keeplived应用
Keeplived简介
keepalived观其名可知，保持存活，在网络里面就是保持在线了，也就是所谓的高可用或热备，用来防止单点故障(单点故障是指一旦某一点出现故障就会导致整个系统架构的不可用)的发生。
Keepalived原理
keepalived也是模块化设计，不同模块复杂不同的功能
core：是keepalived的核心，复杂主进程的启动和维护，全局配置文件的加载解析等
check：负责healthchecker(健康检查)，包括了各种健康检查方式，以及对应的配置的解析包括LVS的配置解析
vrrp：VRRPD子进程，VRRPD子进程就是来实现VRRP协议的
libipfwc：iptables(ipchains)库，配置LVS会用到
libipvs*：配置LVS会用到
注意，keepalived和LVS完全是两码事，只不过他们各负其责相互配合而已
keepalived启动后会有三个进程
父进程：内存管理，子进程管理等等
子进程：VRRP子进程
子进程：healthchecker子进程
有图可知，两个子进程都被系统WatchDog看管，两个子进程各自复杂自己的事，healthchecker子进程复杂检查各自服务器的健康程度，例如HTTP，LVS等等，如果healthchecker子进程检查到MASTER上服务不可用了，就会通知本机上的兄弟VRRP子进程，让他删除通告，并且去掉虚拟IP，转换为BACKUP状态
实验：
主机1 172.16.31.10
主机2 172.16.31.30
1， 下载rpm包至本地，然后安装，此次实验用的是keepalived-1.2.13-1.el6.x86_64.rpm主机1,2都安装
#yum install -y keepalived-1.2.13-1.el6.x86_64.rpm
2， 编辑配置文件在/etc/keepalived/keepalived.conf，一般在配置前最好先把配置文件备份一下
#cp keepalived.conf keepalived.conf.bak
!Configuration File for keepalived //叹号开头的行为注释行
global_defs{ //全局配置
notification_email {
[email protected]
[email protected]
[email protected]
}
notification_email_from [email protected]
smtp_server 192.168.200.1
smtp_connect_timeout 30
router_id LVS_DEVEL
}
vrrp_instanceVI_1 { // 配置实例1
state MASTER //状态为主
interface eth0 //在哪个接口上配置
virtual_router_id 51 // 虚拟router_id 号为51
priority 100 // 优先级
advert_int 1
authentication {
auth_type PASS // 基于密码的认证
auth_pass1111 // 认证密码
}
virtual_ipaddress { // 虚拟地址IP组
172.16.31.22
}
}
3， 配置备份节点etc/keepalived/keepalived.conf
vrrp_instanceVI_1 {
state BACKUP // 状态变为备的，BACKUP
interface eth0
virtual_router_id 101
priority 99 // 优先级比主的小就行
advert_int 1
authentication {
auth_type PASS
auth_pass 123456 // 与主的保持一致
}
virtual_ipaddress {
172.16.31.22
}
}
4， 启动keepalived服务在主机1接点上
# service keepalived start
Starting keepalived: [ OK ]
在主机1上# ip addr show
(1): lo: <LOOPBACK,UP,LOWER_UP> mtu16436 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 pfifo_fast state UP qlen1000
link/ether 00:0c:29:a7:90:c5 brd ff:ff:ff:ff:ff:ff
inet 172.16.31.10/16 brd 172.16.255.255 scope global eth0
inet 172.16.31.22/32 scope global eth0 虚拟IP配置上了
inet6 fe80::20c:29ff:fea7:90c5/64 scope link
valid_lft forever preferred_lft forever
5， 然后再主机1上停止一下服务，查看日志文件
# service keepalived stop
Oct 10 17:00:49 localhost Keepalived[2791]: Stopping Keepalivedv1.2.13 (09/14,2014)
Oct 10 17:00:49 localhost Keepalived_vrrp[2793]:VRRP_Instance(VI_1) removing protocol VIPs.
主机1上# ip addrshow
(1): lo:<LOOPBACK,UP,LOWER_UP> 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:<BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen1000
link/ether 00:0c:29:a7:90:c5 brdff:ff:ff:ff:ff:ff
inet 172.16.31.10/16 brd 172.16.255.255scope global eth0
inet6 fe80::20c:29ff:fea7:90c5/64 scopelink
valid_lftforever preferred_lft forever
此刻172.16.31.22消失了
6， 此刻在启动2节点keepalived服务，并查看日志文件
# service keepalived start
Sep 25 08:15:21 localhost Keepalived_vrrp[3954]: VRRP_Instance(VI_1)Transition to MASTER STATE
Sep 25 08:15:22 localhost Keepalived_vrrp[3954]: VRRP_Instance(VI_1)Entering MASTER STATE
Sep 25 08:15:22 localhost Keepalived_vrrp[3954]: VRRP_Instance(VI_1)setting protocol VIPs.
由于没人和主机2抢，所以主机2就升级为master
# ip addrshow
(1): lo:<LOOPBACK,UP,LOWER_UP> 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:<BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen1000
link/ether 00:0c:29:c8:8f:e1 brdff:ff:ff:ff:ff:ff
inet 172.16.31.30/16 brd 172.16.255.255scope global eth0
inet172.16.31.22/32 scope global eth0 此刻ip在主机2上
inet6 fe80::20c:29ff:fec8:8fe1/64 scopelink
valid_lftforever preferred_lft forever
7， 在2节点上停止服务，在开始一下服务
Oct 10 17:40:59 localhost Keepalived_vrrp[2941]: VRRP_Instance(VI_1)Received lower prio advert, forcing new election 1节点接收到一个比自己优先级低的，然后重新选举，自己变为主
Oct 10 17:41:00 localhost Keepalived_vrrp[2941]: VRRP_Instance(VI_1)Entering MASTER STATE
Oct 10 17:41:00 localhost Keepalived_vrrp[2941]: VRRP_Instance(VI_1)setting protocol VIPs.
Oct 10 17:41:00 localhost Keepalived_healthcheckers[2940]: Netlinkreflector reports IP 172.16.31.22 added
8， 然后开始配置双主模型，在1,2节点上配置，各增加一个实例
vrrp_instanceVI_2 {
state BACKUP
interface eth0
virtual_router_id 200
priority 99
advert_int 1
authentication {
auth_type PASS
auth_pass 123
}
virtual_ipaddress {
172.16.31.222
}
}
节点2上
vrrp_instanceVI_2 {
state MASTER
interface eth0
virtual_router_id 200
priority 100
advert_int 1
authentication {
auth_type PASS
auth_pass 123
}
virtual_ipaddress {
172.16.31.222
}
}
然后节点1上# ip addr show
（1）: lo: <LOOPBACK,UP,LOWER_UP> mtu16436 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 pfifo_fast state UP qlen1000
link/ether 00:0c:29:a7:90:c5 brd ff:ff:ff:ff:ff:ff
inet 172.16.31.10/16 brd 172.16.255.255 scope global eth0
inet 172.16.31.22/32 scope global eth0
inet6 fe80::20c:29ff:fea7:90c5/64 scope link
valid_lft forever preferred_lft forever
节点2上
（2）: eth0:<BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen1000
link/ether 00:0c:29:c8:8f:e1 brd ff:ff:ff:ff:ff:ff
inet 172.16.31.30/16 brd 172.16.255.255 scope global eth0
inet 172.16.31.222/32 scope global eth0
inet6 fe80::20c:29ff:fec8:8fe1/64 scope link
valid_lft forever preferred_lft forever
9， 简单的利用keepalived配置通知nginx
首先在1,2节点上安装nginx，然后再配置文件keepalived.conf配置
vrrp_instanceVI_1 {
state MASTER
interface eth0
virtual_router_id 101
priority 100
advert_int 1
authentication {
auth_type PASS
auth_pass 123456
}
virtual_ipaddress {
172.16.31.22
}
notify_master "/etc/rc.d/init.d/nginxstart" 表示当切换master时，要执行的脚本
notify_backup "/etc/rc.d/init.d/nginxstop"
notify_fault "/etc/rc.d/init.d/nginxstop"
}
2节点上同样的配置，然后1,2节点
#servicekeepalived restart
LISTEN 0 128 *:80 *:*
然后再此基础上扩展，shell脚本来检测
vrrp-script chk_nx{
script "killall -0 nginx"
interval 1 脚本执行间隔
weight -5 脚本结果导致优先级变更，-5优先级-5，+10优先级+10
fall 2
rise 1
}
然后再配置文件最后再加上
track_script{
chk_ nx 这个是调用上面那个脚本
}
2节点上也是这样，然后2节点开启服务
然后开启1节点，这时候由于1节点比2节点的优先级高，就会变为主的
