Container-based service multi-version coexistence implementation method

Abstract

The invention provides a container-based service multi-version coexistence implementation method. The method comprises the following steps: configuring a plurality of service nodes in a server, configuring a plurality of containers in each service node and loading different service sets in each container, thereby not only realizing the multi-version coexistence of the service, but also clearly distinguishing different service types and different versions of the same service type in a service coexistence environment; services of old and new versions are distinguished by the containers, after new services are added or old services are stopped, a client can automatically screen a to-be-called service version according to its own version and obtains a service call result from the corresponding container, thereby reducing the manual controlled call when the client is inconsistent with the called service version; and due to the coexistence of multiple versions of services, the new services do not need to be restarted and loaded in an upgrading process of a service application, thereby simplifying the cumbersome process of service upgrade and realizing hot deployment of services in a distributed environment.

Description

technical field [0001] The invention belongs to the field of distributed services, and in particular relates to a container-based method for realizing multi-version coexistence of services. Background technique [0002] Internet companies, due to factors such as rapid business development and rapid demand changes, require software applications that can quickly realize business and respond to changes in customer needs, so distributed services have been widely used. The use of distributed services has brought the following three major benefits: (1) Independent high-load modules: separate high-load modules into a service and deploy them on independent machines, so that high-load modules can exclusively share hardware resources and avoid high-load modules. The load module affects the normal use of other modules. For example: a product module may be a module with a large number of visits, which can be published as a service, and other modules call its services to access its func...

AI Technical Summary

Problems solved by technology

[0007] Compared with centralized service deployment, distributed service distribution is more complex and difficult

One is because in the distributed service deployment, the centralized application is divided into multiple applications, and each application is deployed on multiple machines, and the release volume increases exponentially, resulting in a double increase in the number of machines that need to be operated when the service is upgraded or upgraded. ; The second is that the release of traditional services needs to restart the service, resulting in a long release time, which may take several minutes to several hours, which further increases the difficulty of distributed service releases; the third is due to the lack of clear marking of services, It is impossible to distinguish between old and new services. When the new service is incompatible with the old service, it is necessary to prevent the old client from accessing the new service

Images