problem to be solved. Summary of the invention

[0004] In view of this, the purpose of the present invention is to provide a load balancing method, device, equipment and medium based on health check, which can alleviate the working pressure of the load balancing system and improve the flexibility and operation and maintenance efficiency of the system. The specific scheme is as follows:

[0005] In a first aspect, the present application provides a load balancing method based on health check, which is applied to a load balancing system, wherein the load balancing system includes a control end, a health check end, and a load balancing end, and the health check end is independent of the load balancing end; wherein the method includes:

[0006] The control end sends corresponding configuration instructions to the health check end and the load balancing end respectively based on the received configuration information, so that the load balancing end performs configuration at the load balancing node according to the configuration instructions to create a backend server;

[0007] When the health check end receives the configuration instruction, it determines a target health check mode and a corresponding check period from a plurality of custom health check modes according to the configuration instruction, and performs a health check on the backend server based on the target health check mode and the check period, so as to send the obtained health check result to the control end;

[0008] The control end adjusts the weight of the back-end server based on the health check result and the preset adjustment strategy to perform load balancing processing.

[0009] Optionally, performing a health check on the backend server based on the target health check mode and the check period includes:

[0010] Determine the response time, number of packets sent, and packet loss rate threshold of the target health check method based on the configuration instruction;

[0011] Taking the inspection period as the health inspection period, and sending echo requests of the network control message protocol to the backend server in sequence based on the number of packets sent, so as to determine whether a response of the backend server to the echo request is received within the response time;

[0012] Counting the number of requests in the echo request that receive the response within the response time to determine the packet loss rate of this health check cycle;

[0013] If the packet loss rate is higher than the packet loss rate threshold, a health check result indicating that the health check status of the backend server is abnormal is obtained;

[0014] If the packet loss rate is lower than the packet loss rate threshold, the average response time of the echo request in this health check cycle is counted, and the average response time is compared with the first preset time threshold;

[0015] If the average response time is higher than the first preset time threshold, a health check result indicating that the health check status of the backend server is sub-healthy is obtained;

[0016] If the average response time is lower than the first preset time threshold, a health check result indicating that the health check status of the backend server is healthy is obtained.

[0017] Optionally, performing a health check on the backend server based on the target health check mode and the check period includes:

[0018] Determine a timeout duration of the target health check mode based on the configuration instruction;

[0019] Taking the inspection period as the health inspection period, initiating a detection of a preset transmission control protocol port to the backend server to determine whether a connection with the backend server is successfully established within the timeout period;

[0020] If the connection with the backend server fails within the timeout period, a health check result indicating that the health check status of the backend server is abnormal is obtained;

[0021] If the connection with the backend server is successfully established within the timeout period, determining the connection duration for successfully establishing the connection with the backend server, and comparing the connection duration with a second preset duration threshold;

[0022] If the connection duration is longer than the second preset duration threshold, a health check result indicating that the health check status of the backend server is sub-healthy is obtained;

[0023] If the connection duration is lower than the second preset duration threshold, a health check result indicating that the health check status of the backend server is healthy is obtained.

[0024] Optionally, performing a health check on the backend server based on the target health check mode and the check period includes:

[0025] Determine the response time, number of packets sent, and packet loss rate threshold of the target health check method based on the configuration instruction;

[0026] Taking the inspection period as the health inspection period, sending data packets to the backend server in sequence based on the number of packets sent through the user datagram protocol to determine whether a response to the data packet from the backend server is received within the response time;

[0027] Counting the number of data packets in the data packets that receive the response within the response time to determine the packet loss rate of this health check cycle;

[0028] If the packet loss rate is higher than the packet loss rate threshold, a health check result indicating that the health check status of the backend server is abnormal is obtained;

[0029] If the packet loss rate is lower than the packet loss rate threshold, and the packet loss rate is higher than a preset numerical threshold, a health check result is obtained indicating that the health check status of the backend server is sub-healthy; and the preset numerical threshold is lower than the packet loss rate threshold;

[0030] If the packet loss rate is lower than the preset numerical threshold, a health check result is obtained indicating that the health check status of the backend server is healthy.

[0031] Optionally, performing a health check on the backend server based on the target health check mode and the check period includes:

[0032] Determine the response duration, request protocol, response information and request initiation information of the target health check method based on the configuration instruction; the request protocol includes Hypertext Transfer Protocol and Hypertext Transfer Protocol Security;

[0033] Taking the inspection period as the health inspection period, sending the request protocol to the backend server based on the request initiation information to determine whether target response information consistent with the response information is received within the response time;

[0034] If the target response information is not received within the response time, a health check result indicating that the health check status of the backend server is abnormal is obtained;

[0035] If the target response information is received within the response time, the target response time for receiving the target response information is counted, and the target response time is compared with a third preset time threshold;

[0036] If the target response time is longer than the third preset time threshold, a health check result indicating that the health check status of the backend server is sub-healthy is obtained;

[0037] If the target response time is lower than the third preset time threshold, a health check result indicating that the health check status of the backend server is healthy is obtained.

[0038] Optionally, adjusting the weight of the backend server based on the health check result and a preset adjustment strategy includes:

[0039] If the health check result indicates that the health check status of the backend server is healthy, the weight of the backend server is increased by a preset amount until the weight reaches a preset maximum weight value;

[0040] If the health check result indicates that the health check status of the backend server is sub-healthy, or the health check result indicates that the health check status of the backend server is abnormal, the weight of the backend server is reduced by a preset amount until the weight reaches a preset minimum weight value.

[0041] Optionally, the health check-based load balancing method further includes:

[0042] If the health check results of a preset number of consecutive check cycles all indicate that the health check status of the backend server is abnormal, the backend server is deleted from the forwarding pool.

[0043] In a second aspect, the present application provides a load balancing device based on health check, which is applied to a load balancing system, wherein the load balancing system includes a control end, a health check end, and a load balancing end, and the health check end is independent of the load balancing end; wherein the device includes:

[0044] An instruction issuing module, used to issue corresponding configuration instructions to the health check end and the load balancing end respectively through the control end based on the received configuration information, so that the load balancing end performs configuration at the load balancing node according to the configuration instructions and creates a backend server;

[0045] A health check module, configured to determine a target health check mode and a corresponding check period from a plurality of custom health check modes according to the configuration instruction after the health check terminal receives the configuration instruction, and perform a health check on the backend server based on the target health check mode and the check period, so as to send the obtained health check result to the control terminal;

[0046] The weight adjustment module is used to adjust the weight of the back-end server based on the health check result and the preset adjustment strategy through the control end to perform load balancing processing.

[0047] In a third aspect, the present application provides an electronic device, including:

[0048] Memory, used to store computer programs;

[0049] The processor is used to execute the computer program to implement the aforementioned load balancing method based on health check.

[0050] In a fourth aspect, the present application provides a computer-readable storage medium for storing a computer program; wherein, when the computer program is executed by a processor, the aforementioned health check-based load balancing method is implemented.

[0051] In this embodiment, the control end sends corresponding configuration instructions to the health check end and the load balancing end respectively based on the received configuration information, so that the load balancing end configures the load balancing node according to the configuration instructions and creates a back-end server; when the health check end receives the configuration instruction, it determines the target health check method and the corresponding inspection cycle from several custom health check methods according to the configuration instruction, and performs a health check on the back-end server based on the target health check method and the inspection cycle, so as to send the obtained health check result to the control end; the control end adjusts the weight of the back-end server based on the health check result and the preset adjustment strategy to perform load balancing processing. As can be seen from the above, the present application first sends corresponding configuration instructions to the health check end and the load balancing end respectively based on the received configuration information through the control end, so that the load balancing end configures the load balancing node according to the configuration instructions, creates a back-end server, and after the health check end receives the configuration instructions, determines the target health check method and the corresponding inspection cycle from several custom health check methods according to the configuration instructions, performs a health check on the back-end server based on the target health check method and the inspection cycle, and sends the obtained health check result to the control end. After the control end receives the health check result, the weight of the back-end server is adjusted based on the health check result and the preset adjustment strategy to perform load balancing processing. In this way, through the above process of this application, the health check function is separated from the load balancer, so that the load balancer can focus on load distribution, reduce the working pressure of the load balancing node, and simplify the process of replacing the load balancer. At the same time, the health status of the back-end server is centrally monitored. Each health check only performs a health status detection on the back-end server once, which can reduce the number of health check requests received by the back-end server and alleviate its work pressure. In addition, a variety of custom health check methods are provided to improve the flexibility of the load balancing system and better meet the needs of users, thereby alleviating the working pressure of the load balancing system and improving the flexibility and operation and maintenance efficiency of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0053] Figure 1 A flow chart of a load balancing method based on health check disclosed in this application;

[0054] Figure 2 This is an overall architecture diagram of a load balancing method based on health check disclosed in this application;

[0055] Figure 3 A schematic diagram of the structure of a load balancing device based on health check disclosed in this application;

[0056] Figure 4 This is a structural diagram of an electronic device disclosed in this application. DETAILED DESCRIPTION

[0057] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0058] Existing load balancing technologies are usually accompanied by a health check mechanism to regularly monitor the back-end servers and dynamically adjust the server weights, thereby significantly improving the reliability and performance of the system. However, since the load balancer not only needs to distribute the load, but also needs to perform health checks, the workload of the load balancing node is relatively high, and when replacing the load balancer, it is necessary to reconfigure or develop its health check logic, which is difficult to maintain and the replacement process is complicated. In addition, since each load balancer may perform a health check on the backend server, it may receive multiple health check requests, which results in high workload and low operation and maintenance efficiency. In addition, the health check method is relatively simple and its flexibility needs to be improved.

[0059] In order to overcome the above technical problems, the present application provides a load balancing method based on health check to alleviate the working pressure of the load balancing system and improve the flexibility and operation and maintenance efficiency of the system.

[0060] See also Figure 1 As shown, an embodiment of the present invention discloses a load balancing method based on health check, which is applied to a load balancing system, wherein the load balancing system includes a control end, a health check end and a load balancing end, and the health check end is independent of the load balancing end; wherein the method includes:

[0061] Step S11: The control end sends corresponding configuration instructions to the health check end and the load balancing end based on the received configuration information, so that the load balancing end configures the load balancing node according to the configuration instructions and creates a backend server.

[0062] In this embodiment, the load balancing system first sends corresponding configuration instructions to the health check end and the load balancing end respectively through the control end based on the received configuration information, so that the load balancing end performs configuration at the load balancing node according to the configuration instructions and creates a backend server and a listener. Among them, the configuration information includes health check configuration information that needs to be sent to the health check end and load balancing configuration information that needs to be sent to the load balancing end.

[0063] It should be noted that the load balancing system of this embodiment includes a control end, a health check end and a load balancing end. Figure 2 The figure shows the overall architecture diagram of a load balancing method based on health check provided by the present application. The control module is located at the control end, responsible for sending configurations such as listeners to the load balancing module, sending health check configurations to the health check module, and dynamically adjusting the weight of the back-end server group of the load balancing module according to the received health check results; the health check module is located at the health check end, responsible for performing a health check on the load balancing module according to the received configuration, and reporting the health check results to the control module; the load balancing module is located at the load balancing end, mainly including load balancing nodes and network devices that provide multi-active, etc., mainly responsible for reverse proxy and traffic distribution functions, and is also responsible for creating listeners and back-end servers in the nginx (an open source web server and reverse proxy server) configuration file according to the received configuration, and completing the load balancing configuration. It should be further pointed out that the health check end is independent of the load balancing end, that is, the load balancing end is only responsible for the reverse proxy and traffic distribution functions, and is not responsible for the status detection of the back-end server. Specifically, a server is deployed separately in the load balancing cluster to use the health check module. Its network configuration is consistent with the configuration of the load balancing node. By receiving the health check configuration instructions issued by the control module, the corresponding timed detection task is registered to perform the health check of the backend server. In this way, this embodiment separates the load balancing end from the health check end, so that the load balancing end only needs to be responsible for the reverse proxy and traffic distribution functions. The entire system does not depend on the selection of the load balancer, has strong portability, and reduces the working pressure of the load balancing node.

[0064] Step S12: After the health check end receives the configuration instruction, it determines a target health check method and a corresponding check cycle from a number of custom health check methods according to the configuration instruction, and performs a health check on the back-end server based on the target health check method and the check cycle, so as to send the obtained health check results to the control end.

[0065] In this embodiment, after the load balancing system receives the configuration instruction at the health check end, it determines the target health check method and the corresponding check period from a number of custom health check methods according to the configuration instruction, and performs a health check on the backend server based on the target health check method and the check period, so as to send the obtained health check result to the control end. Among them, the custom health check method can be a health check method customized by the user according to his own needs. It should be pointed out that in order to obtain the health check result, this embodiment can adopt a variety of different determination methods.

[0066] In a first specific implementation, the target health check method may be a detection method using ICMP (Internet Control Message Protocol), that is, testing the reachability of the backend server by sending an ICMPEcho request (such as ping). Specifically, the health check end determines the response time, number of packets and packet loss rate threshold of the target health check method based on the configuration instruction; takes the inspection period as the health check period, and sends echo requests of the network control message protocol to the back-end server in sequence based on the number of packets to determine whether a response of the back-end server to the echo request is received within the response time; counts the number of requests in the echo request that receive the response within the response time to determine the packet loss rate of this health check period; if the packet loss rate is higher than the packet loss rate threshold, a health check result is obtained that the health check status of the back-end server is abnormal; if the packet loss rate is lower than the packet loss rate threshold, the average response time of the echo request in this health check period is counted, and the average response time is compared with a first preset time threshold; if the average response time is higher than the first preset time threshold, a health check result is obtained that the health check status of the back-end server is sub-healthy; if the average response time is lower than the first preset time threshold, a health check result is obtained that the health check status of the back-end server is healthy. That is, the response duration, number of packets and packet loss rate threshold of the target health check method are determined according to the configuration instructions, the inspection period is used as the health check period, and echo requests of the network control message protocol are sent to the back-end server in sequence based on the number of packets to determine whether a response of the back-end server to the echo request is received within the response duration. If the back-end server is able to respond to the request, it is considered that the back-end server is online, otherwise the load balancer will no longer send requests to the server.

[0067] It is understandable that after the health check end sends the echo request, the number of requests in the echo request that receive the response within the response time is counted to determine the packet loss rate of this health check cycle, and the packet loss rate is compared with the packet loss rate threshold. If it is higher than the packet loss rate threshold, it means that the backend server may have a problem, and an abnormal health check result is obtained. If it is lower than the packet loss rate threshold, it means that the backend server is normal, but further judgment is required to determine the specific status of the backend server, that is, the average response time of the echo request in this health check cycle is counted, and when the average response time is higher than the first preset time threshold, it may be due to the high current load of the server that the interface response time is too long, and a sub-healthy health check result is obtained. When the average response time is lower than the first preset time threshold, a healthy health check result is obtained. Take the inspection cycle of 30 seconds, the response time of 2 seconds, the number of packets sent is 5, the packet loss rate threshold is 20%, and the first preset time threshold is 500 milliseconds as an example. The inspection cycle is 30 seconds, that is, the status of the backend server is evaluated every 30 seconds; the response time is 2 seconds, that is, if there is no response within 2 seconds for each packet, it is considered a packet loss; the number of packets is 5, that is, 5 data packets are sent continuously, with an interval of 1 second between each data packet; the packet loss rate threshold is 20%, that is, if four or more data packets are not lost after sending five packets, the health check result is considered successful, and the average response time is recorded, otherwise it is considered a failure, and the health check result of the backend server being in an abnormal state is obtained. When the health check result is considered successful, the recorded average response time is compared with the first preset time threshold. For example, if the average response time exceeds 500 milliseconds, a sub-healthy health check result is obtained, and if the average response time is less than 500 milliseconds, a healthy health check result is obtained.

[0068] In a second specific implementation, the target health check mode may be a mode of detecting with a TCP (i.e., Transmission Control Protocol) port, that is, attempting to establish a TCP connection with a specific TCP port of a backend server. Specifically, the health check end determines the timeout duration of the target health check mode based on the configuration instruction; takes the inspection period as the health check period, initiates detection of a preset transmission control protocol port to the backend server to determine whether a connection with the backend server is successfully established within the timeout duration; if the connection with the backend server fails within the timeout duration, a health check result is obtained that the health check status of the backend server is abnormal; if the connection with the backend server is successfully established within the timeout duration, the connection duration of the successful connection with the backend server is determined, and the connection duration is compared with a second preset duration threshold; if the connection duration is higher than the second preset duration threshold, a health check result is obtained that the health check status of the backend server is sub-healthy; if the connection duration is lower than the second preset duration threshold, a health check result is obtained that the health check status of the backend server is healthy. That is, the timeout duration of the target health check mode is determined according to the configuration instruction, and the check period is used as the health check period. The preset transmission control protocol port detection is initiated to the back-end server to determine whether the connection with the back-end server is successfully established within the timeout period. If the connection fails, it means that there may be a problem with the back-end server, and an abnormal health check result is obtained. If the connection is successful, it means that the back-end server is normal, and it is necessary to determine the connection duration for the successful connection. When the connection duration is higher than the second preset duration threshold, a sub-healthy health check result is obtained. When the connection duration is lower than the second preset duration threshold, a healthy health check result is obtained. Taking the check period of 30 seconds, the timeout duration of 2 seconds, and the second preset duration threshold of 500 milliseconds as an example, a TCP port detection is performed on all back-end servers every 30 seconds. If the TCP connection fails to be established within 2 seconds, the back-end server is considered abnormal and an abnormal health check result is obtained; if the TCP connection is successfully established within 2 seconds, the specific state of the back-end server is determined according to the connection duration. For example, when performing TCP port detection on a backend server, if a TCP connection is successfully established within 500 milliseconds, a healthy health check result is obtained. If a TCP connection is successfully established within 2 seconds beyond 500 milliseconds, a sub-healthy health check result is obtained.

[0069] In a third specific implementation, the target health check method can be a test method using UDP (User Datagram Protocol) data transmission, that is, sending a data packet to the back-end server through the UDP protocol, and the back-end server responds to the data packet sent by the health check end, and the health check end checks the transmission quality of the data packet. Specifically, the health check end determines the response duration, number of packets and packet loss rate threshold of the target health check method based on the configuration instruction; takes the inspection period as the health check period, and sends data packets to the back-end server in sequence based on the number of packets through the user datagram protocol to determine whether the back-end server receives a response to the data packet within the response period; counts the number of data packets in the data packet that receive the response within the response period to determine the packet loss rate of this health check period; if the packet loss rate is higher than the packet loss rate threshold, a health check result is obtained that the health check status of the back-end server is abnormal; if the packet loss rate is lower than the packet loss rate threshold, and the packet loss rate is higher than a preset numerical threshold, a health check result is obtained that the health check status of the back-end server is sub-healthy; the preset numerical threshold is lower than the packet loss rate threshold; if the packet loss rate is lower than the preset numerical threshold, a health check result is obtained that the health check status of the back-end server is healthy. That is, the response duration, number of packets and packet loss rate threshold of the target health check method are determined according to the configuration instructions, the inspection period is used as the health check period, and data packets are sent to the back-end server in sequence based on the number of packets through the User Datagram Protocol to determine whether a response to the data packet from the back-end server is received within the response duration.

[0070] It is understandable that after the health check end sends the data packet, the number of data packets in the data packet that receive the response within the response time is counted to determine the packet loss rate of this health check cycle, and the packet loss rate is compared with the packet loss rate threshold. If it is higher than the packet loss rate threshold, an abnormal health check result is obtained. If it is lower than the packet loss rate threshold and the packet loss rate is higher than the preset numerical threshold, it means that the server is normal but there may be a high load situation resulting in a long response time, and a sub-healthy health check result is obtained. If it is lower than the preset numerical threshold, a healthy health check result is obtained. Taking the inspection cycle of 30 seconds, the response time of 2 seconds, the number of packets sent is 10, the packet loss rate threshold is 20%, and the preset numerical threshold is 10% as an example, the health check end sends a data packet to all backend servers through the UDP protocol every 30 seconds, and receives the data packet replied by the backend server, and sends 10 data packets continuously. If the number of received data packets is greater than or equal to the difference between the number of packets sent and the packet loss rate threshold, that is, 8 and more than 8, the backend server is considered normal, otherwise it is considered unhealthy and an abnormal health check result is obtained. Among them, the health check result is normal and is also divided into two situations. If the packet loss rate is between 10% and 20%, a sub-healthy health check result is obtained. If the packet loss rate is less than 10%, a healthy health check result is obtained.

[0071] In a fourth specific implementation, the target health check method can be a test method using an HTTP (Hypertext Transfer Protocol) or HTTPS (Hypertext Transfer Protocol Secure) interface, that is, sending an HTTP or HTTPS request to a backend server and checking the returned status code and response content. Specifically, the health check end determines the response duration, request protocol, response information and request initiation information of the target health check method based on the configuration instruction; the request protocol includes Hypertext Transfer Protocol and Hypertext Transfer Security Protocol; the check period is the health check period, and the request protocol is sent to the back-end server based on the request initiation information to determine whether the target response information consistent with the response information is received within the response duration; if the target response information is not received within the response duration, a health check result is obtained that the health check status of the back-end server is abnormal; if the target response information is received within the response duration, the target response duration of receiving the target response information is counted, and the target response duration is compared with a third preset duration threshold; if the target response duration is higher than the third preset duration threshold, a health check result is obtained that the health check status of the back-end server is sub-healthy; if the target response duration is lower than the third preset duration threshold, a health check result is obtained that the health check status of the back-end server is healthy. Wherein, the request initiation information includes but is not limited to domain name, check path, request method, response status code, etc. That is, the response duration, request protocol, response information and request initiation information of the target health check method are determined according to the configuration instruction, the inspection period is used as the health check period, and the request protocol is sent to the back-end server based on the request initiation information to determine whether the target response information consistent with the response information is received within the response duration. If the target response information is not received, an abnormal health check result is obtained. If the target response information is received, the status of the back-end server is further determined, the target response time of receiving the target response information is counted, and the target response time is compared with the third preset time threshold. If it is higher than the third preset time threshold, a sub-healthy health check result is obtained. If it is lower than the third preset time threshold, a healthy health check result is obtained.Taking the inspection cycle of 30 seconds, response time of 2 seconds, domain name www.test.com, inspection path / health-check, request protocol https, request method GET, response status code 2xx, response body success as an example, the health check end sends an https request to all back-end servers every 30 seconds, which is equivalent to making an http interface call. The request URL is https: / / www.test.com / health-check, and the request method is GET. If a response code of 2xx is returned within 2 seconds, and the response body contains success, it can be preliminarily determined that the back-end server is in a normal state. If the interface response time exceeds 2s, or the interface response code is 4xx, 5xx, etc., an abnormal health check result is obtained. At the same time, after determining that the back-end server is normal, the interface response time is recorded. If the response time is between 500 milliseconds and 2 seconds, a sub-healthy health check result is obtained. If the response time is less than 500 milliseconds, a healthy health check result is obtained.

[0072] It should be further pointed out that, in addition to adopting a single method for determining the health check result, this embodiment can also use a combination of multiple methods, that is, select two or more methods from the several custom health check methods as target health check methods, and use them in combination to improve the accuracy and reliability of the health check. For example, when the ICMP detection test is normal, the HTTP interface test can be performed to ensure the integrity and availability of the server. In this way, this embodiment provides a custom health check method. In addition to providing some preset methods, it can also be customized by users, making the health check method more diversified and more flexible to meet the personalized needs of users; at the same time, in addition to health and abnormality, the health check status adds an intermediate state, that is, a sub-healthy state, so as to dynamically adjust the weight for the back-end server in the sub-healthy state, reduce traffic distribution, and thus promote the server to return to normal; in addition, the centralized monitoring program monitors the health status of the back-end server, and each health check only performs a health status detection on the back-end server, which can reduce the number of health check requests received by the back-end server, thereby alleviating the pressure on the back-end server.

[0073] Step S13: The control end adjusts the weight of the back-end server based on the health check result and a preset adjustment strategy to perform load balancing.

[0074] In this embodiment, the load balancing system adjusts the weight of the back-end server based on the health check results and the preset adjustment strategy through the control end to perform load balancing processing, thereby improving the system's adaptability when facing changes in the status of the back-end server, and ensuring efficient processing of user requests and high-quality service experience.

[0075] Specifically, after the control end receives the health check result sent by the health check end, if the health check result indicates that the health check status of the back-end server is healthy, the weight of the back-end server is increased by a preset amount until the weight reaches the preset maximum weight value; if the health check result indicates that the health check status of the back-end server is sub-healthy, or the health check result indicates that the health check status of the back-end server is abnormal, the weight of the back-end server is reduced by a preset amount until the weight reaches the preset minimum weight value. That is, the control end makes corresponding flexible adjustments to the weight of the backend server according to the health check result of the backend server. When the health check status of the backend server is healthy, it means that the performance of the backend server is good, and its weight can be increased to make it bear more requests. Therefore, the weight of the backend server can be increased by a preset amount until the weight reaches the preset maximum weight value; and when the health check status of the backend server is sub-healthy or abnormal, it means that the state of the backend server is not good or there is a fault, and its weight needs to be reduced to reduce the traffic distributed to the server in this state, reduce its load, and then prompt the server to return to normal, ensure the stability of the system and service quality, therefore, the weight of the backend server can be reduced by a preset amount until the weight reaches the preset minimum weight value. For example, assuming that the load balancing end adds five backend servers, and the initial weight of each backend server is 10, if its health check result is healthy, the weight increases by 2 each time until the maximum value of 20 is reached; if the health check result is sub-healthy, the weight decreases by 2 each time until the minimum value of 2 is reached.

[0076] It should be pointed out that the present embodiment can also add a parameter of the maximum number of attempts to limit the maximum number of times the status of the back-end server is abnormal. Specifically, if the health check results of a preset number of inspection cycles continuously indicate that the health check status of the back-end server is abnormal, the back-end server is deleted from the forwarding pool. That is, when the health check status of the back-end server for a preset number of inspection cycles continuously is abnormal, it means that the back-end server may have a failure that cannot be recovered by itself, and it is necessary to directly delete the back-end server from the forwarding pool to achieve stable operation and performance optimization of the system, and avoid the situation where the user's service experience is poor due to the failure of the back-end server to work properly. For example, if the health check results are abnormal within three consecutive maximum number of attempts, the back-end server is directly deleted from the forwarding pool. In this way, this embodiment dynamically adjusts the configuration of the load balancing module according to the results, more reasonably allocates access traffic and number of connections, improves the flexibility and scalability of the load balancing system, so as to achieve stable operation and performance optimization of the system, and significantly improves the system's adaptability in the face of changes in the status of the back-end server, ensuring efficient processing of user requests and high-quality service experience; at the same time, when the back-end server is detected to be in a sub-healthy state, the weight is dynamically adjusted to reduce the traffic distributed to the server to reduce its load, thereby prompting the server to return to normal, effectively maintaining the overall performance and reliability of the load balancing system.

[0077] As can be seen from the above, in the embodiment of the present application, the control end first sends corresponding configuration instructions to the health check end and the load balancing end respectively based on the received configuration information, so that the load balancing end performs configuration at the load balancing node according to the configuration instructions, creates a back-end server, and after the health check end receives the configuration instruction, determines a target health check method and a corresponding inspection cycle from a number of custom health check methods according to the configuration instruction, performs a health check on the back-end server based on the target health check method and the inspection cycle, and sends the obtained health check result to the control end. After the control end receives the health check result, the weight of the back-end server is adjusted based on the health check result and a preset adjustment strategy to perform load balancing processing. In this way, through the above process of the embodiment of the present application, on the one hand, the health check function is separated from the load balancer, so that the load balancer focuses on reverse proxy and traffic distribution. The entire system does not depend on the selection of the load balancer, has strong portability, reduces the working pressure of the load balancing node, and simplifies the process of replacing the load balancer; on the one hand, a variety of custom health check methods are provided. In addition to providing some preset methods, user-defined methods can also be provided, making the health check method more diversified, improving the flexibility of the load balancing system, and better meeting user needs; on the one hand, in addition to healthy and abnormal health check states, an intermediate state, namely a sub-healthy state, is added to dynamically adjust the weight of the back-end server in the sub-healthy state, reduce traffic distribution, and thereby prompt the server to return to normal; on the one hand, the health status of the back-end server is centrally monitored, and each A health check only performs one health status detection on the back-end server, which can reduce the number of health check requests received by the back-end server and relieve its work pressure; on the one hand, the configuration of the load balancing module is dynamically adjusted according to the results, and the access traffic and number of connections are more reasonably allocated, which improves the flexibility and scalability of the load balancing system to achieve stable operation and performance optimization of the system, and significantly improves the system's adaptability in the face of changes in the status of the back-end server, ensuring efficient processing of user requests and high-quality service experience; on the other hand, when the back-end server is detected to be in a sub-healthy state, the weight is dynamically adjusted to reduce the traffic distributed to the server to reduce its load, thereby prompting the server to return to normal, effectively maintaining the overall performance and reliability of the load balancing system, thereby alleviating the work pressure of the load balancing system and improving the system's flexibility and operation and maintenance efficiency.

[0078] Accordingly, see Figure 3 As shown, the embodiment of the present application also provides a load balancing device based on health check, which is applied to a load balancing system, wherein the load balancing system includes a control end, a health check end and a load balancing end, and the health check end is independent of the load balancing end; wherein the device includes:

[0079] An instruction issuing module 11 is used to issue corresponding configuration instructions to the health check end and the load balancing end respectively through the control end based on the received configuration information, so that the load balancing end performs configuration at the load balancing node according to the configuration instructions and creates a backend server;

[0080] A health check module 12 is used for, when the health check terminal receives the configuration instruction, determining a target health check mode and a corresponding check period from a plurality of custom health check modes according to the configuration instruction, and performing a health check on the backend server based on the target health check mode and the check period, so as to send the obtained health check result to the control terminal;

[0081] The weight adjustment module 13 is used to adjust the weight of the backend server based on the health check result and the preset adjustment strategy through the control end to perform load balancing processing.

[0082] As can be seen from the above, in the embodiment of the present application, the control end first sends corresponding configuration instructions to the health check end and the load balancing end respectively based on the received configuration information, so that the load balancing end performs configuration at the load balancing node according to the configuration instructions, creates a back-end server, and after the health check end receives the configuration instruction, determines a target health check method and a corresponding inspection cycle from a number of custom health check methods according to the configuration instruction, performs a health check on the back-end server based on the target health check method and the inspection cycle, and sends the obtained health check result to the control end. After the control end receives the health check result, the weight of the back-end server is adjusted based on the health check result and a preset adjustment strategy to perform load balancing processing. In this way, through the above process of the embodiment of the present application, the health check function is separated from the load balancer, so that the load balancer can focus on load distribution, reduce the working pressure of the load balancing node, and simplify the process of replacing the load balancer. At the same time, the health status of the back-end server is centrally monitored. Each health check only performs a health status detection on the back-end server once, which can reduce the number of health check requests received by the back-end server and alleviate its working pressure. In addition, a variety of custom health check methods are provided, which improves the flexibility of the load balancing system and better meets the needs of users, thereby alleviating the working pressure of the load balancing system and improving the flexibility and operation and maintenance efficiency of the system.

[0083] In some specific implementations, the health check module 12 may specifically include:

[0084] A first threshold determination unit, configured to determine a response time, a number of packets sent, and a packet loss rate threshold of the target health check method based on the configuration instruction;

[0085] A request sending unit, configured to use the check period as a health check period and send echo requests of a network control message protocol to the backend server in sequence based on the number of packets sent, so as to determine whether a response to the echo request from the backend server is received within the response time;

[0086] A first quantity counting unit is used to count the number of requests in the echo request that receive the response within the response time to determine the packet loss rate of this health check cycle;

[0087] A first result determination unit, configured to obtain a health check result indicating that the health check status of the backend server is abnormal if the packet loss rate is higher than the packet loss rate threshold;

[0088] A first duration comparison unit is used to count the average response time of the echo request in this health check cycle if the packet loss rate is lower than the packet loss rate threshold, and compare the average response time with a first preset duration threshold;

[0089] A second result determination unit is used to obtain a health check result that the health check status of the backend server is sub-healthy if the average response time is higher than the first preset time threshold;

[0090] The third result determination unit is configured to obtain a health check result indicating that the health check status of the backend server is healthy if the average response time is lower than the first preset time threshold.

[0091] In some specific implementations, the health check module 12 may specifically include:

[0092] A duration determination unit, configured to determine a timeout duration of the target health check mode based on the configuration instruction;

[0093] A detection initiating unit, used to initiate detection of a preset transmission control protocol port to the backend server with the inspection period as the health inspection period, so as to determine whether a connection with the backend server is successfully established within the timeout period;

[0094] a fourth result determination unit, configured to obtain a health check result indicating that the health check status of the backend server is abnormal if a connection with the backend server fails to be established within the timeout period;

[0095] A second duration comparison unit is used to determine the connection duration for successfully establishing the connection with the backend server if the connection with the backend server is successfully established within the timeout duration, and compare the connection duration with a second preset duration threshold;

[0096] a fifth result determination unit, configured to obtain a health check result indicating that the health check status of the backend server is sub-healthy if the connection duration is longer than the second preset duration threshold;

[0097] The sixth result determination unit is used to obtain a health check result indicating that the health check status of the backend server is healthy if the connection duration is lower than the second preset duration threshold.

[0098] In some specific implementations, the health check module 12 may specifically include:

[0099] A second threshold determination unit, configured to determine a response time, a number of packets sent, and a packet loss rate threshold of the target health check method based on the configuration instruction;

[0100] a data packet sending unit, configured to use the inspection period as a health inspection period, and to sequentially send data packets to the backend server based on the number of packets sent through a user datagram protocol, so as to determine whether a response to the data packet from the backend server is received within the response time;

[0101] A second quantity counting unit is used to count the number of data packets in the data packets that receive the response within the response time to determine the packet loss rate of this health check cycle;

[0102] a seventh result determination unit, configured to obtain a health check result indicating that the health check status of the backend server is abnormal if the packet loss rate is higher than the packet loss rate threshold;

[0103] an eighth result determination unit, for obtaining a health check result that the health check status of the backend server is sub-healthy if the packet loss rate is lower than the packet loss rate threshold and the packet loss rate is higher than a preset numerical threshold; the preset numerical threshold is lower than the packet loss rate threshold;

[0104] The ninth result determination unit is used to obtain a health check result indicating that the health check status of the backend server is healthy if the packet loss rate is lower than the preset numerical threshold.

[0105] In some specific implementations, the health check module 12 may specifically include:

[0106] An information determination unit, configured to determine, based on the configuration instruction, a response duration, a request protocol, response information, and request initiation information of the request protocol of the target health check method; the request protocol includes a hypertext transfer protocol and a hypertext transfer protocol secure protocol;

[0107] A protocol sending unit, configured to send the request protocol to the backend server based on the request initiation information, taking the check period as the health check period, so as to determine whether target response information consistent with the response information is received within the response time;

[0108] a tenth result determination unit, configured to obtain a health check result indicating that the health check status of the backend server is abnormal if the target response information is not received within the response time;

[0109] A third duration comparison unit is configured to count a target response duration of receiving the target response information if the target response information is received within the response duration, and compare the target response duration with a third preset duration threshold;

[0110] an eleventh result determination unit, configured to obtain a health check result indicating that the health check status of the backend server is sub-healthy if the target response time is longer than the third preset time threshold;

[0111] A twelfth result determination unit is configured to obtain a health check result indicating that the health check status of the backend server is healthy if the target response duration is lower than the third preset duration threshold.

[0112] In some specific implementations, the weight adjustment module 13 may specifically include:

[0113] A weight increasing unit, configured to increase the weight of the backend server by a preset amount if the health check result indicates that the health check status of the backend server is healthy, until the weight reaches a preset maximum weight value;

[0114] The weight reduction unit is used to reduce the weight of the backend server by a preset amount if the health check result indicates that the health check status of the backend server is sub-healthy, or if the health check result indicates that the health check status of the backend server is abnormal, until the weight reaches a preset minimum weight value.

[0115] In some specific implementations, the health check-based load balancing device may further include:

[0116] The server deleting unit is configured to delete the backend server from the forwarding pool if the health check results of a preset number of consecutive check cycles all indicate that the health check status of the backend server is abnormal.

[0117] Furthermore, the present application also discloses an electronic device. Figure 4It is a structural diagram of an electronic device 20 shown according to an exemplary embodiment, and the content in the figure cannot be regarded as any limitation on the scope of use of this application. The electronic device 20 may specifically include: at least one processor 21, at least one memory 22, a power supply 23, a communication interface 24, an input and output interface 25 and a communication bus 26. Among them, the memory 22 is used to store a computer program, and the computer program is loaded and executed by the processor 21 to implement the relevant steps in the load balancing method based on health check disclosed in any of the aforementioned embodiments. In addition, the electronic device 20 in this embodiment can specifically be an electronic computer.

[0118] In this embodiment, the power supply 23 is used to provide working voltage for each hardware device on the electronic device 20; the communication interface 24 can create a data transmission channel between the electronic device 20 and the external device, and the communication protocol it follows is any communication protocol that can be applied to the technical solution of the present application, and is not specifically limited here; the input and output interface 25 is used to obtain external input data or output data to the outside world, and its specific interface type can be selected according to specific application needs and is not specifically limited here.

[0119] In addition, the memory 22 as a carrier for storing resources may be a read-only memory, a random access memory, a disk or an optical disk, etc. The resources stored thereon may include an operating system 221, a computer program 222, etc., and the storage method may be temporary storage or permanent storage.

[0120] The operating system 221 is used to manage and control the hardware devices on the electronic device 20 and the computer program 222, which can be Windows Server, Netware, Unix, Linux, etc. In addition to including a computer program that can be used to complete the load balancing method based on health check performed by the electronic device 20 disclosed in any of the aforementioned embodiments, the computer program 222 can further include a computer program that can be used to complete other specific tasks.

[0121] Furthermore, the present application also discloses a computer-readable storage medium for storing a computer program; wherein, when the computer program is executed by a processor, the load balancing method based on health check disclosed above is implemented. For the specific steps of the method, reference can be made to the corresponding contents disclosed in the above embodiments, and no further description will be given here.

[0122] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0123] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0124] The steps of the method or algorithm described in conjunction with the embodiments disclosed herein may be implemented directly using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0125] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0126] The technical solution provided by the present application is introduced in detail above. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea. At the same time, for general technicians in this field, according to the idea of ​​the present application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A load balancing method based on health check, characterized in that: Applied to a load balancing system, the load balancing system comprises a control end, a health check end and a load balancing end, and the health check end is independent of the load balancing end; wherein the method comprises: The control end sends corresponding configuration instructions to the health check end and the load balancing end respectively based on the received configuration information, so that the load balancing end performs configuration at the load balancing node according to the configuration instructions to create a backend server; When the health check end receives the configuration instruction, it determines a target health check mode and a corresponding check period from a plurality of custom health check modes according to the configuration instruction, and performs a health check on the backend server based on the target health check mode and the check period, so as to send the obtained health check result to the control end; The control end adjusts the weight of the back-end server based on the health check result and the preset adjustment strategy to perform load balancing processing.

2. The load balancing method based on health check according to claim 1, characterized in that: The performing a health check on the backend server based on the target health check mode and the check period includes: Determine the response time, number of packets sent, and packet loss rate threshold of the target health check method based on the configuration instruction; Taking the inspection period as the health inspection period, and sending echo requests of the network control message protocol to the backend server in sequence based on the number of packets sent, so as to determine whether a response of the backend server to the echo request is received within the response time; Counting the number of requests in the echo request that receive the response within the response time to determine the packet loss rate of this health check cycle; If the packet loss rate is higher than the packet loss rate threshold, a health check result indicating that the health check status of the backend server is abnormal is obtained; If the packet loss rate is lower than the packet loss rate threshold, the average response time of the echo request in this health check cycle is counted, and the average response time is compared with the first preset time threshold; If the average response time is higher than the first preset time threshold, a health check result indicating that the health check status of the backend server is sub-healthy is obtained; If the average response time is lower than the first preset time threshold, a health check result indicating that the health check status of the backend server is healthy is obtained.

3. The load balancing method based on health check according to claim 1, characterized in that: The performing a health check on the backend server based on the target health check mode and the check period includes: Determine a timeout duration of the target health check mode based on the configuration instruction; Taking the inspection period as the health inspection period, initiating a detection of a preset transmission control protocol port to the backend server to determine whether a connection with the backend server is successfully established within the timeout period; If the connection with the backend server fails within the timeout period, a health check result indicating that the health check status of the backend server is abnormal is obtained; If the connection with the backend server is successfully established within the timeout period, determining the connection duration for successfully establishing the connection with the backend server, and comparing the connection duration with a second preset duration threshold; If the connection duration is longer than the second preset duration threshold, a health check result indicating that the health check status of the backend server is sub-healthy is obtained; If the connection duration is lower than the second preset duration threshold, a health check result indicating that the health check status of the backend server is healthy is obtained.

4. The load balancing method based on health check according to claim 1, characterized in that: The performing a health check on the backend server based on the target health check mode and the check period includes: Determine the response time, number of packets sent, and packet loss rate threshold of the target health check method based on the configuration instruction; Taking the inspection period as the health inspection period, sending data packets to the backend server in sequence based on the number of packets sent through the user datagram protocol to determine whether a response to the data packet from the backend server is received within the response time; Counting the number of data packets in the data packets that receive the response within the response time to determine the packet loss rate of this health check cycle; If the packet loss rate is higher than the packet loss rate threshold, a health check result indicating that the health check status of the backend server is abnormal is obtained; If the packet loss rate is lower than the packet loss rate threshold, and the packet loss rate is higher than a preset numerical threshold, a health check result is obtained indicating that the health check status of the backend server is sub-healthy; and the preset numerical threshold is lower than the packet loss rate threshold; If the packet loss rate is lower than the preset numerical threshold, a health check result is obtained indicating that the health check status of the backend server is healthy.

5. The load balancing method based on health check according to claim 1, characterized in that: The performing a health check on the backend server based on the target health check mode and the check period includes: Determine the response duration, request protocol, response information and request initiation information of the target health check method based on the configuration instruction; the request protocol includes Hypertext Transfer Protocol and Hypertext Transfer Protocol Security; Taking the inspection period as the health inspection period, sending the request protocol to the backend server based on the request initiation information to determine whether target response information consistent with the response information is received within the response time; If the target response information is not received within the response time, a health check result indicating that the health check status of the backend server is abnormal is obtained; If the target response information is received within the response time, the target response time for receiving the target response information is counted, and the target response time is compared with a third preset time threshold; If the target response time is longer than the third preset time threshold, a health check result indicating that the health check status of the backend server is sub-healthy is obtained; If the target response time is lower than the third preset time threshold, a health check result indicating that the health check status of the backend server is healthy is obtained.

6. The load balancing method based on health check according to claim 1, characterized in that: The adjusting the weight of the backend server based on the health check result and a preset adjustment strategy includes: If the health check result indicates that the health check status of the backend server is healthy, the weight of the backend server is increased by a preset amount until the weight reaches a preset maximum weight value; If the health check result indicates that the health check status of the backend server is sub-healthy, or the health check result indicates that the health check status of the backend server is abnormal, the weight of the backend server is reduced by a preset amount until the weight reaches a preset minimum weight value.

7. The load balancing method based on health check according to any one of claims 1 to 6, characterized in that: The adjusting the weight of the backend server based on the health check result and the preset adjustment strategy also includes: If the health check results of a preset number of consecutive check cycles all indicate that the health check status of the backend server is abnormal, the backend server is deleted from the forwarding pool.

8. A load balancing device based on health check, characterized in that: Applied to a load balancing system, the load balancing system includes a control end, a health check end and a load balancing end, and the health check end is independent of the load balancing end; wherein the device includes: An instruction issuing module, used to issue corresponding configuration instructions to the health check end and the load balancing end respectively through the control end based on the received configuration information, so that the load balancing end performs configuration at the load balancing node according to the configuration instructions and creates a backend server; A health check module, configured to determine a target health check mode and a corresponding check period from a plurality of custom health check modes according to the configuration instruction after the health check terminal receives the configuration instruction, and perform a health check on the backend server based on the target health check mode and the check period, so as to send the obtained health check result to the control terminal; The weight adjustment module is used to adjust the weight of the back-end server based on the health check result and the preset adjustment strategy through the control end to perform load balancing processing.

9. An electronic device, characterized in that: include: Memory, used to store computer programs; A processor, configured to execute the computer program to implement the load balancing method based on health check as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: Used to store a computer program; wherein, when the computer program is executed by a processor, the load balancing method based on health check as described in any one of claims 1 to 7 is implemented.

Images