Container cloud service discovery and load balancing method based on OpenResty and K8S

Abstract

The invention relates to a container cloud service discovery and load balancing method based on OpenResty and K8S, and the method comprises the steps: a user issuing an application on a container cloud platform, configuring the internal routing and load balancing of the application, carrying out the cluster information configuration through the self-defined resources of K8S in the process of creating load balancing, and enabling the internal service of a cluster to be exposed to the exterior of the cluster; after a user configures load balancing, a load balancer monitoring resource changes, periodically acquiring custom resource information of K8S, refreshing an Nginx configuration file according to the custom resource information and reloads the Nginx, so that the load balancing configured by the user takes effect; after a user request establishes a connection with OpenResty, performing advanced rule filtering on the request through an improved Lua module, calculating a service instance meeting a load balancing strategy, performing data storage, and then forwarding the user request to the service instance in a K8S cluster. Compared with the prior art, the method has the advantages of reducing the operation complexity, improving the service discovery speed and the like.

Description

technical field [0001] The invention relates to the technical field of container cloud platforms, in particular to a container cloud service discovery and load balancing method based on OpenResty and K8S. Background technique [0002] With the gradual implementation and vigorous development of concepts such as cloud computing, micro-services, and containers, there are more and more machines, more and more applications, more and more micro-services, and the basic environment for application operation is becoming more and more diverse. In the application of cloud business process, after the user publishes the application, K8S creates Pod resources with the same number of copies according to the application configuration, and assigns the access address PodIP to the service instance Pod. In the K8S network model, each Pod has an independent IP, and the service instance Pod has a life cycle, and a new access address PodIP will be generated when it is destroyed and rebuilt. Since...

AI Technical Summary

Problems solved by technology

[0008] 2. In the existing technical solution, the user creates an Ingress resource by writing the yaml file of the Ingress resource. The configuration of this solution is complex and the learning cost is high, requiring the user to write the yaml file proficiently to determine the externally accessible URL, load balancing strategy and virtual host and other configuration parameters

When an application instance is updated, scaled, or deleted, K8S deletes the old version of the Pod instance and creates a new Pod instance. The monitoring collection of the current technical solution does not filter the old version of the Pod, and pushes the monitoring data to Promethues, including the offline Pod information. , resulting in invalid data in the monitoring data

Images