A system for realizing container sleep and wake-up and a sleep and wake-up method

Abstract

The present invention discloses a system and a sleep and wake-up method for realizing container sleep and wake-up. The system includes a container engine, a gateway module, a snapshot management module, a sleep management module and one or more applications. The application consists of one or more containers. Composition; the dormancy method includes that after the traffic enters the gateway module, the gateway module identifies the target application for traffic forwarding according to the traffic request information, and the gateway module forwards the request to the container of the target application; the dormancy management module according to the access request data obtained from the gateway module and The dormancy policy is combined to determine whether there is an application that needs to be dormant, and generates a dormancy command that is transmitted to the snapshot management module; the snapshot management module receives the dormancy command and sleeps the corresponding container according to the dormancy command, and the snapshot management module queries the container from the container management platform The host and invoke the container engine of the host to save the container as a snapshot using CRIU technology; the snapshot management module generates a relational table.

Description

technical field [0001] The present invention relates to the field of cloud native, in particular to a system and a method for realizing dormancy and wakeup of containers. Background technique [0002] In the field of cloud native, serverless (serverless) is an emerging technology architecture technology that has received wide attention. It can reduce operating and development costs and achieve on-demand expansion. Serverless is also the direction of cloud computing. One of the most common application fields is function computing, also known as Faas, the full name is Function as a service, which is a subset of Serverless and the core of the entire Serverless; [0003] Serverless means "serverless" in Chinese, which means that developers run server logic in a stateless computing container, which is completely managed by a third party. Faas has fine-grained calls, real-time scaling, and no need to care about the underlying infrastructure; [0004] In the container-based server...

AI Technical Summary

Problems solved by technology

[0012] For the serverless framework, the basic wake-up process is as follows Figure 8 As shown, it takes a certain amount of time to pull the image according to the size of the image and the speed of the network. Configure the network to prepare to start the service. Usually, there is some initialization process to start the service until it can provide the service. The above process is used in some lightweight application scenarios, such as Golang and Python programs usually only take a few seconds to complete, but they will encounter great challenges in some framework-heavy application scenarios. For example, most of the current microservice scenarios are written based on the Java language. Hot is very time-consuming (for example, Java’s Spring framework will scan all class libraries and load them one by one, and the cold start will take a long time), even up to several minutes. In this scenario, from starting the application container on demand to completing the application startup Responding to the request takes a long time, and it cannot meet the performance requirements of the application response time. How to reduce the time interval from the wake-up trigger to the end of the startup has become a difficult problem;

Images