0x001 修改配置
- 复制两份配置文件，分别命名为
redis_6378.conf、redis_6377.conf。他们将在6378、6377两个端口启动 分别修改两个配置：
# redis_6378.conf ... port 6378 slaveof 127.0.0.1 6379 ...# redis_6377.conf ... port 6377 slaveof 127.0.0.1 6378 ...
说明：这里使用监听6379的节点作为主节点，6378作为6379的子节点，6377作为6378的子节点。
0x002 同步
启动
6379主节点$ redis-server 68931:C 29 May 10:53:54.234 # oO0OoO0OoO0Oo Redis is starting # 省略部分过程 _.-`` `. `_. ''-._ Redis 4.0.8 (00000000/0) 64 bit .-`` .-```. ```\/ _.,_ ''-._ ( ' , .-` | `, ) Running in standalone mode |`-._`-...-` __...-.``-._|'` _.-'| Port: 6379 | `-._ `._ / _.-' | PID: 68931 # 省略部分过程 68931:M 29 May 10:53:54.237 * Ready to accept connections启动
6378子节点$ redis-server ~/Desktop/redis_6378.conf # 省略部分内容 _._ _.-``__ ''-._ _.-`` `. `_. ''-._ Redis 4.0.8 (00000000/0) 64 bit .-`` .-```. ```\/ _.,_ ''-._ ( ' , .-` | `, ) Running in standalone mode |`-._`-...-` __...-.``-._|'` _.-'| Port: 6378 | `-._ `._ / _.-' | PID: 69033 # 省略部分内容 # 这里说明启动成功了 69033:S 29 May 10:56:24.581 * Ready to accept connections # 开始连接主节点 69033:S 29 May 10:56:24.581 * Connecting to MASTER 127.0.0.1:6379 # 开始同步 69033:S 29 May 10:56:24.581 * MASTER <-> SLAVE sync started 69033:S 29 May 10:56:24.581 * Non blocking connect for SYNC fired the event. 69033:S 29 May 10:56:24.581 * Master replied to PING, replication can continue... # 尝试增量同步 69033:S 29 May 10:56:24.581 * Trying a partial resynchronization (request 87953c7dc97cf1192a243fe5d7708c11acbd0e0f:1). # 全量同步 69033:S 29 May 10:56:24.582 * Full resync from master: 98a56521a0a958295e499853621029d19ef6b357:0 69033:S 29 May 10:56:24.582 * Discarding previously cached master state. 69033:S 29 May 10:56:24.684 * MASTER <-> SLAVE sync: receiving 826 bytes from master 69033:S 29 May 10:56:24.685 * MASTER <-> SLAVE sync: Flushing old data 69033:S 29 May 10:56:24.685 * MASTER <-> SLAVE sync: Loading DB in memory 69033:S 29 May 10:56:24.685 * MASTER <-> SLAVE sync: Finished with success此时的
6379主节点# 省略部分内容 68931:M 29 May 10:53:54.237 * Ready to accept connections # 上面是之前的状态，下面是`6378`连接进来以后的状态 68931:M 29 May 10:56:24.581 * Slave 127.0.0.1:6378 asks for synchronization 68931:M 29 May 10:56:24.581 * Partial resynchronization not accepted: Replication ID mismatch (Slave asked for '87953c7dc97cf1192a243fe5d7708c11acbd0e0f', my replication IDs are 'dbf6c826156fa5140d891996afcf8cc4f21b61fd' and '0000000000000000000000000000000000000000') 68931:M 29 May 10:56:24.581 * Starting BGSAVE for SYNC with target: disk 68931:M 29 May 10:56:24.582 * Background saving started by pid 69034 69034:C 29 May 10:56:24.584 * DB saved on disk 68931:M 29 May 10:56:24.684 * Background saving terminated with success # 同步成功 68931:M 29 May 10:56:24.684 * Synchronization with slave 127.0.0.1:6378 succeeded启动
6377子节点redis-server ~/Desktop/redis_6377.conf # 省略部分内容 _._ _.-``__ ''-._ _.-`` `. `_. ''-._ Redis 4.0.8 (00000000/0) 64 bit .-`` .-```. ```\/ _.,_ ''-._ ( ' , .-` | `, ) Running in standalone mode |`-._`-...-` __...-.``-._|'` _.-'| Port: 6377 | `-._ `._ / _.-' | PID: 69207 # 启动完成 69207:S 29 May 11:02:22.569 * Ready to accept connections # 连接上一级节点 69207:S 29 May 11:02:22.569 * Connecting to MASTER 127.0.0.1:6378 69207:S 29 May 11:02:22.570 * MASTER <-> SLAVE sync started 69207:S 29 May 11:02:22.570 * Non blocking connect for SYNC fired the event. 69207:S 29 May 11:02:22.570 * Master replied to PING, replication can continue... 69207:S 29 May 11:02:22.570 * Trying a partial resynchronization (request 98a56521a0a958295e499853621029d19ef6b357:1). 69207:S 29 May 11:02:22.570 * Successful partial resynchronization with master. 69207:S 29 May 11:02:22.570 * MASTER <-> SLAVE sync: Master accepted a Partial Resynchronization.
0x003数据写入
使用
redis-cli向6379主节点写入数据$ redis-cli -h 127.0.0.1 -p 6379 127.0.0.1:6379> set name redis OK 127.0.0.1:6379> get name "redis"使用
redis-cli连接6378子节点并读取在主节点写入的数据$ redis-cli -h 127.0.0.1 -p 6378 127.0.0.1:6378> get name "redis"使用
redis-cli连接637 7子节点并读去在主节点写入的数据$ redis-cli -h 127.0.0.1 -p 6377 127.0.0.1:6377> get name "redis"使用
redis-cli连接6378子节点并尝试写入数据127.0.0.1:6378> set name redis-cli (error) READONLY You can't write against a read only slave.
说明：子节点可以读取到主节点写入的数据，说明同步成功了，子节点无法写入数据，因为子节点是默认只读的，修改配置可以做到子节点可读写
0x004 断开连接
关闭
6378子节点$ redis-cli -h 127.0.0.1 -p 6378 127.0.0.1:6378> shutdown not connected>此时的
6378# 上面是之前的状态，下面是`6378`关机时候的状态 9033:S 29 May 11:09:54.587 # User requested shutdown... 69033:S 29 May 11:09:54.587 * Saving the final RDB snapshot before exiting. 69033:S 29 May 11:09:54.588 * DB saved on disk 69033:S 29 May 11:09:54.588 * Removing the pid file. 69033:S 29 May 11:09:54.589 # Redis is now ready to exit, bye bye...- 此时的
6379
# 上面是之前的状态，下面是`6378`关机之后的状态 68931:M 29 May 11:09:54.589 # Connection with slave 127.0.0.1:6378 lost.- 此时的
此时的
6377... # 6377 会一直尝试重新连接`6378` 69219:S 29 May 11:12:07.132 * Connecting to MASTER 127.0.0.1:6378 69219:S 29 May 11:12:07.132 * MASTER <-> SLAVE sync started 69219:S 29 May 11:12:07.132 # Error condition on socket for SYNC: Connection refused 69219:S 29 May 11:12:07.132 * Connecting to MASTER 127.0.0.1:6378 69219:S 29 May 11:12:07.132 * MASTER <-> SLAVE sync started 69219:S 29 May 11:12:07.132 # Error condition on socket for SYNC: Connection refused ...
0x005 子节点可读
修改配置
# redis_6378.conf slave-read-only no重新启动
6378，并向6378写入数据$ redis-cli -h 127.0.0.1 -p 6378 # 此时配置子节点不是只读的，所以可以写入 127.0.0.1:6378> set name redis-cli OK 127.0.0.1:6378> get name "redis-cli"连接
6377并读取6378写入的key$ redis-cli -h 127.0.0.1 -p 6377 127.0.0.1:6377> get name "redis"连接
6379并读取6378写入的key$ redis-cli -h 127.0.0.1 -p 6379 127.0.0.1:6379> GET name "redis"
说明：虽然此时的6378子节点是可读的，但是并不会传播到其他子节点或者主节点，并且在下次同步的时候，将会被删除。
