A method and device for gray scale downgrading at the server side

Abstract

The embodiments of the invention provide a service end gray scale degrading method and device. The method and device are applied to a client, and the method concretely includes the following steps: determining whether data comes from a cache according to source identification in the data when the data returned by a server is received, and determining update time of the data according to a timestamp in the data; if the data comes from the cache, caching a write operation instruction and the write operation timestamp if the execution of the write operation instruction is successful when a user sends the write operation instruction to the server; and rewriting the data by using the write operation instruction locally cached if the data still comes from the cache and the update time of the data is earlier than the time of the write operation timestamp when the user reads the data repeatedly. Therefore, the load of a page service layer of the server is effectively reduced, and the server does not need to be temporarily or permanently deployed for sudden increases in the load. Therefore, the problem of excessive load under high concurrency condition can be effectively solved without increasing the cost.

Description

technical field [0001] The present invention relates to the field of network technology, in particular to a method and device for gray scale downgrading at a server. Background technique [0002] For websites, at certain time points, such as holidays, new drama launch days, concert live broadcast days, shopping festivals, etc., due to the sudden massive user visits, the QPS will increase instantly, and most of the data and user Behavior-related, in this case, the cache cannot be used, which puts a huge pressure on the server. In particular, the Servlet layer (page service layer) in the middle of the system not only needs to adjust the back-end interface, but also assembles data and returns it to the client, which leads to a sharp increase in the CPU usage of the server, even approaching the dangerous threshold. The current measure is to increase the number of servers to achieve the effect of load balancing, including permanent or temporary increases, but this will lead to a...

AI Technical Summary

Problems solved by technology

[0002] For websites, at certain time points, such as holidays, new drama launch days, concert live broadcast days, shopping festivals, etc., due to the sudden massive user visits, the QPS will increase instantly, and most of the data and user Behavior-related, in this case, the cache cannot be used, which puts a huge pressure on the server

In particular, the Servlet layer (page service layer) in the middle of the system not only needs to adjust the back-end interface, but also assembles data and returns it to the client, which leads to a sharp increase in the CPU usage of the server, even approaching the dangerous threshold

The current measure is to achieve load balancing by increasing the number of servers, including permanent or temporary increases, but this will lead to a sharp increase in the cost of the website, which is not good for the operation of the website

Images