Redis sentinel mode

Regularly detect whether the host fails , If it fails, it will automatically change from machine to host according to voting

      

Adjustment structure   First complete the mode of one master and two servants ,6379 With 80、81

Self defined /myredis New under the directory sentinel.conf file , The name must not be wrong

Deploy sentinels , Fill in the content

sentinel monitor The name of the monitored host ( Name yourself ) 127.0.0.1 6379 1   Such as     

sentinel monitor host6379 127.0.0.1 6379 1

Last on top A number 1, After the host is hung up salve Vote to see who takes over as the host , A large number of votes Of Become a host

cd /usr/local/redis/bin/

Activate the sentry   ./redis-sentinel ../conf/sentinel.conf

Normal master-slave demonstration

The original master Hang up

Vote for the new election

Start again from the master and slave ,info replication see

problem ： If the previous master Restart back , Will it be double master Conflict ？  Can't Because the previous host The data is incomplete

Because all the writing operations are first in Master On the operation , Then sync update to Slave On , So from Master Synchronize to Slave The machine has a certain delay , When the system is busy , The delay problem will be more serious ,Slave The increase in the number of machines will make the problem more serious .