Redis persistence mechanism

How to guarantee  Redis  Data can be recovered by restarting after being hung up ？

Many times we need to persist data, that is, write the data in memory to the hard disk , Most of the reason is to reuse the data later （ For example, restart the machine 、 Recover data after machine failure ）, Or backup data to a remote location to prevent system failure .

Redis  differ  Memcached  It's very important that ,Redis  Support persistence , And it supports two different persistence operations .

Redis  One of the persistence methods of is snapshot （snapshotting,RDB）, Another way is to just append files （append-only file, AOF）

. These two methods have their own advantages , I'll elaborate on what these two persistence methods are , How to use it? , How to choose your own persistence method .

What is?  RDB  Persistence ？

Redis  You can create a snapshot to get a copy of the data stored in memory at a certain point in time .Redis  After creating the snapshot , You can back up the snapshot , You can replicate the snapshot to another server to create a server copy with the same data （Redis  Master-slave structure , Mainly used to improve  Redis  performance ）, You can also leave the snapshot in place for use when you restart the server .

Snapshot persistence is  Redis  The default persistence method , stay  

redis.conf

  The configuration file has this configuration by default ：

save 900 1 # stay 900 second (15 minute ) after , If at least 1 individual key change ,Redis Will trigger automatically BGSAVE Command create snapshot .

save 300 10 # stay 300 second (5 minute ) after , If at least 10 individual key change ,Redis Will trigger automatically BGSAVE Command create snapshot .

save 60 10000 # stay 60 second (1 minute ) after , If at least 10000 individual key change ,Redis Will trigger automatically BGSAVE Command create snapshot .


What is?  AOF  Persistence ？

Compared to snapshot persistence ,AOF  Persistence is better in real time , So it has become a mainstream persistence solution . By default  Redis  It's not turned on  AOF（append only file） Persistence of the way , Can pass  appendonly  Parameter on ：

appendonly yes


Turn on  AOF  Every execution will change after persistence  Redis  Commands for data in ,Redis  The command will be written to the memory cache  

server.aof_buf

  in , And then based on  

appendfsync

  Configuration to determine when to synchronize it to the disk  AOF  file .

AOF  The location and  RDB  The location of the files is the same , It's all through  dir  Parameter setting , The default file name is  

appendonly.aof

.

stay  Redis  There are three different profiles for  AOF  Persistence mode , They are ：

appendfsync always # Write every time a data change occurs AOF file , This will seriously reduce Redis The speed of
appendfsync everysec # Sync every second , Explicitly synchronize multiple write commands to the hard disk
appendfsync no # Let the operating system decide when to synchronize

In order to balance data and write performance , Users can consider  

appendfsync everysec

  Options  , Give Way  Redis  Synchronize once per second  AOF  file ,Redis  Performance is almost unaffected . And even if there is a system crash , Users can only lose data generated within one second at most . When the hard disk is busy performing write operations ,Redis  It will also gracefully slow down its own speed to adapt to the maximum write speed of the hard disk .

expand ： Interview questions ：AOF Rewrite mechanism ,redis The interview will be asked ！！！