基于keepalived实现LVS高可用以及Web服务高可用
VRRP介绍
VRRP（Virtual Routing Redundant Protocol）可以通过在多台路由器组合成一个虚拟路由组(一个VRRP组)之间共享一个虚拟IP(VIP)解决静态配置的问题，此时仅需要客户端以VIP作为其默认网关即可，虚拟IP配置在哪个路由设备上取决于VRRP的工作机制，实现网关地址漂移；MASTER与BACKUP设备之间有优先级设定，数字越大优先级越高，通过各节点之间不停的传送通告信息来确认对方是否在线；当MASTER宕机的时候，虚拟IP会漂移到备用路由上，实现前端网关的高可用
VRRP中一主多从模式
http://blog.运维网.com/attachment/201305/192152268.png
VRRP中多主模式
http://blog.运维网.com/attachment/201305/192336537.png
Keepalived介绍
Keepalived起初是特意为LVS设计，用来监控集群服务中各Real Server节点的状态，根据Layer3,4,5交换机制检查每个服务节点的状态；当某个服务节点出现故障，Keepalived检测到后将这个故障节点从服务中剔除；当故障节点重新恢复到正常状态后，Keepalived可根据自身机制自动将服务加进集群服务添加到集群服务列表中；后来Keepalived加入了VRRP功能，使得现在的Keepalived具有服务器运行检测功能与HA Cluster功能
基本知识点就介绍到这里了，下面让我们一起来看一看keepalived功能实现吧
基于keepalived实现LVS高可用
LVS的实现模型为DR，DR的详细相关配置已经在之前博文有介绍，详细内容请点击DR模型实现
IP地址规划
HA1
HA2
Real Server1
Real Server2
FQDN
node1.magedu.com
FQDN
node2.magedu.com
FQDN
test1.magedu.com
FQDN
test2.magedu.com
IP
172.16.51.76
IP
172.16.51.100
IP
172.16.51.77
IP
172.16.51.78
Resal Server配置准备
安装web服务
# yum install httpd –y
测试web服务
Real Server1
http://blog.运维网.com/attachment/201305/131503553.jpg
Real Server2
http://blog.运维网.com/attachment/201305/131525461.jpg
LVSDR模型配置实现
Rea lServer1上VIP配置
# sysctl -w net.ipv4.conf.lo.arp_ignore=1
# sysctl -w net.ipv4.conf.all.arp_ignore=1
# sysctl -w net.ipv4.conf.lo.arp_announce=2
# sysctl -wnet.ipv4.conf.all.arp_announce=2
# ifconfig lo:0 172.16.51.70 broadcast172.16.51.70 netmask 255.255.255.255 up
# route add -host 172.16.51.70 dev lo:0
Rea lServer2上VIP配置与RealServer1相同，这里就不再列出了
在Rea lServer1与Rea lServer2上安装ipvsadm
# yum -y install ipvsadm
HA1与HA2主机相关操作配置
在HA1与HA2上安装keeplived
# yum -y --nogpgcheck localinstallkeepalived-1.2.7-5.el5.i386.rpm
修改配置文件（HA1上操作）
# cd /etc/keepalived/
# vim keepalived.conf
全局配置更改（关于邮件这里未做太多的更改）
global_defs {
notification_email {
acassen@firewall.loc
failover@firewall.loc
sysadmin@firewall.loc
}
notification_email_from root@localhost
smtp_server 127.0.0.1
smtp_connect_timeout 30
router_id LVS_DEVEL
} 虚拟路由组
vrrp_instance VI_1 {
state MASTER
interface eth0
virtual_router_id 51
priority 101
advert_int 1
authentication {
auth_typePASS
auth_pass password
}
virtual_ipaddress {
172.16.51.70
}
} Virtual_server与real_server设置
virtual_server 172.16.51.70 80 {
delay_loop 6
lb_algo wlc
lb_kind DR
nat_mask 255.255.0.0
persistence_timeout 50
protocol TCP
real_server 172.16.51.77 80 {
weight 1
SSL_GET {
url {
path /
status_code 200
}
connect_timeout 3
nb_get_retry 3
delay_before_retry 3
}
}
virtual_server 172.16.51.70 80 {
delay_loop 6
lb_algo rr
lb_kind DR
nat_mask 255.255.0.0
persistence_timeout 50
protocol TCP
real_server 172.16.51.78 80 {
weight 1
url {
path /
status_code 200
}
connect_timeout 3
nb_get_retry 3
delay_before_retry 3
}
} 更改完成后复制到HA2上相同的一份
# scp /etc/keepalived/keepalived.conf node2:/etc/keepalived/
修改node2上keepalived配置文件，修改项如下
state BACKUP
priority 100
启动keepalived
http://blog.运维网.com/attachment/201305/131544487.jpg
查看配置是否生效
http://blog.运维网.com/attachment/201305/131556920.jpg
查看生成的ipvsadm规则信息
http://blog.运维网.com/attachment/201305/131611572.jpg
使用window主机浏览器进行访问测试
http://blog.运维网.com/attachment/201305/131625963.jpg
刷新一下
http://blog.运维网.com/attachment/201305/131640708.jpg
刷新多次后，查看相关匹配信息
http://blog.运维网.com/attachment/201305/131656260.jpg
基于keepalived实现web服务的高可用
这是在完成上述实验后在HA1与HA2主机实现web服务的高可用
http://blog.运维网.com/attachment/201305/210817872.png
关闭node1与node2的keepalive
#service keepalived stop
在node1与node2上安装web服务
#yum -y install httpd
启动web服务后，分别准备web页面
# echo "node1.magedu.com" >/var/www/html/index.html
# echo "node2.magedu.com" >/var/www/html/index.html
访问web进行测试
http://blog.运维网.com/attachment/201305/131718936.jpg
http://blog.运维网.com/attachment/201305/131731985.jpg
配置文件准备与配置，此rpm制作（马哥专利的哦）时已经提供了（rpm包已经上传到附件了）
# cp keepalived.conf.haproxy_examplekeepalived.conf
# vim keepalived.conf
修改内容
定义一下服务脚本名称httpd（整个配置文件就贴出来了）
! Configuration File for keepalived
global_defs {
notification_email {
linuxedu@foxmail.com
mageedu@126.com
}
notification_email_from kanotify@magedu.com
! Configuration File for keepalived
global_defs {
notification_email {
linuxedu@foxmail.com
mageedu@126.com
}
notification_email_from kanotify@magedu.com
smtp_connect_timeout 3
smtp_server 127.0.0.1
router_id LVS_DEVEL
}
vrrp_script chk_httpd {
script "killall -0 httpd"
interval 2
# check every 2 seconds
weight -2
# if failed, decrease 2 of the priority
fall 2
# require 2 failures for failures
rise 1
# require 1 sucesses for ok
}
vrrp_script chk_schedown {
script "[[ -f /etc/keepalived/down ]]&& exit 1 || exit 0"
interval 2
weight -2
}
vrrp_instance VI_1 {
interface eth0
# interface for inside_network, bound byvrrp
state MASTER
# Initial state, MASTER|BACKUP
# As soon as the other machine(s) come up,
# an election will be held and the machine
# with the highest "priority"will become MASTER.
# So the entry here doesn't matter a wholelot.
priority 101
# for electing MASTER, highest prioritywins.
# to be MASTER, make 50 more than other machines.
virtual_router_id 51
# arbitary unique number 0..255
# used to differentiate multiple instancesof vrrpd
# running on the same NIC (and hence samesocket).
garp_master_delay 1
authentication {
auth_typePASS
auth_pass password
}
track_interface {
eth0
}
# optional, monitor these as well.
# go to FAULT state if any of these godown.
virtual_ipaddress {
172.16.51.70/16 dev eth0 label eth0:0
}
#addresses add|del on change to MASTER, toBACKUP.
#With the same entries on other machines,
#the opposite transition will be occuring.
}
track_interface {
eth0
}
# optional, monitor these as well.
# go to FAULT state if any of these godown.
virtual_ipaddress {
172.16.51.70/16 dev eth0 label eth0:0
}
#addresses add|del on change to MASTER, toBACKUP.
#With the same entries on other machines,
#the opposite transition will be occuring.
#/ brd devscopelabel
track_script {
chk_httpd
chk_schedown
}
notify_master"/etc/keepalived/notify.sh master"
notify_backup"/etc/keepalived/notify.sh backup"
notify_fault"/etc/keepalived/notify.sh fault"
}
#vrrp_instance VI_2 {
# interface eth0
# state MASTER# BACKUP for slaverouters
# priority 101# 100 for BACKUP
# virtual_router_id 52
# garp_master_delay 1
#
# authentication {
# auth_typePASS
# auth_pass password
# }
# track_interface {
# eth0
# }
# virtual_ipaddress {
# 172.16.100.2/16 dev eth0 label eth0:1
# }
# track_script {
# chk_haproxy
# chk_mantaince_down
# }
#
# notify_master "/etc/keepalived/notify.sh master eth0:1"
# notify_backup "/etc/keepalived/notify.sh backup eth0:1"
# notify_fault "/etc/keepalived/notify.sh fault eth0:1"
#
}
复制配置文件到node2主机
# scp /etc/keepalived/keepalived.confnode2:/etc/keepalived/
复制后需要修改内容如下
state BACKUP
priority 100
启动keepalived服务
http://blog.运维网.com/attachment/201305/131753651.jpg
查看MASTER（node1）上主机定义的VIP（eth0别名IP）是否成功
http://blog.运维网.com/attachment/201305/131810185.jpg
访问虚拟IP地址
http://blog.运维网.com/attachment/201305/131835579.jpg
主备转换测试
# cd /etc/keepalived/
# touch downhttp://blog.运维网.com/attachment/201305/131853891.jpg
在备用节点node2上查看http://blog.运维网.com/attachment/201305/131909256.jpg
访问虚拟IP
http://blog.运维网.com/attachment/201305/131928297.jpg
现在把down文件删除，MASTER节点会夺回资源，因为在node1上的优先级（101）比nod2上的优先级（100）高
双主模式实现
两个主机web服务同时开启，配置的VIP不同，让他们互为主从
http://blog.运维网.com/attachment/201305/211831984.png
修改keepalived.conf配置文件（只需修改配置文件末尾“vrrp_instance VI_2”中的内容）
vrrp_instance VI_2 {
interface eth0
state BACKUP# BACKUP forslave routers
priority 100# 100 for BACKUP
virtual_router_id 52
garp_master_delay 1
authentication {
auth_typePASS
auth_pass password
}
track_interface {
eth0
}
virtual_ipaddress {
172.16.51.80/16 dev eth0 label eth0:1
}
track_script {
chk_httpd
chk_schedown
}
notify_master "/etc/keepalived/notify.sh master eth0:1"
notify_backup "/etc/keepalived/notify.sh backup eth0:1"
notify_fault "/etc/keepalived/notify.sh fault eth0:1"
} 在node2上启用“vrrp_instance VI_2”中的内容后，修改一下第二VIP地址与脚本名称（track_httpd, chk_schedown）
修改完成后，启动keepalived服务，查看node1与node2上的VIP地址配置
node1主机
http://blog.运维网.com/attachment/201305/131954453.jpg
node2主机
http://blog.运维网.com/attachment/201305/132010439.jpg
测试
http://blog.运维网.com/attachment/201305/132028394.jpg
http://blog.运维网.com/attachment/201305/132046473.jpg
模拟node1出现故障
# cd /etc/keepalived/
# touch down
查看node2上虚拟IP地址http://blog.运维网.com/attachment/201305/132115618.jpg
此时，访问172.16.51.70与172.16.51.80都是由node2主机返回结果http://blog.运维网.com/attachment/201305/132136796.jpghttp://blog.运维网.com/attachment/201305/132150768.jpg
基于keepalived实现的LVS高可用与web服务的高可用就完成了，希望可以为你带来一些帮助哦
附件：http://down.运维网.com/data/2362850
页:
[1]