Mirror image construction and loading method based on mirror image loading in container environment

Abstract

The invention provides a mirror image construction and loading method based on mirror image loading in a container environment, and the method comprises the following steps: 1, carrying out the decoupling of a basic operation environment part and a service business logic part during mirror image construction; step 2, after decoupling, constructing a first operating system and basic software into a basic mirror image; 3, building the second operating system and the service program part into a service mirror image, and the storage space occupied by the second operating system being smaller than that occupied by the first operating system. The service mirror image and the basic mirror image are separated, the mirror image size of the service logic is relatively small, and storage and distribution of the mirror images are facilitated.

Description

technical field [0001] The invention relates to the field of computer systems, in particular to a mirror image construction and loading method based on mirror image loading. Background technique [0002] Docker image is the storage and distribution carrier for applications, which relies on many underlying technologies, including joint file systems, copy-on-write, and so on. [0003] Image layering: An image is a unified view of a bunch of read-only layers that overlap. Except for the bottom layer, all other layers will have a pointer pointing to the next layer. These layers are implementation details inside Docker and are accessible on the host's filesystem. The unified file system (union file system) technology can integrate different layers into a file system, providing a unified perspective for these layers, thus hiding the existence of multiple layers, from the user's point of view, it is a file system . [0004] Image scheduling management tools: There are many comm...

AI Technical Summary

Problems solved by technology

Larger images will cause problems such as longer time for image construction; more space for image storage; more time for image transmission or distribution, etc.

In particular: the real and effective program size in the image may only be tens of megabytes or even a few megabytes in size, making the proportion of effective content in the image too small

[0010] 2. The image is too large, which poses a huge challenge to the environment during the distribution and storage of the image

Difficulties in the distribution process. For example, in scenarios where real-time requirements are required, the time consumption of image distribution plays a decisive role in the scenario.

Excessively large images will inevitably increase the time cost of image distribution, which in turn puts forward higher requirements for network bandwidth

In an edge environment, a larger image will consume more time, and the possibility of network problems during the distribution process will be higher. It will cause operational difficulties for the deployment and upgrade of the carrying service, and it will also pose more challenges to the stability of the network. high requirements

Difficulties in the storage process Due to the relatively large storage restrictions in edge scenarios, an overly large image will pose challenges to storage management in edge environments

[0011] 3. At present, when building an image, it is necessary to specify the operating system when the service is running, which will cause some inconvenience or problems

For example, if there is no image of the target operating system in the environment where the image is built, then there is no way to build it; once the image is built, it is bound to the operating environment in the image, and there is no way to flexibly adjust the service according to the actual operating environment. For example, there is no way to flexibly adjust the program from the centos operating environment to the lighter alpine operating environment; to adapt the same service to different operating systems, mirroring needs to be made separately, resulting in a waste of time and labor costs

[0012] 4. At present, when building an image, it is necessary to specify the operating environment when the service is running

This may cause some inconvenience or problems

For example: if there is no image of the target operating environment in the environment where the image is built, then there is no way to build the image; after the image has been built, it is bound to the operating environment in the image, and there is no way to flexibly according to the actual operating environment Adjustment, for example, the war package program of the java program cannot be flexibly adjusted from Tomcat version 7.0 to version 8.5, or from Tomcat to Kingdee or Dongfangtong, etc.; the same service needs to make different images to adapt to different operating environments, Causes waste of time and labor costs; in the process of use, if you want to expand the operating environment of the image, you need to rebuild the image, for example, when you add requirements such as injecting remote debugging or link tracking into the image, you need to change the configuration of the operating environment , the traditional image construction method needs to rebuild the image, which is more complicated to update; in the process of use, if a certain operating environment in the image carried on the cloud platform needs to be collectively upgraded, the traditional use method needs to update all the images Reconstruction, time, manpower, security costs are huge, etc.

[0013] 5. The traditional method of pulling the service program from the specified location through the script when the container starts needs to introduce additional components for support

This introduces new inconveniences or problems

For example, the process of pulling services cannot be differentially transmitted, which may take a long time, causing the container to take a long time to provide business functions after startup, which offsets the advantages of fast startup of container technology; the pulled files cannot be guaranteed Correct, there are loopholes in the security of the whole process, and the advantages of security inspection and certification of the image itself are lost

Images