Data double write consistency between redis and MySQL

1. What is? Redis And MySQL Data double write consistency

2. Synchronous write through and asynchronous cache

3. Several update strategies for database and cache consistency *

Don't allow ： Update cache first , The strategy of updating the database

1） Update the database first , Update the cache again

Generally do not use
Update the database first , Update the cache again , If Redis Something unusual happened , There will be a problem that the cache reads dirty data

1 To update mysql The inventory of a commodity , The current stock of goods is 100, Updated to 99 individual .
2 To update mysql It is amended as follows 99 success , And then update redis.
3 At this time, it is assumed that the exception appears , to update redis failed , This leads to mysql The inventory inside is 99 and redis Inside is still 100 .
4 The above happened , Will make the database and cache redis The data inside is inconsistent , Read dirty data

2） So let's delete the cache , Update the database

Low concurrent , Small plant
If the database update fails , Lead to B When a thread requests to access the cache again , Find out redis There's no data in it , Cache missing , Read again mysql when , Read old value from database

Low concurrent ： Write the old data to the cache
High concurrency ： Cache breakdown
technological process ：
（1） request A Write operation , After deleting the cache , The work is in progress …A It hasn't been completely updated yet
（2） request B starts , Inquire about redis Found cache does not exist
（3） request B continue , Go to the database and get myslq The old value in
（4） request B Write the old value to redis cache
（5） request A Write the new value to mysql database
This leads to inconsistencies .

solve ： Delay double delete strategy


3） Update the database first , Delete the cache

If the cache deletion fails or it is too late , Cause a request to visit again redis Cache hit , What is read is the old cache value .
Enterprises often use
Solution ：


1 The cache value to be deleted or the database value to be updated can be temporarily stored in the message queue （ For example, using Kafka/RabbitMQ etc. ）.
2 When the program fails to successfully delete the cache value or update the database value , These values can be re read from the message queue , Then delete or update again .
3 If you can successfully delete or update , We need to remove these values from the message queue , So as not to repeat the operation , here , We can also ensure that the database and cached data are consistent , Otherwise, you need to try again
4 If you try again more than a certain number of times and still fail , We need to send an error message to the business layer , Inform the operation and maintenance personnel .

summary