Scheduling method, device, medium, equipment and system for realizing service-oriented computing resources

Abstract

The present invention relates to a scheduling method for realizing computing resources as a service, including: using an idle physical machine to apply for a CPU-limited virtual machine to build a computing resource pool; when receiving a task request, use a container image to package and encapsulate user program code and Depending on the operating environment and related libraries; scheduling task requests according to the load and task busyness of each virtual machine in the resource pool; automatically increasing and reducing the number of containers according to the amount of real-time concurrent calls. Compared with the prior art, the present invention makes full use of idle resources, and realizes computing resource service through on-demand scheduling and automatic scaling of containers. In addition, the present invention also relates to a scheduling device, medium, equipment and system for realizing computing resource service.

Description

technical field [0001] The invention relates to a scheduling method, device, medium, equipment and system for realizing computing resource service. Background technique [0002] In the operation life cycle of Internet products, it is necessary not only to apply for computing resources to cope with business growth, but also to release computing resources to avoid idle waste during business downturns. Traditionally, operation and maintenance personnel need to order physical servers, select a computer room to install racks, debug hardware, and then deploy server-side software programs, and finally the product can be launched as scheduled. This manual operation and maintenance method is inefficient, has a lot of repetitive work, and the process of each link is cumbersome, and the product launch and iteration are slow. In addition, because physical computing resources are usually not fully utilized, a large number of servers have extremely low loads during off-peak periods, and ...

AI Technical Summary

Problems solved by technology

[0004] However, although the IAAS platform frees operation and maintenance personnel from tedious labor such as purchasing and installing physical equipment, developers still need to install operating systems and various software for virtual machines, and operation and maintenance personnel need to manually focus on computing, network, and storage resources. expansion and contraction of

In addition, virtual machines still have low-load resource waste

At the same time, there are serious limitations in the use scenarios of AWS Lambda, for example, it only supports code written in some languages; the maximum running time of tasks is only 5 minutes, which cannot support long-term offline task programs; it does not support persistent service tasks, and each request It is necessary to create a container, and it is difficult to achieve low-latency (within 10ms) response, etc.

Images