Service privatization method and device, computer equipment and storage medium

Abstract

The invention discloses a service privatization method and device, computer equipment and a storage medium. The method comprises the steps that an obtained service code is compiled; Based on the compiling result, making a mirror image of the service through a docker container engine technology; And according to the mirror image, generating an installation package of the service, and storing the installation package in a predetermined position, so that after the installation package is downloaded to a server of a user, when a one-key deployment request of the user is obtained, the deployment ofthe service on the server is completed according to the installation package. By applying the scheme provided by the invention, the development and maintenance cost and the like can be reduced.

Description

【Technical field】 [0001] The invention relates to computer application technology, in particular to a service privatization method, device, computer equipment and storage medium. 【Background technique】 [0002] At present, artificial intelligence (AI, Artificial Intelligence) technology is being widely used, involving various fields such as business, finance, security, entertainment, etc. The more common service provider solution is public cloud service. Many companies provide AI open platforms, but in banks In scenarios involving private data such as finance and security, privatization and deployment of services to customer (user) computer rooms is the most needed and ideal solution for customers. [0003] At present, in terms of service privatization deployment in the industry, the generally adopted model is to run the service (software) directly on the operating system (OS, Operating System) of the client server, and use different versions of deployment packages or patche...

AI Technical Summary

Problems solved by technology

[0004] However, service privatization faces a large customer base, and the OS versions of each customer's server may be different. The service runs directly on the customer server's OS, and needs to be adapted to centos, Ubuntu, redhat, and suse, etc. version of the operating system, the development and maintenance costs of software compatibility are very high

In addition, due to the complex environment dependencies involved in the deep learning framework, and to a certain extent relying on hardware support such as graphics processing units (GPU, Graphics Processing Unit), the deployment is very difficult, and it is necessary to deploy these environments in real time at the customer site, thereby increasing Increased deployment costs, etc.

Images