keepalived vrrp script|interface weight when positive,nagtive,zero & vrrp's status transition

本文以实例的形式来展示一下脚本的weight对vrrp的影响(状态或优先级).

注意优先级的变化可能导致vrrp master重新选举, 从而导致状态的变化.

而直接导致状态的变化则是当script weight=0时, 可以直接导致状态的变化.(当然最后还是选举产生的, 除非backup配置了不降级, 那么就不会主动让位.)

首先来看一张图 : 

这张图表明了track interface或track script的weight在大于0, 小于0, 等于0这三种状态下, 对VRRP的影响.

(track interface的原理和本文将要讲的track script是一样的, 只是检测的是接口)

weight>0 : 检查成功, 优先级加, 检查失败, 回归基本优先级.

weight=0 : 检查成功, 状态不变, 检查失败, 状态变为FAULT, 再次检查成功, 状态回到初始状态, 可能会需要重新选举master. (能不能选为master就看是否符合vrrp的协定了, 当然还与backup是否配置允许降级有关) 

weight<0 : 检查成功, 优先级不变, 检查失败, 需要降低优先级, 再次检查成功, 回归正常优先级.

[注意]

初始优先级是255的话, 不受weight的动态优先级调整影响. 原因见 : 

http://blog.163.com/digoal@126/blog/static/16387704020147212529175/

测试环境 : 

192.168.173.203

192.168.173.204

虚拟IP, 因为只是测试, 我用了一个其他网段的IP.

172.16.173.100

keepalived的安装参考

http://blog.163.com/digoal@126/blog/static/16387704020147199133921/

vrrp script, 创建3个脚本, 分别用于weight配置为正, 负, 零 : 

脚本返回值0表示成功, 非零表示失败, 我们通过改变文件内容来改变脚本的返回值.

例如pos.sh返回的是/root/pos.var的值. 默认我们都配置为0. 

在测试keepalived的过程中只要修改pos.var, 就可以改变pos.sh脚本的返回值. 

vi /root/pos.sh
#!/bin/bash
res=$(cat /root/pos.var)
exit $res

vi /root/pos.var
0

chmod 500 /root/pos.sh

vi /root/nag.sh
#!/bin/bash
res=$(cat /root/nag.var)
exit $res

vi /root/nag.var
0

chmod 500 /root/nag.sh

vi /root/zero.sh
#!/bin/bash
res=$(cat /root/zero.var)
exit $res

vi /root/zero.var
0

chmod 500 /root/zero.sh

配置文件内容 : 

2个节点唯一不同的配置是优先级priority . 一高一低即可.

192.168.173.203: 100

192.168.173.204: 99 

[root@192_168_173_204 ~]# cd /opt/keepalived/etc/keepalived/
[root@192_168_173_204 keepalived]# cat keepalived.conf
! Configuration File for keepalived  !号或#号开头为注释

global_defs {      # 全局配置除了router_id, 其他都无所谓, 因为我这里的测试只是观察script weight对vrrp的影响.
   notification_email {
     acassen@firewall.loc
     failover@firewall.loc
     sysadmin@firewall.loc
   }
   notification_email_from Alexandre.Cassen@firewall.loc
   smtp_server 192.168.200.1
   smtp_connect_timeout 30
   router_id DIGOAL_TEST   # 两个节点一致
}

vrrp_script pos {   # 脚本配置
    script "/root/pos.sh"
    interval 1
    weight 0   # 配置默认weight 0, 后面可以在instance中覆盖
    fall 1    # 从OK到KO需要1次检测失败.
    rise 1    # 从KO到OK需要1次检测成功.
}

vrrp_script nag {   # 脚本配置
    script "/root/nag.sh"
    interval 1
    weight 0   # 配置默认weight 0, 后面可以在instance中覆盖
    fall 1    # 从OK到KO需要1次检测失败.
    rise 1    # 从KO到OK需要1次检测成功.
}

vrrp_script zero {   # 脚本配置
    script "/root/zero.sh"
    interval 1
    weight 0   # 配置默认weight 0, 后面可以在instance中覆盖
    fall 1    # 从OK到KO需要1次检测失败.
    rise 1    # 从KO到OK需要1次检测成功.
}

vrrp_instance vi_1 {
    state MASTER    # 初始状态
    interface eth0
    virtual_router_id 51
    priority 99   # 初始优先级, 注意两个节点配置要不一样. 一高一低, 高的选举为master.
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass 1111
    }
    track_script {
        pos weight 3
        zero weight 0
        nag weight -5
    }
    unicast_peer {   # 使用多播代替组播发送vrrp心跳.
        192.168.173.203
        192.168.173.204
    }
    virtual_ipaddress {   # 虚拟地址配置, 和ifconfig兼容
        172.16.173.100/24 brd 172.16.173.255 dev eth0 scope link label eth0:1
    }
    debug   # 打开debug, 方便调试, 本例没有用到
}

启动keepalived, -D打开详细日志.

我们先启动优先级低的节点, 这样可以观察到选举master的过程. 

[root@192_168_173_204 ~]# keepalived -f /opt/keepalived/etc/keepalived/keepalived.conf -D

查看日志 

/var/log/messages
Aug 20 16:32:43 192_168_173_204 Keepalived[5920]: Starting Keepalived v1.2.13 (08/19,2014)
Aug 20 16:32:43 192_168_173_204 Keepalived[5921]: Starting Healthcheck child process, pid=5922
Aug 20 16:32:43 192_168_173_204 Keepalived[5921]: Starting VRRP child process, pid=5923
Aug 20 16:32:43 192_168_173_204 Keepalived_vrrp[5923]: Netlink reflector reports IP 192.168.173.204 added
Aug 20 16:32:43 192_168_173_204 Keepalived_vrrp[5923]: Netlink reflector reports IP fe80::f6ce:46ff:fe85:15fc added
Aug 20 16:32:43 192_168_173_204 Keepalived_vrrp[5923]: Registering Kernel netlink reflector
Aug 20 16:32:43 192_168_173_204 Keepalived_healthcheckers[5922]: Netlink reflector reports IP 192.168.173.204 added
Aug 20 16:32:43 192_168_173_204 Keepalived_vrrp[5923]: Registering Kernel netlink command channel
Aug 20 16:32:43 192_168_173_204 Keepalived_vrrp[5923]: Registering gratuitous ARP shared channel
Aug 20 16:32:43 192_168_173_204 Keepalived_healthcheckers[5922]: Netlink reflector reports IP fe80::f6ce:46ff:fe85:15fc added
Aug 20 16:32:43 192_168_173_204 Keepalived_healthcheckers[5922]: Registering Kernel netlink reflector
Aug 20 16:32:43 192_168_173_204 Keepalived_healthcheckers[5922]: Registering Kernel netlink command channel
Aug 20 16:32:43 192_168_173_204 Keepalived_healthcheckers[5922]: Opening file '/opt/keepalived/etc/keepalived/keepalived.conf'.
Aug 20 16:32:43 192_168_173_204 Keepalived_vrrp[5923]: Opening file '/opt/keepalived/etc/keepalived/keepalived.conf'.
Aug 20 16:32:43 192_168_173_204 Keepalived_healthcheckers[5922]: Configuration is using : 7935 Bytes
Aug 20 16:32:43 192_168_173_204 Keepalived_vrrp[5923]: Configuration is using : 69309 Bytes
Aug 20 16:32:43 192_168_173_204 Keepalived_vrrp[5923]: Using LinkWatch kernel netlink reflector...
Aug 20 16:32:43 192_168_173_204 Keepalived_vrrp[5923]: VRRP sockpool: [ifindex(2), proto(112), unicast(1), fd(10,11)]
Aug 20 16:32:43 192_168_173_204 Keepalived_healthcheckers[5922]: Using LinkWatch kernel netlink reflector...
Aug 20 16:32:43 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Script(pos) succeeded
Aug 20 16:32:43 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Script(zero) succeeded
Aug 20 16:32:43 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Script(nag) succeeded
Aug 20 16:32:44 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Instance(vi_1) Transition to MASTER STATE
Aug 20 16:32:45 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Instance(vi_1) Entering MASTER STATE
Aug 20 16:32:45 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Instance(vi_1) setting protocol VIPs.
Aug 20 16:32:45 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Instance(vi_1) Sending gratuitous ARPs on eth0 for 172.16.173.100
Aug 20 16:32:45 192_168_173_204 Keepalived_healthcheckers[5922]: Netlink reflector reports IP 172.16.173.100 added
Aug 20 16:32:50 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Instance(vi_1) Sending gratuitous ARPs on eth0 for 172.16.173.100

通过ifconfig可以看到虚拟IP已经起来了 

eth0:1    Link encap:Ethernet  HWaddr F4:CE:46:85:15:FC
          inet addr:172.16.173.100  Bcast:172.16.173.255  Mask:255.255.255.0
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1

简单的说一下日志, 可以看出keepalived进程结构分成3个部分, 主进程以及两个fork出来的子进程.

vrrp复制路由主节点的选举, vrrp心跳.

healthcheckers负责检测, 例如vrrp script和vrrp interface(本例没有配置vrrp interface check).

虚拟ip起来后, 发送了免费的ARP请求, 宣告IP和MAC地址.

现在, 192.168.173.204的优先级是99+3=102

好了, 现在启动192.168.173.203的keepalived. 

[root@192_168_173_203 ~]# keepalived -f /opt/keepalived/etc/keepalived/keepalived.conf -D

查看日志 

Aug 20 16:38:28 192_168_173_203 Keepalived[23107]: Starting Keepalived v1.2.13 (08/19,2014)
Aug 20 16:38:28 192_168_173_203 Keepalived[23108]: Starting Healthcheck child process, pid=23109
Aug 20 16:38:28 192_168_173_203 Keepalived[23108]: Starting VRRP child process, pid=23110
Aug 20 16:38:28 192_168_173_203 Keepalived_healthcheckers[23109]: Netlink reflector reports IP 192.168.173.203 added
Aug 20 16:38:28 192_168_173_203 Keepalived_vrrp[23110]: Netlink reflector reports IP 192.168.173.203 added
Aug 20 16:38:28 192_168_173_203 Keepalived_healthcheckers[23109]: Netlink reflector reports IP 192.168.173.156 added
Aug 20 16:38:28 192_168_173_203 Keepalived_vrrp[23110]: Netlink reflector reports IP 192.168.173.156 added
Aug 20 16:38:28 192_168_173_203 Keepalived_healthcheckers[23109]: Netlink reflector reports IP fe80::f6ce:46ff:fe86:4234 added
Aug 20 16:38:28 192_168_173_203 Keepalived_vrrp[23110]: Netlink reflector reports IP fe80::f6ce:46ff:fe86:4234 added
Aug 20 16:38:28 192_168_173_203 Keepalived_healthcheckers[23109]: Registering Kernel netlink reflector
Aug 20 16:38:28 192_168_173_203 Keepalived_vrrp[23110]: Registering Kernel netlink reflector
Aug 20 16:38:28 192_168_173_203 Keepalived_healthcheckers[23109]: Registering Kernel netlink command channel
Aug 20 16:38:28 192_168_173_203 Keepalived_vrrp[23110]: Registering Kernel netlink command channel
Aug 20 16:38:28 192_168_173_203 Keepalived_vrrp[23110]: Registering gratuitous ARP shared channel
Aug 20 16:38:28 192_168_173_203 Keepalived_healthcheckers[23109]: Opening file '/opt/keepalived/etc/keepalived/keepalived.conf'.
Aug 20 16:38:28 192_168_173_203 Keepalived_vrrp[23110]: Opening file '/opt/keepalived/etc/keepalived/keepalived.conf'.
Aug 20 16:38:28 192_168_173_203 Keepalived_healthcheckers[23109]: Configuration is using : 7937 Bytes
Aug 20 16:38:28 192_168_173_203 Keepalived_vrrp[23110]: Configuration is using : 69311 Bytes
Aug 20 16:38:28 192_168_173_203 Keepalived_vrrp[23110]: Using LinkWatch kernel netlink reflector...
Aug 20 16:38:28 192_168_173_203 Keepalived_vrrp[23110]: VRRP sockpool: [ifindex(2), proto(112), unicast(1), fd(10,11)]
Aug 20 16:38:28 192_168_173_203 Keepalived_healthcheckers[23109]: Using LinkWatch kernel netlink reflector...
Aug 20 16:38:28 192_168_173_203 Keepalived_vrrp[23110]: VRRP_Script(nag) succeeded
Aug 20 16:38:28 192_168_173_203 Keepalived_vrrp[23110]: VRRP_Script(zero) succeeded
Aug 20 16:38:28 192_168_173_203 Keepalived_vrrp[23110]: VRRP_Script(pos) succeeded

初始设置为MASTER, 所以有选举的过程, 但是刚起来的时候脚本还没开始检查, 优先级是100, 所以选举中变成了backup角色. 

Aug 20 16:38:29 192_168_173_203 Keepalived_vrrp[23110]: VRRP_Instance(vi_1) Transition to MASTER STATE
Aug 20 16:38:29 192_168_173_203 Keepalived_vrrp[23110]: VRRP_Instance(vi_1) Received higher prio advert
Aug 20 16:38:29 192_168_173_203 Keepalived_vrrp[23110]: VRRP_Instance(vi_1) Entering BACKUP STATE

当开始检查脚本后, 优先级提高到了100+3=103, 大于192.168.173.204的102, 所以选举中变成了MASTER.

同时虚拟IP切换到了本节点 

Aug 20 16:38:30 192_168_173_203 Keepalived_vrrp[23110]: VRRP_Instance(vi_1) forcing a new MASTER election
Aug 20 16:38:30 192_168_173_203 Keepalived_vrrp[23110]: VRRP_Instance(vi_1) forcing a new MASTER election
Aug 20 16:38:31 192_168_173_203 Keepalived_vrrp[23110]: VRRP_Instance(vi_1) Transition to MASTER STATE
Aug 20 16:38:32 192_168_173_203 Keepalived_vrrp[23110]: VRRP_Instance(vi_1) Entering MASTER STATE
Aug 20 16:38:32 192_168_173_203 Keepalived_vrrp[23110]: VRRP_Instance(vi_1) setting protocol VIPs.
Aug 20 16:38:32 192_168_173_203 Keepalived_vrrp[23110]: VRRP_Instance(vi_1) Sending gratuitous ARPs on eth0 for 172.16.173.100
Aug 20 16:38:32 192_168_173_203 Keepalived_healthcheckers[23109]: Netlink reflector reports IP 172.16.173.100 added
Aug 20 16:38:37 192_168_173_203 Keepalived_vrrp[23110]: VRRP_Instance(vi_1) Sending gratuitous ARPs on eth0 for 172.16.173.100

查看192.168.173.204节点的日志, 选举为BACKUP后, 移除了虚拟IP. 

Aug 20 16:38:08 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Instance(vi_1) Transition to MASTER STATE
Aug 20 16:38:09 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Instance(vi_1) Entering MASTER STATE
Aug 20 16:38:09 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Instance(vi_1) setting protocol VIPs.
Aug 20 16:38:09 192_168_173_204 Keepalived_healthcheckers[5922]: Netlink reflector reports IP 172.16.173.100 added
Aug 20 16:38:09 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Instance(vi_1) Sending gratuitous ARPs on eth0 for 172.16.173.100
Aug 20 16:38:14 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Instance(vi_1) Sending gratuitous ARPs on eth0 for 172.16.173.100
Aug 20 16:38:29 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Instance(vi_1) Received lower prio advert, forcing new election
Aug 20 16:38:29 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Instance(vi_1) Sending gratuitous ARPs on eth0 for 172.16.173.100
Aug 20 16:38:29 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Instance(vi_1) Received lower prio advert, forcing new election
Aug 20 16:38:29 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Instance(vi_1) Sending gratuitous ARPs on eth0 for 172.16.173.100
Aug 20 16:38:30 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Instance(vi_1) Received higher prio advert
Aug 20 16:38:30 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Instance(vi_1) Entering BACKUP STATE
Aug 20 16:38:30 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Instance(vi_1) removing protocol VIPs.
Aug 20 16:38:30 192_168_173_204 Keepalived_healthcheckers[5922]: Netlink reflector reports IP 172.16.173.100 removed

现在我们把192.168.173.203的pos.var改成1, 那么将降回初始优先级100. 比192.168.173.204的优先级102要低. 

[root@192_168_173_203 ~]# echo 1 > /root/pos.var

我们在192.168.173.204节点可以看到它重新选举变成了MASTER 

Aug 20 16:44:43 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Instance(vi_1) forcing a new MASTER election
Aug 20 16:44:43 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Instance(vi_1) forcing a new MASTER election
Aug 20 16:44:44 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Instance(vi_1) Transition to MASTER STATE
Aug 20 16:44:45 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Instance(vi_1) Entering MASTER STATE
Aug 20 16:44:45 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Instance(vi_1) setting protocol VIPs.
Aug 20 16:44:45 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Instance(vi_1) Sending gratuitous ARPs on eth0 for 172.16.173.100
Aug 20 16:44:45 192_168_173_204 Keepalived_healthcheckers[5922]: Netlink reflector reports IP 172.16.173.100 added
Aug 20 16:44:50 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Instance(vi_1) Sending gratuitous ARPs on eth0 for 172.16.173.100

接下来在192.168.173.204测试一下nag.sh脚本, 把nag.var改成1. 那么nag.sh将返回1, 也就是失败. 优先级需要减5. 变成102-5=97, vrrp重新选举的话, 变成BACKUP. 

[root@192_168_173_204 ~]# echo 1 > /root/nag.var

我们在192.168.173.203上可以看到它选举变成了master. 

Aug 20 16:48:04 192_168_173_203 Keepalived_vrrp[23110]: VRRP_Instance(vi_1) forcing a new MASTER election
Aug 20 16:48:04 192_168_173_203 Keepalived_vrrp[23110]: VRRP_Instance(vi_1) forcing a new MASTER election
Aug 20 16:48:05 192_168_173_203 Keepalived_vrrp[23110]: VRRP_Instance(vi_1) Transition to MASTER STATE
Aug 20 16:48:06 192_168_173_203 Keepalived_vrrp[23110]: VRRP_Instance(vi_1) Entering MASTER STATE
Aug 20 16:48:06 192_168_173_203 Keepalived_vrrp[23110]: VRRP_Instance(vi_1) setting protocol VIPs.
Aug 20 16:48:06 192_168_173_203 Keepalived_vrrp[23110]: VRRP_Instance(vi_1) Sending gratuitous ARPs on eth0 for 172.16.173.100
Aug 20 16:48:06 192_168_173_203 Keepalived_healthcheckers[23109]: Netlink reflector reports IP 172.16.173.100 added
Aug 20 16:48:11 192_168_173_203 Keepalived_vrrp[23110]: VRRP_Instance(vi_1) Sending gratuitous ARPs on eth0 for 172.16.173.100

现在的优先级构成如下 :  

192.168.173.203  MASTER
base=100
pos.sh failed  +0
zero.sh success
nag.sh success  -0
最终优先级: 100

192.168.173.204  BACKUP
base=99
pos.sh success  + 3
zero.sh success
nag.sh failed  -5
最终优先级: 97

接下来把192.168.173.203的nag.sh弄成failed, 那么将再次减5, 变成95. 选举将变成BACKUP. 

[root@192_168_173_203 ~]# echo 1 > /root/nag.var

可以看到192.168.173.204变成MASTER了. 

Aug 20 16:52:29 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Instance(vi_1) forcing a new MASTER election
Aug 20 16:52:29 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Instance(vi_1) forcing a new MASTER election
Aug 20 16:52:30 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Instance(vi_1) Transition to MASTER STATE
Aug 20 16:52:31 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Instance(vi_1) Entering MASTER STATE
Aug 20 16:52:31 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Instance(vi_1) setting protocol VIPs.
Aug 20 16:52:31 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Instance(vi_1) Sending gratuitous ARPs on eth0 for 172.16.173.100
Aug 20 16:52:31 192_168_173_204 Keepalived_healthcheckers[5922]: Netlink reflector reports IP 172.16.173.100 added

最后测试一下zero脚本对状态的影响, 前面我们看到的nag.sh和pos.sh都只改变优先级, 不会直接对状态造成影响, 而是在选举的过程中简介的对状态造成的影响.

zero脚本的weight=0, 所以脚本调用的成功与否直接影响vrrp instance的状态

测试 : 

把192.168.173.204的zero.var改成1, 

[root@192_168_173_204 ~]# echo 1 >/root/zero.var

查看日志 :  

Aug 20 16:54:49 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Script(zero) failed
Aug 20 16:54:50 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Instance(vi_1) Entering FAULT STATE
Aug 20 16:54:50 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Instance(vi_1) removing protocol VIPs.
Aug 20 16:54:50 192_168_173_204 Keepalived_healthcheckers[5922]: Netlink reflector reports IP 172.16.173.100 removed
Aug 20 16:54:50 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Instance(vi_1) Now in FAULT state

192.168.173.203变成了唯一的状态正常的主机, 所以理所当然就是MASTER了. 

Aug 20 16:54:50 192_168_173_203 Keepalived_vrrp[23110]: VRRP_Instance(vi_1) Transition to MASTER STATE
Aug 20 16:54:51 192_168_173_203 Keepalived_vrrp[23110]: VRRP_Instance(vi_1) Entering MASTER STATE
Aug 20 16:54:51 192_168_173_203 Keepalived_vrrp[23110]: VRRP_Instance(vi_1) setting protocol VIPs.
Aug 20 16:54:51 192_168_173_203 Keepalived_vrrp[23110]: VRRP_Instance(vi_1) Sending gratuitous ARPs on eth0 for 172.16.173.100
Aug 20 16:54:51 192_168_173_203 Keepalived_healthcheckers[23109]: Netlink reflector reports IP 172.16.173.100 added
Aug 20 16:54:56 192_168_173_203 Keepalived_vrrp[23110]: VRRP_Instance(vi_1) Sending gratuitous ARPs on eth0 for 172.16.173.100

最后, 把zero.var改回0, 观察状态回到配置的初始状态. 并重新选举.

[root@192_168_173_204 ~]# echo 0 >/root/zero.var

重新选举的过程中依然是先使用初始优先级进行选举, 状态定了以后, 再检查脚本, 动态的改变优先级.

本例192.168.173.204的初始优先级是99, 大于当前192.168.173.203的优先级95, 所以初始优先级就可以当选MASTER.

等检测开始后, 优先级变成99+3-5=97, 依然大于当前192.168.173.203的优先级95, 所以继续MASTER. 

Aug 20 17:02:05 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Script(zero) succeeded
Aug 20 17:02:06 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Instance(vi_1) prio is higher than received advert
Aug 20 17:02:06 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Instance(vi_1) Transition to MASTER STATE
Aug 20 17:02:06 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Instance(vi_1) Received lower prio advert, forcing new election
Aug 20 17:02:07 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Instance(vi_1) Entering MASTER STATE
Aug 20 17:02:07 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Instance(vi_1) setting protocol VIPs.
Aug 20 17:02:07 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Instance(vi_1) Sending gratuitous ARPs on eth0 for 172.16.173.100
Aug 20 17:02:07 192_168_173_204 Keepalived_healthcheckers[5922]: Netlink reflector reports IP 172.16.173.100 added
Aug 20 17:02:12 192_168_173_204 Keepalived_vrrp[5923]: VRRP_Instance(vi_1) Sending gratuitous ARPs on eth0 for 172.16.173.100

[参考]

1. http://blog.163.com/digoal@126/blog/static/16387704020147199133921/

2. http://blog.163.com/digoal@126/blog/static/163877040201472010332545/

3. http://blog.163.com/digoal@126/blog/static/163877040201471992259474/

时间： 2024-12-01 18:11:41

keepalived vrrp script|interface weight when positive,nagtive,zero & vrrp's status transition的相关文章

keepalived instance priority don&#039;t impact by track script|interface weight when 255

前面我在写keepalived优先级动态调整时, 写到关于初始优先级设置为255的instance不被动态调整的内容, 原因参考本文末尾部分. 因为 if (vrrp->base_priority == VRRP_PRIO_OWNER) { /* we will not run a PRIO_OWNER into a non-PRIO_OWNER */ vrrp->effective_priority = VRRP_PRIO_OWNER; 本文主要就是实际验证一下. 用的例子参考 http:/

keepalived track script introduce

前面简单的介绍了一下keepalived track script | interface weight & instance priority & election的关系,  http://blog.163.com/digoal@126/blog/static/163877040201471992259474/ 本文将重点介绍一下keepalived跟踪脚本的配置用法以及参数的解释 :  keepalived VRRP配置中有一个配置项, vrrp_script :  截取自 : kee

keepalived which time exec track script , notify script when vrrp transition

keepalived什么时候调用track script, track interface, notify_* script, notify; 搞清楚这些对设计HA很有必要. 以下这张图大概的画出了跟踪脚本, 跟踪接口, 状态转换, 以及notify脚本的先后关系 :  当然实际的程序设计和这个没有关系, keepalived是3个进程组成的, 主进程, vrrp进程和health_check进程. 这里不作体现.   keepalived.conf配置项 :  跟踪脚本或跟踪接口配置 : tr

keepalived modify notify.c &amp; vrrp.c to enforce waiting notify script execute success

我们在设计数据库HA的时候, 为了防止主备同时读写一份数据, 或者主备同时对外提供服务. 备机在切换成主机前, 需要有一个fence操作, 需先将主节点fence掉, 然后再激活备库为主库. 例如我们这边用的一个HA脚本 :  https://raw.githubusercontent.com/digoal/sky_postgresql_cluster/master/INSTALL.txt 或者RHEL的HA套件, 都有类似的FENCE操作. 但是keepalived目前没有切换前的自定义脚本,

2004 version keepalived.conf introduction

注意由于1.5.3的man keepalived.conf还是用的2004版, 所以少了很多新功能的介绍, 后面再补充. 最新的配置的请参考源代码里面的doc下面的keepalived.conf.SYNOPSIS. NOTE_vrrp_vmac.txt. 截取man keepalived.conf, 简单的描述一下配置文件的使用. keepalived.conf分成3个组成部分 : GLOBAL CONFIGURATION contains subblocks of Global definit

Keepalived双主模型中vrrp_script中的权重改变故障排查

故障重现 keepalived配置如下 # vi /etc/keepalived/keepalived.conf ! Configuration File for keepalived global_defs { notification_email { root@localhost } notification_email_from admin@lnmmp.com smtp_connect_timeout 3 smtp_server 127.0.0.1 router_id LVS_DEVEL

haproxy+keepalived实现高可用负载均衡(理论篇)_Linux

HAProxy相比LVS的使用要简单很多,功能方面也很丰富.当 前,HAProxy支持两种主要的代理模式:"tcp"也即4层(大多用于邮件服务器.内部协议通信服务器等),和7层(HTTP).在4层模式 下,HAProxy仅在客户端和服务器之间转发双向流量.7层模式下,HAProxy会分析协议,并且能通过允许.拒绝.交换.增加.修改或者删除请求 (request)或者回应(response)里指定内容来控制协议,这种操作要基于特定规则. 我现在用HAProxy主要在于它有以下优点,这里我

Keepalived+HAProxy实现MySQL高可用负载均衡的配置_Mysql

 Keepalived 由于在生产环境使用了mysqlcluster,需要实现高可用负载均衡,这里提供了keepalived+haproxy来实现.       keepalived主要功能是实现真实机器的故障隔离及负载均衡器间的失败切换.可在第3,4,5层交换.它通过VRRPv2(Virtual Router Redundancy Protocol) stack实现的.       Layer3:Keepalived会定期向服务器群中的服务器.发送一个ICMP的数据包(既我们平时用的Ping程

keepalived实现服务高可用

第1章 keepalived服务说明 1.1 keepalived是什么? Keepalived软件起初是专为LVS负载均衡软件设计的,用来管理并监控LVS集群系统中各个服务节点的状态,后来又加入了可以实现高可用的VRRP功能.因此,Keepalived除了能够管理LVS软件外,还可以作为其他服务(例如:Nginx.Haproxy.MySQL等)的高可用解决方案软件. Keepalived软件主要是通过VRRP协议实现高可用功能的.VRRP是Virtual Router RedundancyPr