keepalived集群高可用

keepalived是基于vrrp的实现，最初为ipvs提供高可用功能，同时能够检测后端realserver的健康状态，此后又为其他的服务提供高可用功能，它可以调用外部脚本，监控资源状态，从而进行故障转移，比较适用于节点比较少，不使用共享存储等情况下；相对于heartbeat以及corosync和RHCS来说，算是一种轻量级的高可用解决方案。

keepalived包含2个主要组件，一个组件负责VRRP协议的实现和管理，一个负责对资源进行监控；运行VRRP的主机在主机之间发送组播信息，用以通告自身的优先级、以及其他属性，主机的优先级范围从0-255, 0表示不参与选举，255表示最高优先级，优先级高的会成为MASTER，在MASTER上启动VIP和VMAC，所有发送到VIP的数据包，由MASTER进行处理，当MASTER出现故障或资源出现问题，会停止发送心跳信息或降低自身优先级，当BACKUP主机发现自己的优先级高于对方或检测不到对方时，会将自己提升为MASTER，当原先的MASTER恢复时，默认可以把MASTER抢回去，可以配置不抢占。

在CentOS 6.4上有keepalived的rpm包，直接安装就可以

[root@node4 ~]# yum -yinstall keepalived

Installed:

keepalived.x86_64 0:1.2.7-3.el6

Dependency Installed:

lm_sensors-libs.x86_64 0:3.1.1-17.el6 net-snmp-libs.x86_64 1:5.5-44.el6

配置文件在/etc/keepalived/keepalived.conf，服务启动脚本是/etc/init.d/keepalived

[root@node4 keepalived]#cat keepalived.conf

! Configuration File forkeepalived

global_defs {

notification_email {

[email protected]用以定义状态切换时邮件通知对象，可写多行（个）

}

[email protected]；发件人

smtp_server 192.168.200.1 SMTP服务器地址

smtp_connect_timeout 30连接超时时长

router_id LVS_DEVEL VRRP主机标识

vrrp_scriptchk_state_down {

script "[[ -f /etc/keepalived/down ]]"自定义一个检测条件，如果存在此文件

interval 1检测间隔为1s

weight -10 weight 减去10；

}

}

vrrp_instanceVI_1 {定义一个VRRP实例

state MASTER状态为MASTER

interface eth0在eth0上监听VIP

virtual_router_id 10 定义所属的组，同一个组中的主机必须相同；

priority 100优先级

advert_int 1通告间隔（advertising interval）

authentication {认证

auth_type PASS类型

auth_pass 1111密码

}

virtual_ipaddress {

172.16.1.250 定义VIP

}

track_script {

chk_state_down

}

}

virtual_server172.16.1.10080 {

delay_loop 6

lb_algo rr

lb_kind NAT

nat_mask 255.255.255.0

persistence_timeout 50

protocol TCP

real_server 172.16.1.2 80 {

weight 1

HTTP_GET {检测后端realserver的方式为GET

url {

path /

state_code 200 检测报文头部的状态码为200

}

connect_timeout 3检测超时时间为3s

nb_get_retry 3重试的次数为3次

delay_before_retry 3重试的间隔

}

}

}

定义主机发生状态变化时，发送邮件通知

首先定义一个脚本：

#!/bin/bash

#

vip=172.16.1.100

contact='root@localhost'

notify (){

mailsubject="`hostname` became to $1 , $vip floated."

mailbody="`date +"%F %T"`: vrrp status changed.`hostname` became $1"

echo $mailbody | mail -s "$mailsubject" $contact

}

case $1in

master)

notify master ;;

backup)

notify backup;;

fault)

notify fault ;;

*)

echo "Usage: `basename $0` {master|backup|fault}"

exit 1;;

esac

编辑配置文件：

vrrp_instance Instance1 {

state MASTER

interface eth0

virtual_router_id 10

priority 100

advert_int 1

authentication {

auth_type PASS

auth_pass 1119

}

virtual_ipaddress {

172.16.1.100

}

track_script {

chk_state_down

}

notify_master"/etc/keepalived/nofity.sh master"

notify_backup"/etc/keepalived/nofity.sh backup"

notify_fault"/etc/keepalived/nofity.sh fault"

}

[root@node4 keepalived]#cat keepalived.conf

! Configuration File forkeepalived

global_defs {

notification_email {

root@localhost

}

notification_email_from [email protected]

smtp_server 127.0.0.1

smtp_connect_timeout 30

router_id Node4.magedu.com

}

vrrp_scriptchk_state_down {

script "[[ -f /etc/keepalived/down]] && exit 1 || exit 0"

interval 1

weight -10

}

vrrp_instance Instance1 {

state MASTER

interface eth0

virtual_router_id 10

priority 100

advert_int 1

authentication {

auth_type PASS

auth_pass 1119

}

virtual_ipaddress {

172.16.1.100

}

track_script {

chk_state_down

}

notify_master"/etc/keepalived/notify.sh master"

notify_backup"/etc/keepalived/notify.sh backup"

notify_fault"/etc/keepalived/notify.sh fault"

}

virtual_server172.16.1.100 80 {

delay_loop 6

lb_algo rr

lb_kind NAT

nat_mask 255.255.255.0

persistence_timeout 50

protocol TCP

real_server 172.16.1.3 80 {

weight 1

HTTP_GET {

url {

path /

state_code 200

}

connect_timeout 3

nb_get_retry 3

delay_before_retry 3

}

}

real_server 172.16.1.2 80 {

weight 1

HTTP_GET {

url {

path /

status_code 200

}

connect_timeout 3

nb_get_retry 3

delay_before_retry 3

}

}

}

}

如果检测成功就会自动生成ipvs规则

[root@node5 keepalived]#ipvsadm -L -n

IP Virtual Server version1.2.1 (size=4096)

Prot LocalAddress:PortScheduler Flags

-> RemoteAddress:Port Forward Weight ActiveConn InActConn

TCP 172.16.1.100:80 rr persistent 50

-> 172.16.1.2:80 Masq 1 0 0

-> 172.16.1.3:80 Masq 1 0 0

当有一个real server失效以后，会自动修改规则：

[root@node5 keepalived]#ipvsadm -L -n

IP Virtual Server version1.2.1 (size=4096)

Prot LocalAddress:PortScheduler Flags

-> RemoteAddress:Port Forward Weight ActiveConn InActConn

TCP 172.16.1.100:80 rr persistent 50

-> 172.16.1.3:80 Masq 1 0 0

[root@node5 keepalived]#tail /var/log/messages

Sep 25 04:22:20 node5Keepalived_vrrp[16223]: VRRP_Instance(Instance1) Sending gratuitous ARPs oneth0 for 172.16.1.100

Sep 25 04:22:25 node5Keepalived_vrrp[16223]: VRRP_Instance(Instance1) Sending gratuitous ARPs oneth0 for 172.16.1.100

Sep 25 04:22:28 node5Keepalived_vrrp[16223]: VRRP_Instance(Instance1) Received higher prio advert

Sep 25 04:22:28 node5Keepalived_vrrp[16223]: VRRP_Instance(Instance1) Entering BACKUP STATE

Sep 25 04:22:28 node5Keepalived_vrrp[16223]: VRRP_Instance(Instance1) removing protocol VIPs.

Sep 25 04:22:28 node5Keepalived_healthcheckers[16222]: Netlink reflector reports IP 172.16.1.100removed

Sep 25 04:36:11 node5Keepalived_healthcheckers[16222]: Error connecting server [172.16.1.2]:80.

Sep 25 04:36:11 node5Keepalived_healthcheckers[16222]: Removing service [172.16.1.2]:80 from VS[172.16.1.100]:80

Sep 25 04:36:11 node5Keepalived_healthcheckers[16222]: Remote SMTP server [127.0.0.1]:25 connected.

Sep 25 04:36:11 node5Keepalived_healthcheckers[16222]: SMTP alert successfully sent.

# 如何对特定的服务实现高可用：

思路：

让vrrp_script 监控程序的运行状态，当服务停止时，降低优先级，优先级降低引起VRRP的状态切换，切换的同时，系统会发出邮件，管理员收到邮件后及时处理故障，启动服务后，优先级恢复到原来的值，根据实际情况配置preempt；如果不对故障进行处理那么当第二台设备也服务也停止了，整个系统将无法工作；

[root@node4keepalived]# cat keepalived.conf

!Configuration File for keepalived

global_defs {

notification_email {

root@localhost

}

notification_email_from [email protected]

smtp_server 127.0.0.1

smtp_connect_timeout 30

router_id Node4.magedu.com

}

vrrp_script chk_state_down {

script"[[ -f /etc/keepalived/down ]] && exit 1 || exit 0"

interval 1

weight-2

}

vrrp_script chk_httpd {

script"killall -0 httpd"

interval 1

fall 2

rise 1

weight-20

}

vrrp_instanceInstance1 {

state MASTER

interface eth0

virtual_router_id 10

priority 100

advert_int 1

authentication {

auth_type PASS

auth_pass 1119

}

virtual_ipaddress {

172.16.1.100

}

track_script {

chk_state_down

chk_httpd

}

notify_master "/etc/keepalived/notify.sh master"

notify_backup "/etc/keepalived/notify.sh backup"

notify_fault "/etc/keepalived/notify.sh fault"

}

[root@node4keepalived]# cat notify.sh

#!/bin/bash

vip=172.16.1.100

contact='root@localhost'

notify () {

mailsubject="`hostname` became to$1 , $vip floated."

mailbody="`date +"%F%T"`: vrrp status changed. `hostname` became $1"

echo $mailbody | mail -s"$mailsubject" $contact

}

case $1 in

master)

notify master

/etc/init.d/httpd start;;

backup)

notify backup

/etc/init.d/httpd stop;;

fault)

notify fault

/etc/init.d/httpd stop;;

*)

echo "Usage: `basename $0`{master|backup|fault}"

exit 1;;

esac

keepalived双主模型：

提供2个vrrp实例，2个vip；

vrrp_instance Instance1 {

state MASTER

interface eth0

virtual_router_id 10

priority 100

advert_int 1

authentication {

auth_type PASS

auth_pass 1119

}

virtual_ipaddress {

172.16.1.100

}

track_script {

chk_state_down

chk_httpd

}

notify_master"/etc/keepalived/notify.sh master"

notify_backup"/etc/keepalived/notify.sh backup"

notify_fault"/etc/keepalived/notify.sh fault"

}

vrrp_instance Instance2 {

state BACKUP

interface eth0

virtual_router_id 20

priority 95

advert_int 1

authentication {

auth_type PASS

auth_pass 2119

}

virtual_ipaddress {

172.16.1.101

}

track_script {

chk_state_down

chk_httpd

}

notify_master"/etc/keepalived/notify.sh master"

notify_backup"/etc/keepalived/notify.sh backup"

notify_fault"/etc/keepalived/notify.sh fault"

}

重启keepalived服务

# servicekeepalived restart

[root@node4 keepalived]#ip addr show

1: lo: mtu 16436 qdisc noqueue state UNKNOWN

link/loopback 00:00:00:00:00:00 brd00:00:00:00:00:00

inet 127.0.0.1/8 scope host lo

inet6 ::1/128 scope host

valid_lft forever preferred_lft forever

2: eth0: mtu 1500 qdisc pfifo_fast state UP qlen1000

link/ether 00:0c:29:85:22:ac brdff:ff:ff:ff:ff:ff

inet 172.16.1.4/16 brd 172.16.255.255 scopeglobal eth0

inet 172.16.1.100/32 scope global eth0

inet6 fe80::20c:29ff:fe85:22ac/64 scopelink

valid_lft forever preferred_lft forever

当另一台主机发生故障时：

[root@node4 keepalived]#ip addr show

1: lo: mtu 16436 qdisc noqueue state UNKNOWN

link/loopback 00:00:00:00:00:00 brd00:00:00:00:00:00

inet 127.0.0.1/8 scope host lo

inet6 ::1/128 scope host

valid_lft forever preferred_lft forever

2: eth0: mtu 1500 qdisc pfifo_fast state UP qlen1000

link/ether 00:0c:29:85:22:ac brdff:ff:ff:ff:ff:ff

inet 172.16.1.4/16 brd 172.16.255.255 scopeglobal eth0

inet 172.16.1.100/32 scope global eth0

inet 172.16.1.101/32 scope global eth0

inet6 fe80::20c:29ff:fe85:22ac/64 scopelink

valid_lft forever preferred_lft forever