Redis - cache penetration, cache breakdown, cache avalanche

brief introduction

about Redis Cache , Use Redis The cache of , It improves application performance and efficiency , And caching plays a very important role in high concurrency scenarios , For data consistency ,Redis Caching is a very fatal problem , There are three such problems .
1、 Cache penetration
2、 Cache breakdown
3、 Cache avalanche
Then we will explain these three situations respectively , And give the solution ：

Cache penetration

By default , When a user requests data , It's going to cache first (Redis) Search for , If not found, the cache misses , Then search in the database , A small number may not be a problem , But once a lot of data is requested （ For example, the second kill scene ） If the cache misses , It will all be transferred to the database , It puts a lot of pressure on the database , It may cause the database to crash . In network security, some people maliciously use this means to attack, which is called flood attack .

Solution ：
1、 The bloon filter
For all possible query parameters, use Hash Form storage of , In order to quickly determine whether this value exists , In the control layer, the interception verification is performed first , If the verification fails, call back directly , Reduces the pressure on the storage system .

2、 Caching empty objects
If a request is not found in the cache or database , An empty object in the cache is used to process the subsequent request .

​ Be careful ：
There is a flaw in doing so , Storing empty objects also requires space , A large number of empty objects will consume a certain amount of space , Storage efficiency is not high . The solution to this flaw is to set a shorter expiration time , Even if expiration time is set for null value , There will be some inconsistency between the data of cache layer and storage layer for a period of time , This has an impact on businesses that need to be consistent .

Cache breakdown

​ Compared to cache penetration , It's more purposeful , An existing key, At the moment the cache expires , There are a lot of requests at the same time , All of these requests go to DB, Cause instantaneous DB A lot of requests 、 The pressure surged . This is the cache breakdown , Just for one of them key The cache of is not available and causes a breakdown , But the others key You can still use cached responses , For example, in the hot search ranking , When a hot news is accessed by a large number of people at the same time, it may lead to cache breakdown .

Solution ：
1、 Never expired data settings
In this way, the hot data will not be out of date , But when Redis When the memory space is full, some data will also be cleaned up , And it takes up space , Once there's more hot data , It's going to take up part of the space , So it's not recommended .
2、 Add mutex lock ( Distributed lock )
During a visit to key Before , use SETNX（set if not exists） To set up another short term key To lock in the present key The interview of , Delete the short term after the visit key. Ensure that only one thread can access at the same time . In this way, the requirements for locks are very high .

Cache avalanche

Cache avalanche means , At the same time , A large number of caches , Collective failure , It's like it hasn't been cached , At this time, it is assumed that there is such a second kill activity of double 11 and double 12 , Just at this moment, the cache is invalidated in batches , Then the cache fails , These requests are also directly pressed on the database , If the server is unlocked , This flow reaches a peak almost instantaneously , For a server , There will also be downtime .

Solution ：
1、redis High availability
The meaning of this idea is , since redis It's possible to hang up , I'll add more redis, After this one goes down, others can continue to work , In fact, it's a cluster built .
2、 Current limiting the drop
The idea of this solution is , After cache failure , Control the number of threads that read the database write cache by locking or queuing . For example, to some key Only one thread is allowed to query data and write cache , Other threads wait .
3、 Data preheating
Data heating means before deployment , I'll go through the possible data first , In this way, some of the data that may be accessed in large amounts will be loaded into the cache . Manually trigger loading cache before large concurrent access occurs key, Set different expiration times , Make the cache failure time as uniform as possible .