A method for centralized back-to-origin and origin station abnormal state protection

Abstract

The invention discloses a centralized back-to-source and source station abnormal state protection method, including: centralized back-to-source: through data center scheduling, scheduling requests belonging to a designated user space to a designated transit node back to the source or multiple transit nodes Centralized back-to-source; source site abnormality protection: By recording the status code data of each user space, for users with abnormal data, calculate the abnormal cache time that needs to be set, and set the abnormal cache to protect the abnormal user's source site. The present invention is different from the traditional ways of modifying routes or DNS resolution, and realizes dynamic centralized back-to-source configuration through the Redis cluster master-slave synchronization mode, and realizes the goals of high-speed back-to-source, high hit rate, and high reliability. As well as the source site abnormal state protection, by collecting the number, frequency, and duration of the abnormal status codes responded by the source site, calculate and set the exception cache, reduce the pressure on the source site, and protect the source site where abnormal errors occur.

Description

technical field [0001] The present invention relates to the technical field of content distribution network, in particular to a method for centralized back-to-source and source station abnormal state protection. Background technique [0002] CDN (Content Delivery Network), that is, content distribution network, aims to solve the bottleneck of Internet data transmission. By caching resource data at each edge node, users can directly obtain resources in the cache of the nearest edge node, avoiding going to the source station, and reducing user traffic. The time spent in network transmission. [0003] At present, the selection of the traditional CDN back-to-source line is mostly based on the network line speed measurement results, the user's back-to-source IP address and other information. When the amount of user requests is extremely large, even if a second-level cache is implemented at the transit node, due to the uneven distribution of the cache, the amount of requests arri...

AI Technical Summary

Problems solved by technology

This technical solution cannot reasonably reduce the request to return to the origin when there are many abnormalities in the origin station, and reduce the pressure on the origin station and useless CDN resource consumption.

Images