下载redis
wget http://download.redis.io/releases/redis-5.0.5.tar.gz
创建用户及目录
groupadd redis
useradd -g redis redis
mkdir -p /u01/redis
解压及安装
gzip -d redis-5.0.5.tar
tar -xf redis-5.0.5.tar -C /u01/redis
cd /u01/redis/redis-5.0.5
mkdir data log run
make
maie install (root用户执行否则报错）
redis 配置
编辑 .bash_profile
PATH=$PATH:$REDIS_HOME/bin:$HOME/bin:/usr/local/bin
export PATH
alias c='cd /u01/redis/redis-5.0.5'
编辑redis.conf
1 daemonize yes #后台启动 2 port 7001 #修改端口号，从7001到7006 3 maxmemory 100M #设置最大内存 4 maxmemory-policy allkeys-lru #内存达到最大值时策略 5 pidfile /u01/redis/redis-5.0.5/run/redis_7001.pid 6 logfile /u01/redis/redis-5.0.5/log/7001.log 7 dir /u01/redis/redis-5.0.5/data/7001 8 appendonly yes
复制redis.conf到对应的目录
/u01/redis/redis-5.0.5/conf
[redis@redis01 conf]$ ls
7001 7002 7003 7004 redis.conf redis-sentinel-26379.conf
[redis@redis01 conf]$ cd 7001
[redis@redis01 7001]$ ls
7001.conf nodes.conf
编辑启动脚本
[redis@redis01 ~]$ cat s.sh
pkill -9 redis
redis-server /u01/redis/redis-5.0.5/conf/7001/7001.conf
redis-server /u01/redis/redis-5.0.5/conf/7002/7002.conf
ps -ef|grep redis-server|grep -v grep
在复制节点启动复制
slaveof 192.168.20.201 7001
启动日志如下
1 9713:S 16 Sep 2019 13:28:34.459 * REPLICAOF 192.168.20.201:7001 enabled (user request from 'id=3 addr=192.168.20.202:33566 fd=9 name= age=107 idle=0 flags=N db=0 sub=0 psub=0 multi=-1 qbuf=48 qbuf-free=32720 obl=0 oll=0 omem=0 events=r cmd=slaveof') ---建立连接套接字 2 9713:S 16 Sep 2019 13:28:35.309 * Connecting to MASTER 192.168.20.201:7001 --连接master 3 9713:S 16 Sep 2019 13:28:35.310 * MASTER <-> REPLICA sync started 4 9713:S 16 Sep 2019 13:28:35.310 * Non blocking connect for SYNC fired the event. 5 9713:S 16 Sep 2019 13:28:35.310 * Master replied to PING, replication can continue... --从节点发送ping命令返回 6 9713:S 16 Sep 2019 13:28:35.312 * Trying a partial resynchronization (request ce82246b3aecfacf28105aba18eecbb7407e7679:1). 7 9713:S 16 Sep 2019 13:28:35.314 * Full resync from master: 5ad71d8d15caad30b77a50fd6c22eaf088f936dd:0 --开始全量同步数据
查看复制
[redis@redis01 ~]$ redis-cli -h 192.168.20.201 -p 7001 info replication # Replication role:master connected_slaves:2 slave0:ip=192.168.20.202,port=7003,state=online,offset=504,lag=0 slave1:ip=192.168.20.202,port=7004,state=online,offset=504,lag=1 master_replid:15fac3f45a2af5a4421a9906e7e1898abd6713d4 master_replid2:0000000000000000000000000000000000000000 master_repl_offset:504 second_repl_offset:-1 repl_backlog_active:1 repl_backlog_size:1048576 repl_backlog_first_byte_offset:1 repl_backlog_histlen:504 [redis@redis01 ~]$ redis-cli -h 192.168.20.202 -p 7003 info replication # Replication role:slave master_host:192.168.20.201 master_port:7001 master_link_status:up master_last_io_seconds_ago:4 master_sync_in_progress:0 slave_repl_offset:560 slave_priority:100 slave_read_only:1 connected_slaves:0 master_replid:15fac3f45a2af5a4421a9906e7e1898abd6713d4 master_replid2:0000000000000000000000000000000000000000 master_repl_offset:560 second_repl_offset:-1 repl_backlog_active:1 repl_backlog_size:1048576 repl_backlog_first_byte_offset:1 repl_backlog_histlen:560
这是正常情况下复制，如果在同步过程当中出现网络中断或者主机重启等可以用psync部分复制
psync{runId}{offset}，参数含义如下：
·runId：从节点所复制主节点的运行id。
·offset：当前从节点已复制的数据偏移量。
redis@redis01 ~]$ redis-cli -h 192.168.20.201 -p 7001 info server|grep run
run_id:92127d4880d2c7a4c5c6da3dcb73d584fc002817
psync 92127d4880d2c7a4c5c6da3dcb73d584fc002817 -1
优化
如果RDB文件超过6GB以上时要格外小心。传输文件这一步操作非常耗时，需要加大repl-timeout所配置的值（默认60秒）
停止同步
192.168.20.202:7004> slaveof no one OK 10701:S 16 Sep 2019 21:36:43.615 # CONFIG REWRITE executed with success. 10701:S 16 Sep 2019 21:36:44.138 * Connecting to MASTER 192.168.20.202:7003 10701:S 16 Sep 2019 21:36:44.138 * MASTER <-> REPLICA sync started 10701:S 16 Sep 2019 21:36:44.138 * Non blocking connect for SYNC fired the event. 10701:S 16 Sep 2019 21:36:44.139 * Master replied to PING, replication can continue... 10701:S 16 Sep 2019 21:36:44.139 * Trying a partial resynchronization (request a32990ebbf807015eb5186ce88e049b1cae7cee9:4129548). 10701:S 16 Sep 2019 21:36:44.141 * Full resync from master: e9e7b0c3af955da1330327c91c1ca8f7705fae6d:4127868 10701:S 16 Sep 2019 21:36:44.142 * Discarding previously cached master state. 10701:S 16 Sep 2019 21:36:44.239 * MASTER <-> REPLICA sync: receiving 256 bytes from master 10701:S 16 Sep 2019 21:36:44.239 * MASTER <-> REPLICA sync: Flushing old data 10701:S 16 Sep 2019 21:36:44.261 * MASTER <-> REPLICA sync: Loading DB in memory 10701:S 16 Sep 2019 21:36:44.262 * MASTER <-> REPLICA sync: Finished with success 10701:S 16 Sep 2019 21:36:44.264 * Background append only file rewriting started by pid 14688 10701:S 16 Sep 2019 21:36:44.355 * AOF rewrite child asks to stop sending diffs. 14688:C 16 Sep 2019 21:36:44.355 * Parent agreed to stop sending diffs. Finalizing AOF... 14688:C 16 Sep 2019 21:36:44.355 * Concatenating 0.00 MB of AOF diff received from parent. 14688:C 16 Sep 2019 21:36:44.403 * SYNC append only file rewrite performed 14688:C 16 Sep 2019 21:36:44.404 * AOF rewrite: 8 MB of memory used by copy-on-write 10701:S 16 Sep 2019 21:36:44.446 * Background AOF rewrite terminated with success 10701:S 16 Sep 2019 21:36:44.446 * Residual parent diff successfully flushed to the rewritten AOF (0.00 MB) 10701:S 16 Sep 2019 21:36:44.446 * Background AOF rewrite finished successfully