Method for implementing distributed application current limiting based on redis

The purpose of current limiting is to protect the system by limiting the speed of concurrent access / requests or requests within a time window. Once the limit rate is reached, the service can be refused.

A few days ago, in the official account of DD, I saw a program about using the redis to realize single application flow restriction. The original version of the in action is a jedis version. All of them belong to the business level restriction. Current limiting strategies commonly used in actual scenarios:

Nginx access layer current limiting

According to certain rules, such as account number, IP, system call logic, etc., the current is limited at the nginx level

Current limiting of business application system

Control traffic through business code. This traffic can be called semaphore, which can be understood as a lock, which can limit the maximum number of processes that can access a resource at the same time.

code implementation

call

Test interface call

optimization

Optimize code using interceptors + annotations

Interceptor

Define annotation

use

Concurrent testing

Tool: apache-jmeter-3.2

Note: the interface that does not get the semaphore returns 500, the status is red, the interface that gets the semaphore returns 200, and the status is green.

When the limit request semaphore is 2, 5 threads are concurrent:

When the limit request semaphore is 5, 10 threads are concurrent:

data

Implementation based on reids + Lua