阿里云-云小站（无限量代金券发放中）
【腾讯云】云服务器、云数据库、COS、CDN、短信等热卖云产品特惠抢购

通过Keepalived实现Redis双机

204次阅读
没有评论

共计 17463 个字符，预计需要花费 44 分钟才能阅读完成。

Redis 是一个高性能的 kv 数据库，我们用 Redis 构建了一套消息系统。

Redis 架构为 Master Slave 结构，通过 keepalived 进行状态检查，故障迁移，达到高可用的目的。

下载 keepalived

#wget http://www.keepalived.org/software/keepalived-1.2.9.tar.gz

编译安装 keepalived，编译时需要指定内核目录，否则 Use IPVS Framework 为 NO。

#./configure --prefix=/usr/local/keepalived \
--with-kernel-dir=/usr/src/kernels/3.9.10-100.fc17.x86_64 \
--enable-snmp \
--enable-sha1

配置 keepalived 信息如下图：

通过 Keepalived 实现 Redis 双机

编译与安装：

#make && make install

结果报错了，重新指定配置参数，去掉 –with-kernel-dir

配置信息如下图：

通过 Keepalived 实现 Redis 双机

在次编译安装，OK，顺利通过安装成功！

现在可以将 /usr/local/keepalived 下相应文件拷贝到系统对应目录

通过 Keepalived 实现 Redis 双机

配置 keepalived

在配置 keepalived.conf 配置文件之前先备份一份 keepalived.conf 文件后在进行配置 

#cp /etc/keepalived/keepalived.conf /etc/keepalived/keepalived.conf.bak

在 Master 机器配置 keepalived

[root@node3 etc]# more /etc/keepalived/keepalived.conf
global_defs {
   notification_email {admin@qixing318.com}
   notification_email_from keepalived@node3
   router_id node3
}

vrrp_instance mes_Redis {
    state MASTER
    interface eth0
    garp_master_delay 10
    smtp_alert
    virtual_router_id 3
    priority 100
    nopreempt
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass 1111
    }
    virtual_ipaddress {192.168.172.100}
}

在 Slave 机器配置 keepalived

[root@node4 etc]# more /etc/keepalived/keepalived.conf
global_defs {
   notification_email {admin@qixing318.com}
   notification_email_from keepalived@node4
   router_id node4
}

vrrp_instance mes_Redis {
    state BACKUP
    interface eth0
    garp_master_delay 10
    smtp_alert
    virtual_router_id 3
    priority 90
    nopreempt
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass 1111
    }
    virtual_ipaddress {192.168.172.100}
}

配置防火墙

打开 Master 和 Slave 机器上的防火墙 

[root@node3 ~]# more /etc/sysconfig/iptables
# Firewall configuration written by system-config-firewall
# Manual customization of this file is not recommended.
*filter
:INPUT ACCEPT [0:0]
:FORWARD ACCEPT [0:0]
:OUTPUT ACCEPT [0:0]
-A INPUT -m state --state ESTABLISHED,RELATED -j ACCEPT
-A INPUT -p icmp -j ACCEPT
-A INPUT -i lo -j ACCEPT
-A INPUT -m state --state NEW -m tcp -p tcp --dport 22 -j ACCEPT
-A INPUT -m state --state NEW -m tcp -p tcp --dport 9022 -j ACCEPT
-A INPUT -m state --state NEW -m tcp -p tcp --dport 6379 -j ACCEPT
-A INPUT -p 112 -j ACCEPT
-A INPUT -p 255 -j ACCEPT
-A INPUT -j REJECT --reject-with icmp-host-prohibited
-A FORWARD -j REJECT --reject-with icmp-host-prohibited
COMMIT
[root@node3 ~]# iptables-restore /etc/sysconfig/iptables

 

Keepalived 的详细介绍 ：请点这里
Keepalived 的下载地址 ：请点这里

推荐阅读：

CentOS 6.3 下 Haproxy+Keepalived+Apache 配置笔记 http://www.linuxidc.com/Linux/2013-06/85598.htm

Haproxy + KeepAlived 实现 WEB 群集 on CentOS 6 http://www.linuxidc.com/Linux/2012-03/55672.htm

Keepalived+Haproxy 配置高可用负载均衡 http://www.linuxidc.com/Linux/2012-03/56748.htm

Haproxy+Keepalived 构建高可用负载均衡 http://www.linuxidc.com/Linux/2012-03/55880.htm

启动 Keepalived 服务

启动 Master 和 Slave 机器上的 keepalived 服务 

#service keepalived start

查看 vip 是否生效并看日志信息是否有报错 

[root@node3 keepalived]# ip addr
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue state UNKNOWN 
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
    inet6 ::1/128 scope host 
       valid_lft forever preferred_lft forever
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP qlen 1000
    link/ether 00:23:7d:33:2f:32 brd ff:ff:ff:ff:ff:ff
    inet 192.168.172.3/24 brd 192.168.172.255 scope global eth0
    inet 192.168.172.100/32 scope global eth0
    inet6 fe80::223:7dff:fe33:2f32/64 scope link 
       valid_lft forever preferred_lft forever
3: eth1: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN qlen 1000
    link/ether 00:23:7d:33:2f:4c brd ff:ff:ff:ff:ff:ff

查看系统日志 

[root@node3 keepalived]# tail /var/log/nodes
Nov 11 18:02:32 node3 Keepalived_vrrp: Using LinkWatch kernel netlink reflector...
Nov 11 18:02:32 node3 Keepalived_healthcheckers: Opening file '/etc/keepalived/keepalived.conf'.
Nov 11 18:02:32 node3 Keepalived_healthcheckers: Configuration is using : 7114 Bytes
Nov 11 18:02:32 node3 Keepalived_vrrp: VRRP sockpool: [ifindex(2), proto(112), fd(10,11)]
Nov 11 18:02:32 node3 Keepalived_healthcheckers: Using LinkWatch kernel netlink reflector...
Nov 11 18:02:33 node3 Keepalived_vrrp: VRRP_Instance(mes_Redis) Transition to MASTER STATE
Nov 11 18:02:34 node3 Keepalived_vrrp: VRRP_Instance(mes_Redis) Entering MASTER STATE
Nov 11 18:02:34 node3 Keepalived_vrrp: VRRP_Instance(mes_Redis) setting protocol VIPs.
Nov 11 18:02:34 node3 Keepalived_vrrp: VRRP_Instance(mes_Redis) Sending gratuitous ARPs on eth0 for 192.168.172.100
Nov 11 18:02:44 node3 Keepalived_vrrp: VRRP_Instance(mes_Redis) Sending gratuitous ARPs on eth0 for 192.168.172.100

在网络中其它机器 ping vip

这时 keepalived 就实现了基本功能了，下面来配置 redis 服务。

Redis 实现 Master-Slave 架构，Master 只使用内存存储数据，对外提供读写服务，Slave 仅作为 Standby，进行数据持久化。

配置 Redis Master 并启动，Master 采用内存工作的方式。

[root@node3 etc]# more redis.master.conf 
daemonize yes
port 6379
#bind 10.25.28.18
pidfile /tmp/redis.pid
loglevel verbose 
logfile /data/logs/redis/redis_master.log 
timeout 300
databases 16
vm-enabled no
maxmemory-policy volatile-ttl
slave-serve-stale-data yes
appendonly no
appendfsync everysec
no-appendfsync-on-rewrite yes
hash-max-zipmap-entries 64
hash-max-zipmap-value 512
list-max-ziplist-entries 512
list-max-ziplist-value 64
set-max-intset-entries 512
activerehashing yes
[root@node3 etc]# /usr/local/redis/bin/redis-server /etc/redis/redis.master.conf

配置 Redis Slave 并启动 

[root@node4 etc]# more redis.slave.conf 
daemonize yes
port 6379
#bind 10.25.28.18
slaveof 192.168.172.3 6379
pidfile /tmp/redis.pid
loglevel verbose 
logfile /data/logs/redis/redis_slave.log 
dir /data/redis
timeout 300
databases 16
vm-enabled no
maxmemory-policy volatile-ttl
rdbcompression yes
slave-serve-stale-data yes
appendonly yes
appendfsync everysec
no-appendfsync-on-rewrite yes
hash-max-zipmap-entries 64
hash-max-zipmap-value 512
list-max-ziplist-entries 512
list-max-ziplist-value 64
set-max-intset-entr
[root@node4 etc]# /usr/local/redis/bin/redis-server /etc/redis/redis.slave.conf

进行测试

测试数据同步

在 Master 上插入一些测试数据 

[root@node3 etc]# redis-cli 
redis 127.0.0.1:6379> set foo bar
OK
redis 127.0.0.1:6379> get foo
"bar"

然后在 Slave 上获取测试数据 

[root@node4 etc]# redis-cli 
redis 127.0.0.1:6379> get foo
"bar"

配置状态检查

配置 keepalived 对 Redis 进行状态检查

在 Redis Master 上配置如下，修改 keepalived.conf

[root@node3 redis]# more /etc/keepalived/keepalived.conf
global_defs {
   notification_email {admin@qixing318.com}
   notification_email_from keepalived@node3
   router_id node3
}

vrrp_script chk_redis {
    script "/etc/keepalived/redis/redis_check.sh"
    interval 1 
}

vrrp_instance mes_Redis {
    state MASTER
    interface eth0
    garp_master_delay 10
    smtp_alert
    virtual_router_id 3
    priority 100
    nopreempt
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass 1111
    }
    virtual_ipaddress {192.168.172.100}
    track_script {chk_redis}
    notify_master /etc/keepalived/redis/master.sh
    notify_backup /etc/keepalived/redis/backup.sh
    notify_fault  /etc/keepalived/redis/fault.sh
    notify_stop   /etc/keepalived/redis/stop.sh
}

创建以下脚本：

[root@node3 redis]# pwd
/etc/keepalived/redis
[root@node3 redis]# ll
total 16
-rwxr-xr-x 1 root root 442 Nov 12 10:03 backup.sh
-rwxr-xr-x 1 root root 280 Nov 12 10:01 fault.sh
-rwxr-xr-x 1 root root 564 Nov 12 10:03 master.sh
-rwxr-xr-x 1 root root 149 Nov 12 09:24 redis_check.sh
[root@node3 redis]# more *
::::::::::::::
backup.sh
::::::::::::::
#!/bin/sh

KLOG=/data/logs/redis/keepalived-state.log

RLOG=/data/logs/redis/redis-state.log
echo "[backup]" >> $KLOG
date >> $KLOG
echo "----------" >> $KLOG

echo "[backup]" >> $RLOG
date >> $RLOG
echo "being slave...." >> $RLOG 2>&1
echo "run slaveof cmd ..." >> $RLOG
echo SLAVEOF 192.168.172.4 6379 | /usr/local/redis/bin/redis-cli >>$RLOG  2>&1
echo "run sync cmd ..." >> $RLOG
echo SYNC | /usr/local/redis/bin/redis-cli >> $RLOG 2>&1 
::::::::::::::
fault.sh
::::::::::::::
#!/bin/sh

KLOG=/data/logs/redis/keepalived-state.log
echo "[fault]" >> $KLOG
date >> $KLOG
echo "----------" >> $KLOG

[root@CentOSV1 redis]# more stop.sh 
#!/bin/sh

KLOG=/data/logs/redis/keepalived-state.log
echo "[stop]" >> $KLOG
date >> $KLOG
echo "----------" >> $KLOG
::::::::::::::
master.sh
::::::::::::::
#!/bin/sh

KLOG=/data/logs/redis/keepalived-state.log

RLOG=/data/logs/redis/redis-state.log
echo "[master]" >> $KLOG
date >> $KLOG
echo "----------" >> $KLOG

echo "[master]" >> $RLOG
date >> $RLOG
echo "being master...." >> $RLOG 2>&1
echo "run slaveof cmd ..." >> $RLOG
echo SLAVEOF 192.168.172.4 6379 | /usr/local/redis/bin/redis-cli >>$RLOG  2>&1
sleep 1
echo "run sync cmd ..." >> $RLOG
echo SYNC | /usr/local/redis/bin/redis-cli >>$RLOG 2>&1
sleep 1
echo "run slaveof no one cmd ..." >> $RLOG
echo SLAVEOF NO ONE | /usr/local/redis/bin/redis-cli >>$RLOG 2>&1
::::::::::::::
redis_check.sh
::::::::::::::
#!/bin/sh
alive=`/usr/local/redis/bin/redis-cli PING`
if ["$alive" == "PONG"]; then
  echo "OK"
  exit 0
fi
  #echo "$alive"
  echo "NG"
  exit 1

在 Redis Slave 机器上配置如下：

[root@node4 keepalived]# vi /etc/keepalived/keepalived.conf

global_defs {
   notification_email {admin@qixing318.com}
   notification_email_from keepalived@node4
   router_id node4
}

vrrp_script chk_redis {
    script "/etc/keepalived/redis/redis_check.sh"
    interval 1
}

vrrp_instance mes_Redis {
    state BACKUP
    interface eth0
    garp_master_delay 10
    smtp_alert
    virtual_router_id 3
    priority 90
    nopreempt
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass 1111
    }
    virtual_ipaddress {192.168.172.100}
    track_script {chk_redis}
    notify_master /etc/keepalived/redis/master.sh
    notify_backup /etc/keepalived/redis/backup.sh
    notify_fault  /etc/keepalived/redis/fault.sh
    notify_stop   /etc/keepalived/redis/stop.sh
}

创建脚本如下：

[root@node4 redis]# cd /etc/keepalived/redis/
[root@node4 redis]# ls
backup.sh  fault.sh  master.sh  redis_check.sh
[root@node4 redis]# more *
::::::::::::::
backup.sh
::::::::::::::
#!/bin/sh
KLOG=/data/logs/redis/keepalived-state.log
RLOG=/data/logs/redis/redis-state.log
echo "[backup]" >> $KLOG
date >> $KLOG
echo "----------" >> $KLOG
echo "[backup]" >> $RLOG
date >> $RLOG
sleep 10
echo "being slave...." >> $RLOG 2>&1
echo "run slaveof cmd ..." >> $RLOG
echo SLAVEOF 192.168.172.3 6379 | /usr/local/redis/bin/redis-cli >>$RLOG  2>&1
echo "run sync cmd ..." >> $RLOG
echo SYNC | /usr/local/redis/bin/redis-cli >> $RLOG 2>&1 
::::::::::::::
fault.sh
::::::::::::::
#!/bin/sh
KLOG=/data/logs/redis/keepalived-state.log
echo "[fault]" >> $KLOG
date >> $KLOG
echo "----------" >> $KLOG

[root@CentosV1 redis]# more stop.sh 
#!/bin/sh
KLOG=/data/logs/redis/keepalived-state.log
echo "[stop]" >> $KLOG
date >> $KLOG
echo "----------" >> $KLOG
::::::::::::::
master.sh
::::::::::::::
#!/bin/sh

KLOG=/data/logs/redis/keepalived-state.log

RLOG=/data/logs/redis/redis-state.log
echo "[master]" >> $KLOG
date >> $KLOG
echo "----------" >> $KLOG

echo "[master]" >> $RLOG
date >> $RLOG
echo "being master...." >> $RLOG 2>&1
echo "run slaveof cmd ..." >> $RLOG
echo SLAVEOF 192.168.172.3 6379 | /usr/local/redis/bin/redis-cli >>$RLOG  2>&1
sleep 1
echo "run sync cmd ..." >> $RLOG
echo SYNC | /usr/local/redis/bin/redis-cli >>$RLOG 2>&1
sleep 1
echo "run slaveof no one cmd ..." >> $RLOG
echo SLAVEOF NO ONE | /usr/local/redis/bin/redis-cli >>$RLOG 2>&1
::::::::::::::
redis_check.sh
::::::::::::::
#!/bin/sh
alive=`/usr/local/redis/bin/redis-cli PING`
if ["$alive" == "PONG"]; then
  echo "OK"
  exit 0
fi
  #echo "$alive"
  echo "NG"
  exit 1

Redis 是一个高性能的 kv 数据库，我们用 Redis 构建了一套消息系统。

Redis 架构为 Master Slave 结构，通过 keepalived 进行状态检查，故障迁移，达到高可用的目的。

下载 keepalived

#wget http://www.keepalived.org/software/keepalived-1.2.9.tar.gz

编译安装 keepalived，编译时需要指定内核目录，否则 Use IPVS Framework 为 NO。

#./configure --prefix=/usr/local/keepalived \
--with-kernel-dir=/usr/src/kernels/3.9.10-100.fc17.x86_64 \
--enable-snmp \
--enable-sha1

配置 keepalived 信息如下图：

通过 Keepalived 实现 Redis 双机

编译与安装：

#make && make install

结果报错了，重新指定配置参数，去掉 –with-kernel-dir

配置信息如下图：

通过 Keepalived 实现 Redis 双机

在次编译安装，OK，顺利通过安装成功！

现在可以将 /usr/local/keepalived 下相应文件拷贝到系统对应目录

通过 Keepalived 实现 Redis 双机

配置 keepalived

在配置 keepalived.conf 配置文件之前先备份一份 keepalived.conf 文件后在进行配置 

#cp /etc/keepalived/keepalived.conf /etc/keepalived/keepalived.conf.bak

在 Master 机器配置 keepalived

[root@node3 etc]# more /etc/keepalived/keepalived.conf
global_defs {
   notification_email {admin@qixing318.com}
   notification_email_from keepalived@node3
   router_id node3
}

vrrp_instance mes_Redis {
    state MASTER
    interface eth0
    garp_master_delay 10
    smtp_alert
    virtual_router_id 3
    priority 100
    nopreempt
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass 1111
    }
    virtual_ipaddress {192.168.172.100}
}

在 Slave 机器配置 keepalived

[root@node4 etc]# more /etc/keepalived/keepalived.conf
global_defs {
   notification_email {admin@qixing318.com}
   notification_email_from keepalived@node4
   router_id node4
}

vrrp_instance mes_Redis {
    state BACKUP
    interface eth0
    garp_master_delay 10
    smtp_alert
    virtual_router_id 3
    priority 90
    nopreempt
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass 1111
    }
    virtual_ipaddress {192.168.172.100}
}

配置防火墙

打开 Master 和 Slave 机器上的防火墙 

[root@node3 ~]# more /etc/sysconfig/iptables
# Firewall configuration written by system-config-firewall
# Manual customization of this file is not recommended.
*filter
:INPUT ACCEPT [0:0]
:FORWARD ACCEPT [0:0]
:OUTPUT ACCEPT [0:0]
-A INPUT -m state --state ESTABLISHED,RELATED -j ACCEPT
-A INPUT -p icmp -j ACCEPT
-A INPUT -i lo -j ACCEPT
-A INPUT -m state --state NEW -m tcp -p tcp --dport 22 -j ACCEPT
-A INPUT -m state --state NEW -m tcp -p tcp --dport 9022 -j ACCEPT
-A INPUT -m state --state NEW -m tcp -p tcp --dport 6379 -j ACCEPT
-A INPUT -p 112 -j ACCEPT
-A INPUT -p 255 -j ACCEPT
-A INPUT -j REJECT --reject-with icmp-host-prohibited
-A FORWARD -j REJECT --reject-with icmp-host-prohibited
COMMIT
[root@node3 ~]# iptables-restore /etc/sysconfig/iptables

 

Keepalived 的详细介绍 ：请点这里
Keepalived 的下载地址 ：请点这里

推荐阅读：

CentOS 6.3 下 Haproxy+Keepalived+Apache 配置笔记 http://www.linuxidc.com/Linux/2013-06/85598.htm

Haproxy + KeepAlived 实现 WEB 群集 on CentOS 6 http://www.linuxidc.com/Linux/2012-03/55672.htm

Keepalived+Haproxy 配置高可用负载均衡 http://www.linuxidc.com/Linux/2012-03/56748.htm

Haproxy+Keepalived 构建高可用负载均衡 http://www.linuxidc.com/Linux/2012-03/55880.htm

测试关闭 Redis Master

[root@node3 redis]# redis-cli 
redis 127.0.0.1:6379> get foo
"bar"
redis 127.0.0.1:6379> shutdown

查看 keepalived backup 信息，此时 keepalvied 进入 master 状态 

Nov 12 18:47:25 message4 Keepalived_vrrp: VRRP_Instance(mes_Redis) Transition to MASTER STATE
Nov 12 18:47:26 message4 Keepalived_vrrp: VRRP_Instance(mes_Redis) Entering MASTER STATE
Nov 12 18:47:26 message4 Keepalived_vrrp: VRRP_Instance(mes_Redis) setting protocol VIPs.
Nov 12 18:47:26 message4 Keepalived_vrrp: VRRP_Instance(mes_Redis) Sending gratuitous ARPs on eth0 for 192.168.172.100
Nov 12 18:47:36 message4 Keepalived_vrrp: VRRP_Instance(mes_Redis) Sending gratuitous ARPs on eth0 for 192.168.172.100

查看 Redis slave 信息，此时 redis 进入 master 状态 

[root@node4 redis]# ip addr
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue state UNKNOWN 
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
    inet6 ::1/128 scope host 
       valid_lft forever preferred_lft forever
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP qlen 1000
    link/ether d4:85:64:50:47:48 brd ff:ff:ff:ff:ff:ff
    inet 192.168.172.4/24 brd 192.168.172.255 scope global eth0
    inet 192.168.172.100/32 scope global eth0
    inet6 fe80::d685:64ff:fe50:4748/64 scope link 
       valid_lft forever preferred_lft forever
3: eth1: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN qlen 1000
    link/ether d4:85:64:50:47:4a brd ff:ff:ff:ff:ff:ff
4: eth2: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN qlen 1000
    link/ether d4:85:64:50:47:44 brd ff:ff:ff:ff:ff:ff
5: eth3: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state

测试恢复 Redis Master，此时 keepalived 状态变为 backup，redis 变为 slave

Nov 12 10:56:23 node3 Keepalived_vrrp: VRRP_Script(chk_redis) succeeded
Nov 12 10:56:23 node3 Keepalived_vrrp: VRRP_Instance(mes_Redis) Entering BACKUP STATE
role:slave
master_host:192.168.172.4
master_port:6379
master_link_status:up
master_last_io_seconds_ago:4
master_sync_in_progress:0

数据同步正常

测试关闭 Keepalived backup

[root@node4 redis]# service keepalived stop

查看 keepalived master 信息 

Nov 12 11:15:49 node3 Keepalived_vrrp: VRRP_Instance(mes_Redis) Transition to MASTER STATE
Nov 12 11:15:50 node3 Keepalived_vrrp: VRRP_Instance(mes_Redis) Entering MASTER STATE
Nov 12 11:15:50 node3 Keepalived_vrrp: VRRP_Instance(mes_Redis) setting protocol VIPs.
Nov 12 11:15:50 node3 Keepalived_vrrp: VRRP_Instance(mes_Redis) Sending gratuitous ARPs on eth0 for 192.168.172.7
Nov 12 11:16:00 node3 Keepalived_vrrp: VRRP_Instance(mes_Redis) Sending gratuitous ARPs on eth0 for 192.168.172.7

测试关闭 Redis Slave 重启数据是否丢失 

[root@node4 redis]# /dragon/bin/redis/bin/redis-cli 
redis 127.0.0.1:6379> get foo
"bar"

用 redis-benchmark 测试 Redis 性能 

[root@node2 ~]# redis-benchmark -h 192.168.172.7 -t set -n 1000000 -r 100000000
====== SET ======
  1000000 requests completed in 18.96 seconds
  50 parallel clients
  3 bytes payload
  keep alive: 1

99.28% <= 1 milliseconds
99.85% <= 2 milliseconds
99.96% <= 3 milliseconds
99.97% <= 8 milliseconds
99.97% <= 9 milliseconds
99.98% <= 14 milliseconds
99.98% <= 15 milliseconds
99.99% <= 17 milliseconds
99.99% <= 18 milliseconds
100.00% <= 87 milliseconds
100.00% <= 87 milliseconds
52748.18 requests per second

压力测试一段时间，master 内存数据量达到 15.9G，slave 内存数据量达到 9G，此时 slave 持久化时间消耗很大，查询时延很大。

master 的 info

redis 127.0.0.1:6379> info
redis_version:2.4.14
redis_git_sha1:00000000
redis_git_dirty:0
arch_bits:64
multiplexing_api:epoll
gcc_version:4.4.6
process_id:24839
uptime_in_seconds:8348
uptime_in_days:0
lru_clock:2086437
used_cpu_sys:3687.96
used_cpu_user:1733.84
used_cpu_sys_children:0.01
used_cpu_user_children:0.00
connected_clients:1
connected_slaves:1
client_longest_output_list:495810
client_biggest_input_buf:0
blocked_clients:0
used_memory:17087636096
used_memory_human:15.91G
used_memory_rss:16042332160
used_memory_peak:17119264664
used_memory_peak_human:15.94G
mem_fragmentation_ratio:0.94
mem_allocator:jemalloc-2.2.5
loading:0
aof_enabled:0
changes_since_last_save:300459493
bgsave_in_progress:0
last_save_time:1342063376
bgrewriteaof_in_progress:0
total_connections_received:8839
total_commands_processed:301293452
expired_keys:0
evicted_keys:0
keyspace_hits:425469
keyspace_misses:1
pubsub_channels:0
pubsub_patterns:0
latest_fork_usec:1616
vm_enabled:0
role:master
slave0:192.168.172.4,38292,online
db0:keys=94571396,expires=0

slave 的 info

redis 127.0.0.1:6379> info
redis_version:2.4.14
redis_git_sha1:00000000
redis_git_dirty:0
arch_bits:64
multiplexing_api:epoll
gcc_version:4.4.6
process_id:21820
uptime_in_seconds:6894
uptime_in_days:0
lru_clock:2089328
used_cpu_sys:695.35
used_cpu_user:1090.20
used_cpu_sys_children:24.67
used_cpu_user_children:188.16
connected_clients:5
connected_slaves:0
client_longest_output_list:0
client_biggest_input_buf:4
blocked_clients:0
used_memory:9740498160
used_memory_human:9.07G
used_memory_rss:8099332096
used_memory_peak:9740500440
used_memory_peak_human:9.07G
mem_fragmentation_ratio:0.83
mem_allocator:jemalloc-2.2.5
loading:0
aof_enabled:1
changes_since_last_save:160162637
bgsave_in_progress:0
last_save_time:1342094071
bgrewriteaof_in_progress:1
total_connections_received:6799
total_commands_processed:226063438
expired_keys:0
evicted_keys:0
keyspace_hits:13
keyspace_misses:0
pubsub_channels:0
pubsub_patterns:0
latest_fork_usec:142125
vm_enabled:0
role:slave
aof_current_size:7735856830
aof_base_size:3791924084
aof_pending_rewrite:0
aof_buffer_length:16400
aof_pending_bio_fsync:0
master_host:192.168.172.3
master_port:6379
master_link_status:up
master_last_io_seconds_ago:1
master_sync_in_progress:0
db0:keys=88586614,expires=0
(1.76s)
redis 127.0.0.1:6379> get foo
"bar"
(2.33s)

 参考 http://www.linuxidc.com/Linux/2014-02/96083.htm

正文完
星哥玩云-微信公众号
post-qrcode
Apache Hadoop Keepalived Linux服务器 Nginx Tomcat 分布式 分布式系统 服务器 通过Keepalived实现Redis双机 集群
发表至： 服务器应用
2022-01-20
 0
星锅
版权声明：本站原创文章，由 星锅 于2022-01-20发表，共计17463字。
转载说明：除特殊说明外本站文章皆由CC-4.0协议发布，转载请注明出处。
【腾讯云】推广者专属福利，新客户无门槛领取总价值高达2860元代金券，每种代金券限量500张，先到先得。
Nginx实现反向代理和负载均衡的配置及优化
保护Linux服务器的常用方法
Nginx压力测试工具之WebBench
Ubuntu 16.04上安装Lighttpd+PHP 7(PHP-FPM)+MySQL 5.7
Mongo-connector集成MongoDB到Solr实现增量索引
在Ubuntu 16.04上运行ASP.NET Core
Google Authenticator实现Linux用户登录双认证
Squid 服务启动脚本
阿里云-最新活动爆款每日限量供应
评论（没有评论）
验证码
星哥说事
星哥说事
分享互联网知识
阅读量
3462131
评论数
4
文章搜索
热门文章
基于开源MaxKB构建大语言模型的本地知识库系统

基于开源MaxKB构建大语言模型的本地知识库系统

基于开源 MaxKB 构建大语言模型的本地知识库系统 什么是 MaxKB MaxKB 是一款基于 LLM（La...
获取各大人工智能AI工具通过API和KEY调用的方法

获取各大人工智能AI工具通过API和KEY调用的方法

获取各大人工智能 AI 工具通过 API 和 KEY 调用的方法 前言 随着人工智能技术的飞速发展，AI 工具...
安装开源软件ChatALL（齐叨）来聚合各大人工智能工具

安装开源软件ChatALL（齐叨）来聚合各大人工智能工具

安装开源软件 ChatALL（齐叨）来聚合各大人工智能工具 前言 如果你跟我一样无论遇到什么问题都要询问人工智...
给你的NAS无限可能，安装小晓雅全家桶影音库

给你的NAS无限可能，安装小晓雅全家桶影音库

给你的 NAS 无限可能，安装小晓雅全家桶影音库 家庭存储和娱乐的需求日益增长。无论是高清电影、电视剧，还是音...
vmware下的网卡分配问题

vmware下的网卡分配问题

导读 vmware 虚拟机下 linux centos6.6 只有 lo，没有 eth0 网卡、随机分配 ip...
CDN
阿里云CDN-提高用户访问的响应速度和成功率
随机文章
Linux之rar文件解压之路

Linux之rar文件解压之路

导读 相信大家在使用 Linux 系统时有时候会碰到有关于 rar 文件的解压缩。网上下载 rar 压缩文件的...
入门MySQL数据库导入与导出及重置root密码

入门MySQL数据库导入与导出及重置root密码

如何导入和导出数据库 要导出数据库，打开终端，确保您未登录 MySQL 并键入， mysqldump -u [...
了解ansible架构与工作原理

了解ansible架构与工作原理

Ansible 是一个模型驱动的配置管理器，支持多节点发布、远程任务执行。默认使用 SSH 进行远程连接。无需...
Ubuntu之jdk安装

Ubuntu之jdk安装

导读 JDK 是 Java 语言的软件开发工具包，主要用于移动设备、嵌入式设备上的 java 应用程序。JDK...
Zabbix 3.4 Source code compilation installation

Zabbix 3.4 Source code compilation installation

导读 官介：Zabbix 是一个企业级的分布式开源监控解决方案。它可以监控各种网络应用以及服务器健康性和完整性...
一言一句话
-「」
【腾讯云】云服务器、云数据库、COS、CDN、短信等云产品特惠热卖中