分类: 数据库开发技术
2012-05-20 20:32:13
redis主从复制配置和使用都非常简单。通过主从复制可以允许多个slave server拥有和master server相同的数据库副本。
(1)、master可以拥有多个slave
(2)、多个slave可以连接同一个master外,还可以连接到其他slave
(3)、主从复制不会阻塞master,在同步数据时,master可以继续处理client请求
(4)、提高系统的伸缩性
当配置好slave后,slave与master建立连接,然后发送sync命令。无论是第一次连接还是重新连接,master都会启动一个后台进程,将数据库快照保存到文件中,同时master主进程会开始收集新的写命令并缓存。后台进程完成写文件后,master就发送文件给slave,slave将文件保存到硬盘上,再加载到内存中,接着master就会把缓存的命令转发给slave,后续master将收到的写命令发送给slave。如果master同时收到多个slave发来的同步连接命令,master只会启动一个进程来写数据库镜像,然后发送给所有的slave。
配置slave服务器很简单,只需要在slave的配置文件中加入如下配置
slaveof 192.168.1.1 6379 #指定master的ip和端口 |
下面我们做一个实验来演示如何搭建一个主从环境:
# slaveof slaveof localhost 6379 |
我们在一台机器上启动主库(端口6379),从库(端口6378)
启动后主库控制台日志如下:
[root@localhost redis-2.2.12]# src/redis-server redis.conf [7064] 09 Aug 20:13:12 * Server started, Redis version 2.2.12 [7064] 09 Aug 20:13:12 # WARNING overcommit_memory is set to 0! Background save may fail under low memory condition. To fix this issue add 'vm.overcommit_memory = 1' to /etc/sysctl.conf and then reboot or run the command 'sysctl vm.overcommit_memory=1' for this to take effect. [7064] 09 Aug 20:13:12 * The server is now ready to accept connections on port 6379 [7064] 09 Aug 20:13:13 - 0 clients connected (0 slaves), 539512 bytes in use [7064] 09 Aug 20:13:18 - 0 clients connected (0 slaves), 539512 bytes in use [7064] 09 Aug 20:13:20 - Accepted 127.0.0.1:37789 [7064] 09 Aug 20:13:20 * Slave ask for synchronization [7064] 09 Aug 20:13:20 * Starting BGSAVE for SYNC [7064] 09 Aug 20:13:20 * Background saving started by pid 7067 [7067] 09 Aug 20:13:20 * DB saved on disk [7064] 09 Aug 20:13:20 * Background saving terminated with success [7064] 09 Aug 20:13:20 * Synchronization with slave succeeded [7064] 09 Aug 20:13:23 - 0 clients connected (1 slaves), 547380 bytes in use |
启动后从库控制台日志如下:
[root@localhost redis-2.2.12]# src/redis-server redis.slave [7066] 09 Aug 20:13:20 * Server started, Redis version 2.2.12 [7066] 09 Aug 20:13:20 # WARNING overcommit_memory is set to 0! Background save may fail under low memory condition. To fix this issue add 'vm.overcommit_memory = 1' to /etc/sysctl.conf and then reboot or run the command 'sysctl vm.overcommit_memory=1' for this to take effect. [7066] 09 Aug 20:13:20 * The server is now ready to accept connections on port 6378 [7066] 09 Aug 20:13:20 - 0 clients connected (0 slaves), 539548 bytes in use [7066] 09 Aug 20:13:20 * Connecting to MASTER... [7066] 09 Aug 20:13:20 * MASTER <-> SLAVE sync started: SYNC sent [7066] 09 Aug 20:13:20 * MASTER <-> SLAVE sync: receiving 10 bytes from master [7066] 09 Aug 20:13:20 * MASTER <-> SLAVE sync: Loading DB in memory [7066] 09 Aug 20:13:20 * MASTER <-> SLAVE sync: Finished with success [7068] 09 Aug 20:13:20 * SYNC append only file rewrite performed [7066] 09 Aug 20:13:20 * Background append only file rewriting started by pid 7068 [7066] 09 Aug 20:13:21 * Background append only file rewriting terminated with success [7066] 09 Aug 20:13:21 * Parent diff flushed into the new append log file with success (0 bytes) [7066] 09 Aug 20:13:21 * Append only file successfully rewritten. [7066] 09 Aug 20:13:21 * The new append only file was selected for future appends. [7066] 09 Aug 20:13:25 - 1 clients connected (0 slaves), 547396 bytes in use |
我们在主库上设置一对键值对
redis 127.0.0.1:6379> set name HongWan OK redis 127.0.0.1:6379> |
在从库上取一下这个键
redis 127.0.0.1:6378> get name "HongWan" redis 127.0.0.1:6378> |
说明主从是同步正常的.
那么我们如何判断哪个是主哪个是从呢?我们只需调用info这个命令就可以得到主从的信息了,我们在从库上执行info命令
redis 127.0.0.1:6378> info . . . role:slave master_host:localhost master_port:6379 master_link_status:up master_last_io_seconds_ago:10 master_sync_in_progress:0 db0:keys=1,expires=0 redis 127.0.0.1:6378> |
里面有一个角色标识,来判断是主库还是从库,对于本例是一个从库,同时还有一个master_link_status用于标明主从是否异步,如果此值=up,说明同步正常;如果此值=down,说明同步异步;