A kind of server automatic intelligent expansion and contraction method and system

Abstract

The present application provides a method and system for automatic and intelligent expansion and contraction of servers. The method includes: obtaining the capacity of the current server, expansion and contraction evaluation index data, and historical expansion and contraction record data; wherein, the expansion and contraction evaluation index data includes the server's Basic index data and business index data; obtain auxiliary reference index data that affect server capacity based on historical expansion and contraction record data; calculate server expectations based on current server capacity, expansion and contraction evaluation index data, and auxiliary reference index data Capacity: According to the current server capacity and the server's expected capacity, calculate the current capacity difference that needs to be expanded and contracted; perform server expansion and contraction operations based on the capacity difference. This application can evaluate the server capacity through various indicators, analyze whether the server needs to be expanded and reduced, and the size of the expansion and contraction, and can automatically expand and contract in time according to the analysis results, so that the server can run safely, stably and continuously.

Description

technical field [0001] The present application relates to the technical field of electronic digital data processing, and in particular to a method and system for automatic and intelligent expansion and contraction of servers. Background technique [0002] Automatic scaling algorithms are currently mainly divided into Kubernetes HPA (horizontal Pod automatic scaling) and Knative (a platform for building, deploying, and managing modern serverless applications) automatic scaling KPA algorithm. [0003] HPA (Horizontal Pod Auto Scaling) Pods are automatically scaled horizontally. Through this function, with simple configuration, you can use monitoring indicators (cpu usage, disk, memory, etc.) to automatically expand or shrink the number of Pods in the service. [0004] Knative Serving injects a QUEUE proxy container (queue-proxy) for each POD, which is responsible for reporting user container concurrency metrics to Autoscaler. After receiving these indicators, the Autoscaler w...

AI Technical Summary

Problems solved by technology

[0005] The implementation of the HPA algorithm is simple. By default, it does not support shrinking to 0 or cold start of workloads, and it is not suitable for traffic burst scenarios, so there are certain hidden dangers.

[0006] KPA cannot access other custom indicators, lacks scalability and customizability

[0007] The servers in the prior art cannot expand and shrink automatically in a timely and reasonable manner, and there are certain security risks in their operation, which may easily cause waste of resources

Images