Redis master-slave architecture lock failure problem (master-slave)

Hello everyone , I meet you again , I'm your friend, Quan Jun .

One 、 Get ready 1、 modify pidfile And port

2、 close RDB Persistent modify the storage location of the persistent file

3、 start-up Redis

redis-server /etc/redis.conf

Copy

4、 Use client connection Redis

redis-cli

Copy

Two 、 Master slave copy ( Read / write separation ) redis The master-slave replication function of is very powerful , One master You can have multiple slave, And one slave You can have more than one slave, Go on like this , Formed a powerful multi-level server cluster architecture . You can avoid redis A single point of failure , Build a read-write separation architecture , Meet the application scenario of reading more and writing less . 1、 Master slave replication principle

①slave Initiate request and master Establishing a connection ,master After verification, the connection can be established . ②slave Send synchronization sync The order of , At this time, the main database will start a new sub process , Import data into by snapshot rdb In file , And transfer it to the slave library ③ From the library rdb The file is imported into the database , And load it into memory ④ When synchronizing later ,master Will give all orders first buffer get up , Can't write to disk , Direct to slave. remarks ： If master restart , The backup data will be re dump.

2、Redis Several important aspects of replication

① One Master There can be multiple Slave; ②Redis Use asynchronous replication . from 2.8 Start ,Slave It will be cyclical （ Once a second ） To launch a Ack Confirm the copy stream （replication stream） The progress of being processed ; ③ Not only the master server can have slave servers , The slave server can also have its own slave server , Multiple slave servers can form a graphical structure ④ Copy on Master The end is non blocking , This means that even multiple Slave When performing the first synchronization ,Master You can still provide query services ; ⑤ Copy on Slave The end is also non blocking ： If you are in the redis.conf Set it up ,Slave When performing the first synchronization, you can still use the old dataset to provide queries ; You can also configure to be Master And Slave When you lose contact , Give Way Slave Return an error prompt to the client ; ⑥ When Slave Delete the old data set , And reload the new version of data ,Slave Will block connection requests （ It usually happens with Master Recovery phase after disconnection and reconnection ）; ⑦ Replication can be used for data redundancy only （dataredundancy）, You can also improve scalability by having multiple read-only command requests processed from the server （scalability）： for instance , Heavy SORT Commands can be given to affiliated nodes to run . ⑧ It can be modified by Master Terminal redis.config To avoid Master Perform persistence operation on the end （Save）, from Slave End to end persistence .

3、 ... and 、 Master slave architecture

1、 Get ready 3 The configuration file ports are

6379 (Master)
6380 (Slave)
6381 (Slave)

Copy

2、 Modify the original redis.conf file , Copy out 2 individual redis.conf file

mv /etc/redis.conf /etc/redis.6379.conf
cp /etc/redis.6379.conf /etc/redis.6380.conf 
cp /etc/redis.6379.conf /etc/redis.6381.conf

Copy

3、 modify 6380 and 6381 The configuration file vim /etc/redis.6380.conf

Replace by command 6379 by 6380

:%s/6379/6380/g

Copy

Appear at the bottom

Indicates the modification is successful , wq Exit and save

4、 Modify in the same way 6381 Configuration file for 5、 start-up 3 individual redis example

redis-server /etc/redis.6379.conf
redis-server /etc/redis.6380.conf
redis-server /etc/redis.6381.conf

Copy

6、 adopt ps Command view redis process

ps -ef | grep redis

Copy

7、 The master-slave configuration includes 2 Methods : ① On all slave nodes redis.conf(redis.6380.conf and redis.6381.conf) Set in slaveof

② Use redis-cli The client connects to Redis In service , perform slaveof command This method will lose the master-slave replication relationship after restart

8、 View master-slave information ：INFO replication ① Main database query

② Information displayed from the Library

③ Test the master-slave relationship Write data in the main library , Then read data from the Library

Publisher ： Full stack programmer stack length , Reprint please indicate the source ：https://javaforall.cn/127911.html Link to the original text ：https://javaforall.cn