Problems and solutions of several concurrent scenarios of redis

brief introduction

 

redis As an intermediate cache between application and database , When users access the data source, they will first access redis, If there is no data in the query, directly query the database , When you find out , The returned data will be loaded into redis Inside . In use ,redis In the concurrent scenario , Cache penetration exists , Cache breakdown and cache avalanche .

Cache penetration

problem

When accessing key When it doesn't exist , The request will always be on the database , Increased database pressure .

Solutions

One You can cache null values first key

Two Filter abnormal requests

Cache breakdown

problem

When key After expired , A large number of requests are directly pressed on the database , May cause the database to crash

Solutions

One Preload popular key

Two To extend properly key Time for

3、 ... and Set lock （ Low efficiency ）

Cache avalanche

problem

The cache key A lot of overdue , It may cause the data cache update to fail

Solutions

One take key The time of failure is scattered

Two Use locks or queues

3、 ... and Regularly update key