CN116781701A - Message processing method, device, electronic equipment and storage medium - Google Patents

Abstract

The message processing method provided by the embodiment of the present disclosure is executed by a load balancing node. The method includes: receiving the first request sent by the client; determining the target container instance to allocate the first request according to the load balancing rules; according to the target container The Internet Protocol IP address of the instance and the port number of the service port, modify the target address of the first request and obtain the second request; and send the second request to the container node where the target container instance is located. Here, the target container instance to which the first request is allocated is determined according to the load balancing rules, the IP address and port number of the target container instance are directly determined during load balancing, and the first request is written according to the determined IP address and port number to obtain the first request. Second request message, directly send the second request message, and it will be automatically routed to the target container instance. There is no need for secondary load balancing and address modification in the request message, which can improve the sending speed of the request and thereby improve the response rate of the request.

Description

Message processing method, device, electronic equipment and storage medium

Technical field

The present disclosure relates to, but is not limited to, the technical field of cloud computing, and in particular, to a message processing method, device, electronic device, and storage medium.

Background technique

With the rapid development of cloud computing technology, more and more services and applications are provided through container clusters that containerize software applications. Container clusters can be orchestrated through the container cluster orchestration system to achieve continuous delivery of container clusters. Rapid deployment and intelligent operation and maintenance, etc.

In existing container orchestration systems, the way of providing external services often relies on container orchestration system nodes to forward loads, which cannot cope with large-scale generation needs. When the client sends a lot of external requests, there are problems such as long external request processing time and low access efficiency.

Contents of the invention

In view of this, embodiments of the present disclosure disclose a message processing method, device, electronic device, and storage medium.

According to a first aspect of an embodiment of the present disclosure, a message processing method is provided, which is executed by a load balancing node. The method includes: receiving a first request sent by a client; and determining, according to a load balancing rule, the number of times a message is allocated to the first request. Target container instance; modify the target address of the first request and obtain the second request according to the Internet Protocol IP address of the target container instance and the port number of the service port; send the target container node to the container node where the target container instance is located. Second request.

In one embodiment, the method further includes: receiving a first response to the second request; wherein the source address of the first response is the IP address of the target container instance, and the The target address is the IP address of the load balancing node; modify the source address of the first response to the IP address of the load balancing node, and modify the target address of the first response to the IP address of the client , obtain a second response; send the second response to the client.

In one embodiment, the method further includes: obtaining deployment information of alternative container instances in one or more container nodes in the container cluster, wherein the deployment information includes: one or more alternative container instances deployed by the container node. Select the IP address of the container instance and the port number of the service port.

In one embodiment, obtaining the deployment information of alternative container instances in one or more container nodes in the container cluster includes: receiving the deployment information broadcast by one or more container nodes in the container cluster; or, receiving load balancing The controller of the node obtains the deployment information from the server of the container cluster.

In a second aspect, embodiments of the present disclosure provide a message processing method, which is executed by a container node. The method includes: receiving a second request sent by a load balancing node; wherein the second request includes: the Internet protocol of the target container instance. The IP address and the port number of the service port; the second request is generated and provided by the load balancing node based on the first request provided by the client; the target address of the first request points to the load balancing node; according to the The IP address of the target container instance and the port number of the service port are used to determine the network namespace of the target container instance; and the second request is sent to the network namespace of the target container instance.

In one embodiment, the method further includes: receiving the target container instance to send a first response based on the second request; wherein the target address of the first response is the IP address of the load balancing node, and The source address is the IP address of the target container instance and the port number of the service port; the first response is sent to the load balancing node, where the first response is used by the load balancing node to The client provides a second response; wherein the target address of the second response is the IP address of the client and the source address is the IP address of the load balancing node.

In a third aspect, embodiments of the present disclosure provide a message processing method, which is executed by a controller of a load balancing node. The method includes: monitoring deployment information, where the deployment information at least includes: one or more containers in a container cluster. The IP address and service port number of one or more alternative container instances deployed by the node; send the deployment information to the load balancing node, where the deployment information is used for the load balancing node to modify the client The endpoint receiving the target address of the first request gets the second request routed directly to the target container instance.

In a fourth aspect, embodiments of the present disclosure provide a message processing device. The device includes: a receiving module, configured to receive the first request sent by the client; and a determining module, configured to: determine to allocate the first request according to load balancing rules. A requested target container instance; a modification module configured to: modify the target address of the first request and obtain a second request according to the Internet Protocol IP address of the target container instance, the target container instance and the port number of the service port; send Module configured to send the second request to the container node where the target container instance is located.

In a fifth aspect, embodiments of the present disclosure provide a message processing device. The device includes: a receiving module, configured to: receive a second request sent by a load balancing node; wherein the second request includes: the Internet address of the target container instance. The protocol IP address and the port number of the service port; the second request is generated and provided by the load balancing node based on the first request provided by the client; the target address of the first request points to the load balancing node; determine A module configured to: determine the network namespace of the target container instance according to the IP address of the target container instance and the port number of the service port; a sending module configured to send the second request to the target container instance network namespace.

In a sixth aspect, embodiments of the present disclosure provide a message processing device. The device includes: a monitoring module configured to: monitor deployment information, wherein the deployment information at least includes: deployment of one or more container nodes in the container cluster. The IP address and service port number of one or more alternative container instances; a sending module, used to: send the deployment information to the load balancing node, where the deployment information is used for the load balancing node Modify the target address of the first request received from the client to route the second request directly to the target container instance.

In a seventh aspect, an embodiment of the present disclosure provides an electronic device. The electronic device includes: a processor and a memory for storing a computer program capable of running on the processor; wherein when the processor runs the computer program , perform the steps of the method described in one or more of the foregoing technical solutions.

In an eighth aspect, embodiments of the present disclosure provide a computer-readable storage medium that stores computer-executable instructions; after the computer-executable instructions are executed by a processor, one or more of the foregoing technical solutions can be implemented described method.

In the message processing method provided by the embodiment of the present disclosure, the load balancing node determines the target container instance to which the first request is allocated according to the load balancing rules, directly determines the IP address and port number of the target container instance during load balancing, and determines the target container instance based on the determined IP address and port number. Write the port number into the first request and get the second request message. Send the second request message directly and it will be automatically routed to the target container instance. There is no need for secondary load balancing and address modification in the request message, which can improve the efficiency of the request. Sending speed, thereby improving the response rate of requests.

Description of drawings

Figure 1 is a schematic flowchart of a message processing method provided by an embodiment of the present disclosure.

Figure 2 is a schematic flowchart of a message processing method provided by an embodiment of the present disclosure.

Figure 3 is a schematic flowchart of a message processing method provided by an embodiment of the present disclosure.

Figure 4 is a schematic flowchart of a message processing method provided by an embodiment of the present disclosure.

Figure 5 is a schematic flowchart of a message processing method provided by an embodiment of the present disclosure.

Figure 6 is a schematic flowchart of a message processing method for a container cluster provided by an embodiment of the present disclosure.

Figure 7 is a schematic diagram of a message processing system provided by an embodiment of the present disclosure.

Figure 8 is a schematic flowchart of a request processing method provided by an embodiment of the present disclosure.

Figure 9 is a schematic flowchart of a request processing method provided by an embodiment of the present disclosure.

Figure 10 is a schematic diagram of a message processing device provided by an embodiment of the present disclosure.

Figure 11 is a schematic diagram of a message processing device provided by an embodiment of the present disclosure.

Figure 12 is a schematic diagram of a message processing device provided by an embodiment of the present disclosure.

Detailed ways

In order to make the purpose, technical solutions and advantages of the present disclosure clearer, the present disclosure will be further described in detail below in conjunction with the accompanying drawings. The described embodiments should not be regarded as limiting the present disclosure. Those of ordinary skill in the art will not make any All other embodiments obtained under the premise of creative work belong to the scope of protection of this disclosure.

In the following description, reference is made to "some embodiments" which describe a subset of all possible embodiments, but it is understood that "some embodiments" may be the same subset or a different subset of all possible embodiments, and Can be combined with each other without conflict.

In the following description, the terms "first\second\third" are only used to distinguish similar objects and do not represent a specific ordering of objects. It is understandable that "first\second\third" is used in Where permitted, the specific order or sequence may be interchanged so that the disclosed embodiments described herein can be practiced in other sequences than illustrated or described herein.

Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure belongs. The terminology used herein is for the purpose of describing embodiments of the disclosure only and is not intended to limit the disclosure.

In order to better understand the embodiments of the present disclosure, some scenario examples are described below:

In one embodiment, as shown in Figure 6, the container cluster message processing method may include: a host port (HostPort) mechanism, a node port (NodePort) mechanism, and a load balancing service (Load Balance Service) mechanism.

The host port mechanism maps the port of the container instance to the port of the container node, and receives the request sent by the client according to the container node IP address and the container node port.

The node port mechanism performs load balancing according to the request sent by the client received by the container node; performs load balancing according to the kernel module of the container node, and forwards the request to the target container instance.

The load balancing service mechanism is based on adding a load balancer outside the container cluster, performing first-layer load balancing through the equal-cost multi-routing protocol based on the load balancer, receiving requests sent by the client and forwarding the requests to the kernel forwarding module of the container node. ; Perform second-layer load balancing based on the kernel forwarding module and forward the request to the target container instance. The load balancer may include a third-party load balancer or a container cluster network plug-in, and the network plug-in may include Calico network plug-in (Calico), Metal Load Balancer (MetalLB), etc. The second layer load balancing may include a cluster Internet protocol (ClusterIP) mechanism or a node port (NodePort) mechanism.

Here, load balancing through the kernel of the container node cannot cope with large-scale requests. Two-layer load balancing is performed through the Load Balance Service mechanism. The container has problems of uneven load and large network performance loss.

As shown in Figure 1, an embodiment of the present disclosure provides a message processing method, which is executed by a load balancing node. The method includes:

Step S101: Receive the first request sent by the client;

Step S102: Determine the target container instance to which the first request is allocated according to the load balancing rules;

Step S103: Modify the target address of the first request and obtain a second request according to the Internet Protocol IP address of the target container instance and the port number of the service port;

Step S104: Send the second request to the container node where the target container instance is located.

In one embodiment, the load balancing node can be in a load balancing cluster (SLB, Server LoadBalance), and the load balancing cluster can evenly distribute requests from external clients to multiple back-end servers for execution according to load balancing rules. deal with.

Illustratively, the SLB can select the target container instance that responds to the first request according to a load balancing strategy such as a polling strategy and/or a random strategy. Specifically, any method of related technologies can be used to implement load balancing. Herein, No specific restrictions are made.

The target container instance may be in a container cluster, and the container cluster may be automatically deployed, expanded, and managed by a container orchestration system, where the container orchestration system may include a Kubernetes cluster. A Kubernetes cluster executes workloads by placing container instances on nodes. A Kubernetes cluster includes at least one node. The node can be a virtual machine or a physical machine. The node includes at least one container instance.

In one embodiment, the step S101 may include: setting one or more virtual load balancing node Internet Protocol (IP, Internet Protocol) addresses as external client access addresses according to all container clusters; receiving the client sent to the load The first request for the IP address of the balancing node, wherein the source IP address of the first request is the client IP address, and the destination address of the first request is the IP address of the load balancing node.

In one embodiment, the first request may include: a request from the client to operate or access an application or service provided according to the container cluster.

In one embodiment, the load balancing rules in step S102 may include but are not limited to at least one of the following:

Polling rules, each request is assigned to different target container instances one by one in chronological order;

Weight rules, set a weight for each container node and/or container instance, and determine the target container instance for allocation requests based on the weight. The higher the weight, the more requests allocated to the container node and/or container instance;

Random rules to randomly distribute requests to target container instances;

The minimum weighted connection number rule counts the number of active connections between container nodes and container instances in real time, and allocates new requests to target container instances with fewer active connections in proportion to the weight;

The fastest response time rule is allocated based on the response time of the container instance, and requests are allocated first to the container instance with the fastest response time, etc.

In one embodiment, step S103 may include: modifying the target IP address in the first requested target address to the IP address of the target container instance, and modifying the target port in the first requested target address to the target The port number of the service port of the container instance. Modify the first request target address to obtain the second request.

In one embodiment, step S104 may include: obtaining the best routing path from the load balancing node to the container node where the target container instance is located according to a dynamic routing protocol, and sending the best routing path to the container node according to the best routing path. The container node where the target container instance is located sends the second request.

In this way, the load balancing node sends the second request through the best routing path obtained by the dynamic routing protocol. Compared with sending according to the fixed static routing path, the second request can be sent through the best routing path, which improves the efficiency of sending the second request. Speed and accuracy.

In some embodiments, the method further includes: modifying the source address of the first request according to the IP address of the load balancing node to obtain the second request whose source address is the IP address of the load balancing node.

In one embodiment, the source address of the first request is the client IP address, and the second request is obtained by modifying the source address of the first request to the IP address of the load balancing node.

In one embodiment, a load balancing cluster may include one or more load balancing nodes, and the IP address of the load balancing node may be the IP address of the load balancing node that processes the current first request and the second request. In this way, the source address of the second request is the IP address of the load balancing node, which can facilitate the response of the second request to be returned to the same load balancing node for processing according to the source address of the second request.

When there are multiple load balancing nodes, the load balancing node that handles the first request can be determined according to the service type of the first request, and the source address of the first request is modified to the IP address of the load balancing node to obtain the second request. It is convenient for the same load balancing node to process the second request and the response to the second request. In this way, compared with the load balancing cluster randomly determining the load balancing node for the first request, requests of the same business type can be processed through the fixed load balancing node. Realize the concurrent processing of load balancing clusters in business and improve the efficiency of load balancing.

As shown in Figure 2, in some embodiments, the method further includes:

Step S201: Receive the first response of the second request; wherein the source address of the first response is the IP address of the target container instance, and the target address of the first response is the load balancing node. IP address;

Step S202: Modify the source address of the first response to the IP address of the load balancing node, and modify the target address of the first response to the IP address of the client to obtain a second response;

Step S203: Send the second response to the client.

In one embodiment, the method further includes: after modifying the source address of the first request according to the IP address of the load balancing node and obtaining the second request, recording the source address of the first request. The client’s IP address;

receiving a first response to the second request;

Modify the source address of the first response to the IP address of the load balancing node, and modify the target address of the first response to the IP address of the client to obtain a second response;

The second response is sent to the client according to the target address of the second response. In this way, the processing of the request message sent by the client is completed.

In some embodiments, the method further includes:

Obtain deployment information of alternative container instances in one or more container nodes in the container cluster, where the deployment information includes: the IP address and the port number of the service port of the one or more alternative container instances deployed by the container node.

In one embodiment, the load balancing node can obtain the deployment information of alternative container instances in one or more container nodes in the container cluster in real time; and determine the allocation of the alternative container instances according to the load balancing rules in real time. The target container instance of the request; according to the Internet protocol IP address and the port number of the service port of the target container instance in the deployment information of the alternative container instance, modify the target address of the first request and obtain the second request.

In one embodiment, an elastic scaling operation can be set within the container cluster according to the kubernetes cluster. The elastic scaling operation can include comparing the central processing unit (CPU, central processing unit) usage of the container node with a predetermined CPU usage threshold. Yes, determine whether to increase or decrease the number of container instances.

When the container cluster performs elastic scaling operations, the deployment information of alternative container instances in one or more container nodes in the container cluster can be obtained in real time through the load balancing node, and the new or reduced alternatives in the container nodes in the container cluster can be updated in a timely manner. Deployment information of container instances, and enable the load balancing node to determine the target container instance of the allocation request based on the newly added or reduced alternative container instances, which improves the cooperation between the load balancing node and the container cluster, and improves the processing of the load balancing cluster and the container cluster. Request efficiency.

In some embodiments, obtaining the deployment information of alternative container instances in one or more container nodes in the container cluster includes:

Receive the deployment information broadcast by one or more container nodes in the container cluster;

or,

Receive the deployment information obtained by the controller of the load balancing node from the server of the container cluster.

In one embodiment, receiving the deployment information broadcast by one or more container nodes in the container cluster may include: a dynamic router broadcasts the deployment information of one or more container nodes in the container cluster to A load balancing node receives the deployment information broadcast by the container node.

In one embodiment, the dynamic routing protocol can generate and maintain routing tables required for forwarding, automatically update the routing tables when the network topology changes, and be responsible for determining the best routing path for data transmission. When one or more container nodes add, modify, or delete one or more candidate container instances, the load balancing node may receive the deployment information through a dynamic routing protocol.

As shown in Figure 3, an embodiment of the present disclosure provides a message processing method, which is executed by a container node. The method includes:

Step S301: Receive a second request sent by the load balancing node; wherein the second request includes: the Internet Protocol IP address of the target container instance and the port number of the service port; the second request is the load balancing node based on the client Generated and provided by the first request provided by the client; the target address of the first request points to the load balancing node;

Step S302: Determine the network namespace of the target container instance according to the IP address of the target container instance and the port number of the service port;

Step S303: Send the second request to the network namespace of the target container instance.

In one embodiment, according to cloud computing technology, multiple cloud virtual machines can be obtained based on physical machine virtualization. The physical machines or cloud virtual machines can be container nodes, and a container cluster can be formed based on the one or more container nodes. .

For example, the container node may pre-store a mapping relationship between the IP address and port number of each candidate container and the identifier of the network namespace corresponding to the candidate container. in this way. In step S302, the mapping relationship may be queried according to the IP address and the port number of the service port, and the network namespace of the target container instance may be determined.

The identification of the network namespace may include but is not limited to the name of the network namespace.

One or more container instances can be deployed in a container node, and the container instances in a container node share the operating system kernel of the container node where it is located. Containers can package and isolate applications and running environments, and can be used to develop, deliver and deploy software or services.

Multiple container instances within a container node can be isolated from each other based on namespaces. Network namespaces can enable container instances to have independent network environments, such as network cards, routing tables, and/or firewalls.

In one embodiment, the container node may uniformly receive requests from one or more container instances within the container node, where the request may include a second request sent by the load balancing node.

In one embodiment, the container node can include an extended Berkeley Packet Filter (eBPF, extendedBerkeley Packet Filter) program. The eBPF program can process data packets in the container node and has compilability and security. The steps S302 and step S303 may be operated according to the eBPF program.

At the same time, the eBPF program processes the data packets before they enter the kernel network protocol stack. The eBPF program can work at the network card driver layer or network interface controller (NIC, network interface controller) without going through the node's kernel network protocol. The stack processes data packets, reducing the workload of the kernel, reducing the delay of container access, and improving the efficiency of container node data processing.

In one embodiment, the method may include: obtaining deployment information of an alternative container instance in a container node, where the deployment information may include: an IP address of an alternative container instance, a port number of a service port, and the backup information. The identification of the network namespace of the selected container instance; according to the IP address and service port number of the target container instance of the second request, determine the network namespace of the target container instance from the obtained deployment information of the alternative container instance; The second request is sent to the network namespace of the target container instance.

In one embodiment, the eBPF program can obtain the IP address of the alternative container instance, the port number of the service port and the alternative container of the alternative container instance in the container node from the Extended Berkeley Packet Filtering Rules (eBPF MAP) The identifier of the network namespace of the instance; wherein the identifier of the network namespace may include the name of the network namespace and the path of the network namespace; wherein the deployment information may be obtained by the Extended Berkeley Packet Filtering Agent (eBPF Agent) and The deployment information is added to the eBPF MAP.

In one embodiment, the eBPF program determines the network namespace of the target container instance based on the IP address and the port number of the service port.

The eBPF program determines whether the network card in the network namespace of the target container instance is the same as the network card of the container node host where the eBPF program is located. If the network card of the network namespace of the target container instance is the same as the network card of the container node host, the second request is forwarded directly according to the IP address and service port number of the target container instance. If the network card of the network namespace of the target container instance is different from the network card of the container node host, redirect the second request to the network card of the network namespace of the target container instance according to the network namespace path.

In one embodiment, the method further includes: the eBPF program determines whether to discard the second request based on predetermined control rules in the eBPF MAP and the IP address of the second request and the port number of the service port.

In one embodiment, the control rules may include but are not limited to at least one of the following:

When the IP address of the second request is not among the IP addresses of the alternative container instances of the current container node, discard the second request;

Match the communication protocol of the second request with the port number of the service port of the target container instance to determine whether to discard the second request.

For example, when there is a Domain Name System (DNS) attack, the control rules in the MAP can be set to discard requests whose communication protocol is User Datagram Protocol (UDP) and whose target port number is 53. .

Here, the second request can be filtered by determining whether to discard the second request according to the control rules. Compared with forwarding the request directly, the request can be discarded in time when there is a security attack or there is no target container instance of the second request in the container node, improving the container node Improve the efficiency of processing requests and improve the security of container nodes and container instances.

As shown in Figure 4, in some embodiments, the method further includes:

Step S401: Receive the target container instance and send a first response based on the second request; wherein the target address of the first response is the IP address of the load balancing node, and the source address is the IP address of the target container instance. IP address and service port number;

Step S402: Send the first response to the load balancing node, where the first response is used for the load balancing node to provide a second response to the client; where the second response The destination address is the IP address of the client and the source address is the IP address of the load balancing node.

In one embodiment, the target container instance processes the second request and generates a first response to the second request, and sends the first response; the eBPF program receives the target container instance and sends the first response based on the second request. A response is sent to the load balancing node.

In one embodiment, the method further includes: after sending the second request to the network namespace of the target container instance, recording the connection status of the second request.

The connection status may include: the source IP address of the second request, the source service port, the communication protocol, the IP address of the target container instance, the port number of the service port of the target container instance, etc.

Send the first response to the second request to the load balancing node according to the connection status of the second request. Wherein, the connection status may be recorded into the eBPF MAP according to the eBPF program, and the eBPF program sends the first response through the connection status obtained from the eBPF MAP.

In this way, the eBPF program can determine the target IP address of the first response based on the source IP address in the connection status of the second request, that is, the IP address of the load balancing node. Before the first response is sent to the container node network namespace, it can be directly determined based on the load balancing The IP address of the node forwards the first response without forwarding it through the container node network namespace, and there is no need to analyze and check the target IP address in the first response, which improves the processing efficiency of the first response.

In some embodiments, the method further includes:

When an alternative container instance is created or the address of the container node is updated, deployment information is broadcast; wherein the deployment information at least includes: the IP address of the alternative container instance and the port number of the service port.

In one embodiment, when an alternative container instance is created or the address of the container node is updated, the dynamic router broadcasts the deployment information to the load balancing node through the dynamic routing protocol and the network connection between the container node and the load balancing node. in network equipment.

Wherein, the dynamic router may include a dynamic software router, and the dynamic software router may be deployed in each container node. The routing information of the dynamic software router of the container node may include: the IP address of the alternative container instance within the container node. The next routing node is the IP address of the container node.

In this way, compared with not broadcasting deployment information, it is easier for the load balancing node and the network device connected between the container node and the load balancing node to obtain the changed deployment information in a timely manner, so that the load balancing node can timely obtain the changed deployment information according to the changed deployment information. Adjust the load balancing rules so that the network device updates the routing table in a timely manner, transmits request and response data more quickly, and improves the efficiency of processing request and response data.

In some embodiments, the method further includes:

Locally store the IP address of the alternative container instance, the port number of the service port, and the identification of the network namespace of the alternative container instance.

In one embodiment, the method may include: combining the IP address of the alternative container instance, the port number of the service port, the name of the network namespace of the alternative container instance and the network namespace according to the eBPF Agent. The path is stored into the eBPF MAP.

In one embodiment, the eBPF program can execute specific events in the kernel space, and the eBPF MAP can store the information in the eBPF MAP in the user space. In the user space, the eBPF program can also be compiled or the eBPF MAP can be modified. The filtering rules can be stored or compiled in the user space through the eBPF component, reducing the load on the kernel space and improving the message processing efficiency of the container node.

As shown in Figure 5, the embodiment of the present disclosure provides a message processing method, which is executed by the controller of the load balancing node. The method includes:

Step S501: Monitor deployment information, where the deployment information at least includes: the IP address and the port number of the service port of one or more alternative container instances deployed by one or more container nodes in the container cluster;

Step S502: Send the deployment information to the load balancing node, where the deployment information is used for the load balancing node to modify the target address of the first request received from the client to obtain a second direct route to the target container instance. ask.

In one embodiment, the controller of the load balancing node may include a load balancing node service controller (SLB-SVC, Server Load Balance-Service).

In one embodiment, the method may include: monitoring deployment information of one or more alternative container instances in one or more container nodes in the container cluster, and sending the deployment information to the load balancing node. In this way, the load balancing cluster can obtain the latest deployment information in a timely manner, respond to changes in the container cluster, and adjust load balancing based on the latest deployment information, thereby improving the efficiency of load balancing.

In one embodiment, the method may further include: monitoring the deployment information of one or more container instances in one or more container nodes that provide the same service function in the container cluster, where the container instances that provide the same service function may include It is composed of different container instances in different container nodes in the container cluster.

In one embodiment, the load balancing node may send the deployment information of one or more container instances in one or more container nodes that provide the same service function to the load balancing cluster. The load balancing cluster may allocate an address for external service access to one or more container instances that provide the same service function. The address for external access may include the IP address of the load balancing node and the service port number of the load balancing node.

In one embodiment, the controller of the load balancing node may also clear invalid deployment information in the configuration information of the load balancing cluster within a predetermined time range.

In this way, the controller based on the load balancing node listens to the deployment information in the container cluster and clears the invalid deployment information in the configuration information in the load balancing cluster. Compared with the load balancing cluster without the controller of the load balancing node, it can be configured in the container cluster. When there are frequent changes, the load balancing cluster can obtain the deployment information of the container cluster in time, and clear the invalid deployment information in the load balancing cluster in time. This can realize the two-way synchronization of the container deployment information and the load balancing configuration information, and improve the processing efficiency of the load balancing message. accuracy.

In some embodiments, the monitoring deployment information includes:

Receive the deployment information from the server of the container cluster;

And/or, monitor the deployment information broadcast by container nodes in the container cluster.

In one embodiment, the controller of the load balancing node can monitor the server of the container node in the container cluster, and when the container node is invalid, delete the deployment information of the corresponding container instance in the container node.

In one embodiment, a message processing system can be shown in Figure 7 . Wherein, the message processing system includes: load balancing SLB cluster, load balancing node controller SLB-SVC, kubernetes cluster, and container cluster. The kubernetes cluster may include a Kube-APIServer component, the container cluster may include one or more container nodes, the container cluster may include a dynamic software router, and the dynamic software router may be deployed where the container node is. In the virtual machine, the container node may also include: eBPF program, eBPF Agent, eBPF MAP, container instance and node kernel protocol stack.

In one embodiment, a method for message processing according to the message processing system may include:

Receive the first request sent by the client according to the SLB cluster;

The SLB cluster determines the target container instance to which the first request is allocated based on the load balancing rules;

The SLB cluster modifies the target address of the first request and obtains the second request based on the IP address of the target container instance and the port number of the service port;

The SLB cluster sends the second request to the container node where the target container instance is located according to the dynamic routing protocol;

The container node receives the second request sent by the load balancing node;

According to the eBPF program in the container node, the IP address of the alternative container instance in the container node, the port number of the service port and the network namespace identifier stored in the eBPF MAP are combined with the IP address and the IP address of the target container instance included in the second request. The port number of the service port is used to determine the network namespace of the target container instance;

The second request is sent to the network namespace of the target container instance according to the eBPF program in the container node.

In one embodiment, the load balancing node controller SLB-SVC and the eBPF Agent can obtain the deployment information of alternative container instances in the container nodes in the container cluster through the Kube-APIServer component in the kubernetes cluster.

In one embodiment, a request processing flow can be shown in Figure 8, and the steps can include:

Step a1: The load balancing node receives the first request sent by the client;

Step a2: The load balancing node determines the target container instance to which the first request is allocated according to the load balancing rules;

Step a3: The load balancing node modifies the target address of the first request and obtains the second request according to the Internet Protocol IP address of the target container instance and the port number of the service port;

Step a4: The load balancing node sends the second request to the container node where the target container instance is located;

Step a5: The container node receives the second request sent by the load balancing node; wherein the second request includes: the Internet Protocol IP address of the target container instance and the port number of the service port; the second request is the load balancing node Generated and provided based on the first request provided by the client; the target address of the first request points to the load balancing node;

Step a6: The container node determines the network namespace of the target container instance based on the IP address of the target container instance and the port number of the service port;

Step a7: The container node sends the second request to the network namespace of the target container instance;

Step a8: The target container instance receives and processes the second request.

In one embodiment, the flow of processing a request response may be as shown in Figure 9, and the steps may include:

Step b1: The target container instance generates a first response according to the second request and sends it;

Step b2: The container node receives the target container instance and sends a first response based on the second request; wherein the target address of the first response is the IP address of the load balancing node, and the source address is the target container. The IP address of the instance and the port number of the service port;

Step b3: The container node sends the first response to the load balancing node, where the first response is used for the load balancing node to provide a second response to the client; where the first response The target address of the second response is the IP address of the client and the source address is the IP address of the load balancing node;

Step b4: The load balancing node receives the first response of the second request; wherein the source address of the first response is the IP address of the target container instance, and the target address of the first response is the load The IP address of the balancing node;

Step b5: The load balancing node modifies the source address of the first response to the IP address of the load balancing node, and modifies the target address of the first response to the IP address of the client to obtain a second response;

Step b6: The load balancing node sends the second response to the client.

As shown in Figure 10, an embodiment of the present disclosure provides a message processing device, which includes:

The receiving module 10 is used to receive the first request sent by the client;

Determining module 20, configured to: determine the target container instance to which the first request is allocated according to the load balancing rules;

Modification module 30, configured to: modify the target address of the first request and obtain the second request according to the Internet Protocol IP address of the target container instance and the port number of the service port;

The sending module 40 is configured to send the second request to the container node where the target container instance is located.

In one embodiment, the modification module 30 is further configured to modify the source address of the first request according to the IP address of the load balancing node to obtain the source address of the load balancing node. Second request.

In one implementation, the receiving module 10 is further configured to: receive a first response to the second request; wherein the source address of the first response is the IP address of the target container instance, and the third The target address of a response is the IP address of the load balancing node; the modification module 30 is also configured to: modify the source address of the first response to the IP address of the load balancing node, and modify the first response to the IP address of the load balancing node. The target address of a response is modified to the IP address of the client to obtain a second response; the sending module 40 is also used to: send the second response to the client.

In one embodiment, the device further includes: an acquisition module 50; the acquisition module 50 is configured to: acquire deployment information of alternative container instances in one or more container nodes in the container cluster, where the deployment information Including: the IP address and the port number of the service port of one or more alternative container instances deployed by the container node.

In one embodiment, the receiving module 10 is also configured to receive the deployment information broadcast by one or more container nodes in the container cluster; or, receive the deployment information obtained by the controller of the load balancing node from the server of the container cluster. The deployment information.

As shown in Figure 11, an embodiment of the present disclosure provides a message processing device, which includes:

The receiving module 110 is configured to: receive a second request sent by the load balancing node; wherein the second request includes: the Internet Protocol IP address of the target container instance and the port number of the service port; the second request is the load The balancing node is generated and provided based on the first request provided by the client; the target address of the first request points to the load balancing node;

Determining module 120, configured to: determine the network namespace of the target container instance according to the IP address of the target container instance and the port number of the service port;

The sending module 130 is configured to send the second request to the network namespace of the target container instance.

In one embodiment, the receiving module 110 is further configured to: receive a first response sent by the target container instance based on the second request; wherein the target address of the first response is that of the load balancing node. IP address, and the source address is the IP address of the target container instance and the port number of the service port; the sending module 130 is also used to: send the first response to the load balancing node, wherein, The first response is used by the load balancing node to provide a second response to the client; wherein the target address of the second response is the IP address of the client and the source address is the IP of the load balancing node. address.

As shown in Figure 12, an embodiment of the present disclosure provides a message processing device, which includes:

The monitoring module 310 is configured to: monitor deployment information, where the deployment information at least includes: the IP address and the port number of the service port of one or more alternative container instances deployed by one or more container nodes in the container cluster;

The sending module 320 is configured to send the deployment information to the load balancing node, where the deployment information is used for the load balancing node to modify the target address of the first request received from the client to obtain a direct route to the target container. Second request for instance.

In one embodiment, the monitoring module 310 further includes: a receiving module 311, configured to receive the deployment information from the server of the container cluster; and/or the monitoring module 310 is also configured to monitor the content of the container cluster. The deployment information broadcast by the server node.

It should be noted that those skilled in the art can understand that the methods provided in the embodiments of the present disclosure can be executed alone or together with some methods in the embodiments of the present disclosure or some methods in related technologies.

An embodiment of the present disclosure also provides an electronic device. The electronic device includes: a processor and a memory for storing a computer program that can be run on the processor. When the processor runs the computer program, it executes one or more of the foregoing The steps of the method described in the technical solution.

Embodiments of the present disclosure also provide a computer-readable storage medium that stores computer-executable instructions. After the computer-executable instructions are executed by the processor, the method described in one or more of the foregoing technical solutions can be implemented. .

The computer storage medium provided in this embodiment may be a non-transient storage medium. In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are only illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods, such as: multiple units or components may be combined, or can be integrated into another system, or some features can be ignored, or not implemented. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed may be through some interfaces, and the indirect coupling or communication connection of the devices or units may be electrical, mechanical, or other forms. of.

The units described above as separate components may or may not be physically separated. The components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units; Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

In addition, each functional unit in each embodiment of the present disclosure can be all integrated into one processing module, or each unit can be separately used as a unit, or two or more units can be integrated into one unit; the above-mentioned integration The unit can be implemented in the form of hardware or in the form of hardware plus software functional units.

In some cases, if any two of the above technical features do not conflict, they can be combined into a new method and technical solution.

In some cases, if any two of the above technical features do not conflict, they can be combined into a new equipment technical solution.

Those of ordinary skill in the art can understand that all or part of the steps to implement the above method embodiments can be completed by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, It includes the steps of the above method embodiment; and the aforementioned storage media includes: mobile storage devices, read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), magnetic disks or optical disks, etc. A medium on which program code can be stored.

The above are only specific embodiments of the present disclosure, but the protection scope of the present disclosure is not limited thereto. Any person familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present disclosure. should be covered by the protection scope of this disclosure. Therefore, the protection scope of the present disclosure should be subject to the protection scope of the claims.

Claims (12)

1. A message processing method, characterized in that it is executed by a load balancing node, and the method includes: Receive the first request sent by the client; Determine the target container instance to which the first request is allocated according to the load balancing rules; Modify the target address of the first request and obtain the second request according to the Internet Protocol IP address of the target container instance and the port number of the service port; Send the second request to the container node where the target container instance is located. 2. The method according to claim 1, characterized in that, the method further comprises: Receive the first response of the second request; wherein the source address of the first response is the IP address of the target container instance, and the target address of the first response is the IP address of the load balancing node; Modify the source address of the first response to the IP address of the load balancing node, and modify the target address of the first response to the IP address of the client to obtain a second response; Send the second response to the client. 3. The method according to any one of claims 1 to 2, characterized in that the method further includes: Obtain the deployment information of alternative container instances in one or more container nodes in the container cluster, where the deployment information includes: the IP address and the port number of the service port of one or more alternative container instances deployed by the container node. . 4. The method according to claim 3, wherein obtaining the deployment information of alternative container instances in one or more container nodes in the container cluster includes: Receive the deployment information broadcast by one or more container nodes in the container cluster; or, Receive the deployment information obtained by the controller of the load balancing node from the server of the container cluster. 5. A message processing method, characterized in that it is executed by a container node, and the method includes: Receive a second request sent by the load balancing node; wherein the second request includes: the Internet Protocol IP address of the target container instance and the port number of the service port; the second request is provided by the load balancing node based on the client The first request is generated and provided; the target address of the first request points to the load balancing node; Determine the network namespace of the target container instance according to the IP address of the target container instance and the port number of the service port; Send the second request to the network namespace of the target container instance. 6. The method according to claim 5, characterized in that, the method further comprises: Receive the target container instance and send a first response based on the second request; wherein the target address of the first response is the IP address of the load balancing node, and the source address is the IP address of the target container instance and The port number of the service port; Send the first response to the load balancing node, where the first response is used for the load balancing node to provide a second response to the client; where the target address of the second response is the IP address of the client and the source address is the IP address of the load balancing node. 7. A message processing method, characterized in that it is executed by a controller of a load balancing node, and the method includes: Monitor deployment information, wherein the deployment information at least includes: the IP address and the port number of the service port of one or more alternative container instances deployed by one or more container nodes in the container cluster; The deployment information is sent to the load balancing node, where the deployment information is used for the load balancing node to modify the target address of the first request received from the client to obtain a second request that is directly routed to the target container instance. 8. A message processing device, characterized in that the device includes: The receiving module is used to receive the first request sent by the client; Determining module, configured to: determine the target container instance to which the first request is allocated according to the load balancing rules; Modifying module, configured to: modify the target address of the first request and obtain the second request according to the Internet Protocol IP address of the target container instance, the port number of the target container instance and the service port; A sending module, configured to send the second request to the container node where the target container instance is located. 9. A message processing device, characterized in that the device includes: A receiving module, configured to: receive a second request sent by the load balancing node; wherein the second request includes: the Internet Protocol IP address of the target container instance and the port number of the service port; the second request is the load balancing node The node is generated and provided based on the first request provided by the client; the target address of the first request points to the load balancing node; Determining module, configured to: determine the network namespace of the target container instance according to the IP address of the target container instance and the port number of the service port; A sending module, configured to send the second request to the network namespace of the target container instance. 10. A message processing device, characterized in that the device includes: A monitoring module, configured to: monitor deployment information, wherein the deployment information at least includes: the IP address and the port number of the service port of one or more alternative container instances deployed by one or more container nodes in the container cluster; A sending module, configured to: send the deployment information to a load balancing node, where the deployment information is used for the load balancing node to modify the target address of the first request received from the client to obtain a direct route to the target container instance. second request. 11. An electronic device, characterized in that the electronic device includes: a processor and a memory for storing a computer program capable of running on the processor, wherein when the processor runs the computer program, claim 1 is executed Go to the steps of the message processing method described in any one of 7. 12. A computer-readable storage medium, characterized in that the computer-readable storage medium stores computer-executable instructions; after the computer-executable instructions are executed by a processor, the computer-readable storage medium can implement any one of claims 1 to 7 The message processing method described in the item.