一 环境介绍
1.操作系统
CentOS Linux release 7.2.1511 (Core)
2.服务
keepalived+lvs双主高可用负载均衡集群及LAMP应用
keepalived-1.2.13-7.el7.x86_64
ipvsadm-1.27-7.el7.x86_64
httpd-2.4.6-45.el7.centos.x86_64
mariadb-5.5.52-1.el7.x86_64
php-5.4.16-42.el7.x86_64
二 原理及拓扑图
1.vrrp协议
vrrp(Virtual Redundant Routing Protocol)协议：
在现实的网络环境中，两台需要通信的主机大多数情况下并没有直接的物理连接。对于这样的情况，它们之间路由怎样选择？主机如何选定到达目的主机的下一跳路由，这个问题通常的解决方法有两种：
在主机上使用动态路由协议(RIP、OSPF等)
在主机上配置静态路由
很明显，在主机上配置动态路由是非常不切实际的，因为管理、维护成本以及是否支持等诸多问题。配置静态路由就变得十分流行，但路由器(或者说默认网关default gateway)却经常成为单点故障。VRRP的目的就是为了解决静态路由单点故障问题，VRRP通过一竞选(election)协议来动态的将路由任务交给LAN中虚拟路由器中的某台VRRP路由器。
2.keepalived简介
Keepalived 是一个基于VRRP协议来实现的LVS服务高可用方案，可以利用其来避免单点故障。一个LVS服务会有2台服务器运行Keepalived，一台为主服务器（MASTER），一台为备份服务器（BACKUP），但是对外表现为一个虚拟IP，主服务器会发送特定的消息给备份服务器，当备份服务器收不到这个消息的时候，即主服务器宕机的时候，备份服务器就会接管虚拟IP，继续提供服务，从而保证了高可用性。Keepalived是VRRP的完美实现。
3.lvs-dr
Direct Routing，直接路由,通过为请求报文重新封装一个MAC首部进行转发，源MAC是DIP所在的接口的MAC，目标MAC是某挑选出的RS的RIP所在接口的MAC地址；源IP/PORT，以及目标IP/PORT均保持不变。
Director和各RS都得配置使用VIP；
(1) 确保前端路由器将目标IP为VIP的请求报文发往Director；
(a) 在前端网关做静态绑定；
(b) 在RS上使用arptables；
(c) 在RS上修改内核参数以限制arp通告及应答级别；
arp_announce
arp_ignore
(2) RS的RIP可以使用私网地址，也可以是公网地址；RIP与DIP在同一IP网络；RIP的网关不能指向DIP，以确保响应报文不会经由Director；
(3) RS跟Director要在同一个物理网络；
(4) 请求报文要经由Director，但响应不能经由Director，而是由RS直接发往Client；
(5) 不支持端口映射。
4.IP分配
VIP1:172.18.67.66
VIP2:172.18.67.88
DIP1:172.18.67.13
DIP2:172.18.67.14
RIP1:172.18.67.11
RIP2:172.18.67.12
CIP:172.18.67.3
5.拓扑图
三 keepalived配置
1.安装keepalived
[root@inode2 ~]# yum install -y keepalived [root@inode3 ~]# yum install -y keepalived
2.高可用的ipvs双主集群配置
第一个节点：
[root@inode2 ~]# cd /etc/keepalived/
[root@inode2 keepalived]# vim keepalived.conf
! Configuration File for keepalived
global_defs {
notification_email {
root@localhost #接受通知的邮件地址
}
notification_email_from kaadmin@localhost
smtp_server 127.0.0.1
smtp_connect_timeout 30
router_id node2 #路由节点
vrrp_mcast_group4 224.0.67.67 #多播地址，范围224.0.0.0~239.255.255.255
}
vrrp_instance myr1 {
state MASTER
interface eno16777736 #网卡接口
virtual_router_id 167 #虚拟路由ID号，0~255
priority 100 #优先级，MASTER比BACKUP优先级高
advert_int 1
authentication {
auth_type PASS
auth_pass f1bf7fda
}
virtual_ipaddress {
172.18.67.66/16 dev eno16777736 label eno16777736:0
}
track_interface {
eno16777736
}
notify_master "/etc/keepalived/notify.sh master" #调用通知脚本
notify_backup "/etc/keepalived/notify.sh backup"
notify_fault "/etc/keepalived/notify.sh fault"
}
vrrp_instance myr2 {
state BACKUP
interface eno16777736
virtual_router_id 168
priority 98
advert_int 1
authentication {
auth_type PASS
auth_pass f2bf7ade
}
virtual_ipaddress {
172.18.67.88/16 dev eno16777736 label eno16777736:1
}
track_interface {
eno16777736
}
notify_master "/etc/keepalived/notify.sh master"
notify_backup "/etc/keepalived/notify.sh backup"
notify_fault "/etc/keepalived/notify.sh fault"
}
virtual_server 172.18.67.66 80 { #VIP
delay_loop 2
lb_algo wrr #lvs负载均衡调度算法
lb_kind DR #负载均衡类型
protocol TCP #传输协议
sorry_server 127.0.0.1 80 #localhost
real_server 172.18.67.11 80 { #后端RIP
weight 1 #调度权重
HTTP_GET { #http请求方式
url {
path /
status_code 200 #状态码
}
connect_timeout 2 #连接超时
nb_get_retry 3
delay_before_retry 3
}
}
real_server 172.18.67.12 80 {
weight 1
HTTP_GET {
url {
path /
status_code 200
}
connect_timeout 2
nb_get_retry 3
delay_before_retry 3
}
}
}
第二个节点：
[root@inode3 ~]# cd /etc/keepalived/
[root@inode3 keepalived]# vim keepalived.conf
! Configuration File for keepalived
global_defs {
notification_email {
root@localhost
}
notification_email_from kaadmin@localhost
smtp_server 127.0.0.1
smtp_connect_timeout 30
router_id node3
vrrp_mcast_group4 224.0.67.67
}
vrrp_instance myr1 {
state BACKUP
interface eno16777736
virtual_router_id 167
priority 98
advert_int 1
authentication {
auth_type PASS
auth_pass f1bf7fda
}
virtual_ipaddress {
172.18.67.66/16 dev eno16777736 label eno16777736:0
}
track_interface {
eno16777736
}
notify_master "/etc/keepalived/notify.sh master"
notify_backup "/etc/keepalived/notify.sh backup"
notify_fault "/etc/keepalived/notify.sh fault"
}
vrrp_instance myr2 {
state MASTER
interface eno16777736
virtual_router_id 168
priority 100
advert_int 1
authentication {
auth_type PASS
auth_pass f2bf7ade
}
virtual_ipaddress {
172.18.67.88/16 dev eno16777736 label eno16777736:1
}
track_interface {
eno16777736
}
notify_master "/etc/keepalived/notify.sh master"
notify_backup "/etc/keepalived/notify.sh backup"
notify_fault "/etc/keepalived/notify.sh fault"
}
virtual_server 172.18.67.88 80 {
delay_loop 2
lb_algo wrr
lb_kind DR
protocol TCP
sorry_server 127.0.0.1 80
real_server 172.18.67.11 80 {
weight 1
HTTP_GET {
url {
path /
status_code 200
}
connect_timeout 2
nb_get_retry 3
delay_before_retry 3
}
}
real_server 172.18.67.12 80 {
weight 1
HTTP_GET {
url {
path /
status_code 200
}
connect_timeout 2
nb_get_retry 3
delay_before_retry 3
}
}
}
3.邮件通知脚本
当双主高可用集群主备切换时可通过邮件通知管理员，此时在配置文件中可自动调用实现编辑好的脚本
[root@inode2 ~]# vim notify.sh
#!/bin/bash
#
contact='root@localhost'
notify() {
mailsubject="$(hostname) to be $1, vip floating"
mailbody="$(date +'%F %T'): vrrp transition, $(hostname) changed to be $1"
echo "$mailbody" | mail -s "$mailsubject" $contact
}
case $1 in
master)
notify master
;;
backup)
notify backup
;;
fault)
notify fault
;;
*)
echo "Usage: $(basename $0) {master|backup|fault}"
exit 1
;;
esac
节点二同样配置。
四 后端RS服务器的配置
1.配置LAMP环境
[root@inode4 ~]# yum install httpd mariadb-server php -y [root@inode5 ~]# yum install httpd mariadb-server php -y
2.简单编辑测试网页
[root@inode4 ~]# echo "RS1:172.18.67.11" > /var/www/html/index.html [root@inode5 ~]# echo "RS2:172.18.67.12" > /var/www/html/index.html
3.修改RS内核参数
dr模型中，各主机上均需要配置VIP，解决地址冲突的方式有三种：
(1) 在前端网关做静态绑定；
(2) 在各RS使用arptables；
(3) 在各RS修改内核参数，来限制arp响应和通告的级别；
限制响应级别：arp_ignore
0：默认值，表示可使用本地任意接口上配置的任意地址进行响应；
1: 仅在请求的目标IP配置在本地主机的接收到请求报文接口上时，才给予响应；
限制通告级别：arp_announce
0：默认值，把本机上的所有接口的所有信息向每个接口上的网络进行通告；
1：尽量避免向非直接连接网络进行通告；
2：必须避免向非本网络通告。
可通过编辑脚本实现：
[root@inode4 ~]# vim dr.sh #!/bin/bash # case $1 in start) echo 1 > /proc/sys/net/ipv4/conf/all/arp_ignore echo 1 > /proc/sys/net/ipv4/conf/eth0/arp_ignore echo 2 > /proc/sys/net/ipv4/conf/all/arp_announce echo 2 > /proc/sys/net/ipv4/conf/eth0/arp_announce ;; stop) echo 0 > /proc/sys/net/ipv4/conf/all/arp_ignore echo 0 > /proc/sys/net/ipv4/conf/eth0/arp_ignore echo 0 > /proc/sys/net/ipv4/conf/all/arp_announce echo 0 > /proc/sys/net/ipv4/conf/eth0/arp_announce ;; *) echo "Usage $(basename $0) start|stop" exit 1 ;; esac
同理第二个RS需同样配置
4.添加路由
节点一：
[root@inode4 ~]# ifconfig lo:0 172.18.67.66 netmask 255.255.255.255 broadcast 172.18.67.66 up [root@inode4 ~]# ifconfig lo:1 172.18.67.88 netmask 255.255.255.255 broadcast 172.18.67.88 up [root@inode4 ~]# route add -host 172.18.67.66 dev lo:0 [root@inode4 ~]# route add -host 172.18.67.88 dev lo:1
节点二：
[root@inode5 ~]# ifconfig lo:0 172.18.67.88 netmask 255.255.255.255 broadcast 172.18.67.88 up [root@inode5 ~]# ifconfig lo:1 172.18.67.66 netmask 255.255.255.255 broadcast 172.18.67.66 up [root@inode5 ~]# route add -host 172.18.67.88 dev lo:0 [root@inode5 ~]# route add -host 172.18.67.66 dev lo:1
五 测试
1.启动服务
RS:
[root@inode4 ~]# systemctl start httpd [root@inode5 ~]# systemctl start httpd
节点一：
[root@inode2 ~]# systemctl start keepalived.service [root@inode2 ~]# systemctl status -l keepalived.service ● keepalived.service - LVS and VRRP High Availability Monitor Loaded: loaded (/usr/lib/systemd/system/keepalived.service; disabled; vendor preset: disabled) Active: active (running) since Sun 2017-05-14 01:19:27 CST; 17s ago Process: 2120 ExecStart=/usr/sbin/keepalived $KEEPALIVED_OPTIONS (code=exited, status=0/SUCCESS) Main PID: 2121 (keepalived) CGroup: /system.slice/keepalived.service ├─2121 /usr/sbin/keepalived -D ├─2122 /usr/sbin/keepalived -D └─2123 /usr/sbin/keepalived -D May 14 01:19:29 inode2 Keepalived_vrrp[2123]: Opening script file /etc/keepalived/notify.sh May 14 01:19:29 inode2 Keepalived_healthcheckers[2122]: Netlink reflector reports IP 172.18.67.66 added May 14 01:19:31 inode2 Keepalived_vrrp[2123]: VRRP_Instance(myr2) Transition to MASTER STATE May 14 01:19:32 inode2 Keepalived_vrrp[2123]: VRRP_Instance(myr2) Entering MASTER STATE May 14 01:19:32 inode2 Keepalived_vrrp[2123]: VRRP_Instance(myr2) setting protocol VIPs. May 14 01:19:32 inode2 Keepalived_vrrp[2123]: VRRP_Instance(myr2) Sending gratuitous ARPs on eno16777736 for 172.18.67.88 May 14 01:19:32 inode2 Keepalived_vrrp[2123]: Opening script file /etc/keepalived/notify.sh May 14 01:19:32 inode2 Keepalived_healthcheckers[2122]: Netlink reflector reports IP 172.18.67.88 added May 14 01:19:34 inode2 Keepalived_vrrp[2123]: VRRP_Instance(myr1) Sending gratuitous ARPs on eno16777736 for 172.18.67.66 May 14 01:19:37 inode2 Keepalived_vrrp[2123]: VRRP_Instance(myr2) Sending gratuitous ARPs on eno16777736 for 172.18.67.88 [root@inode2 ~]# ip a l 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 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 valid_lft forever preferred_lft forever inet6 ::1/128 scope host valid_lft forever preferred_lft forever 2: eno16777736: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000 link/ether 00:0c:29:8b:08:6f brd ff:ff:ff:ff:ff:ff inet 172.18.67.13/16 brd 172.18.255.255 scope global eno16777736 valid_lft forever preferred_lft forever inet 172.18.67.66/16 scope global secondary eno16777736:0 valid_lft forever preferred_lft forever inet 172.18.67.88/16 scope global secondary eno16777736:1 valid_lft forever preferred_lft forever inet6 fe80::20c:29ff:fe8b:86f/64 scope link tentative dadfailed valid_lft forever preferred_lft forever
节点二：
[root@inode3 ~]# systemctl start keepalived.service [root@inode3 ~]# systemctl status -l keepalived.service ● keepalived.service - LVS and VRRP High Availability Monitor Loaded: loaded (/usr/lib/systemd/system/keepalived.service; disabled; vendor preset: disabled) Active: active (running) since Sun 2017-05-14 01:20:25 CST; 6s ago Process: 2110 ExecStart=/usr/sbin/keepalived $KEEPALIVED_OPTIONS (code=exited, status=0/SUCCESS) Main PID: 2111 (keepalived) CGroup: /system.slice/keepalived.service ├─2111 /usr/sbin/keepalived -D ├─2112 /usr/sbin/keepalived -D └─2113 /usr/sbin/keepalived -D May 14 01:20:25 inode3 Keepalived_vrrp[2113]: VRRP sockpool: [ifindex(2), proto(112), unicast(0), fd(10,11)] May 14 01:20:26 inode3 Keepalived_vrrp[2113]: VRRP_Instance(myr2) Transition to MASTER STATE May 14 01:20:26 inode3 Keepalived_vrrp[2113]: VRRP_Instance(myr2) Received lower prio advert, forcing new election May 14 01:20:26 inode3 Keepalived_vrrp[2113]: VRRP_Instance(myr2) Received lower prio advert, forcing new election May 14 01:20:26 inode3 Keepalived_vrrp[2113]: VRRP_Instance(myr2) Received lower prio advert, forcing new election May 14 01:20:27 inode3 Keepalived_vrrp[2113]: VRRP_Instance(myr2) Entering MASTER STATE May 14 01:20:27 inode3 Keepalived_vrrp[2113]: VRRP_Instance(myr2) setting protocol VIPs. May 14 01:20:27 inode3 Keepalived_vrrp[2113]: VRRP_Instance(myr2) Sending gratuitous ARPs on eno16777736 for 172.18.67.88 May 14 01:20:27 inode3 Keepalived_vrrp[2113]: Opening script file /etc/keepalived/notify.sh May 14 01:20:27 inode3 Keepalived_healthcheckers[2112]: Netlink reflector reports IP 172.18.67.88 added [root@inode3 ~]# ip a l 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 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 valid_lft forever preferred_lft forever inet6 ::1/128 scope host valid_lft forever preferred_lft forever 2: eno16777736: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000 link/ether 00:0c:29:78:24:c3 brd ff:ff:ff:ff:ff:ff inet 172.18.67.14/16 brd 172.18.255.255 scope global eno16777736 valid_lft forever preferred_lft forever inet 172.18.67.88/16 scope global secondary eno16777736:1 valid_lft forever preferred_lft forever inet6 fe80::20c:29ff:fe78:24c3/64 scope link tentative dadfailed valid_lft forever preferred_lft forever
2.访问测试
[root@inode1 ~]# curl http://172.18.67.66 RS2:172.18.67.12 [root@inode1 ~]# curl http://172.18.67.66 RS1:172.18.67.11 [root@inode1 ~]# curl http://172.18.67.66 RS2:172.18.67.12 [root@inode1 ~]# curl http://172.18.67.66 RS1:172.18.67.11 [root@inode1 ~]# curl http://172.18.67.88 RS2:172.18.67.12 [root@inode1 ~]# curl http://172.18.67.88 RS1:172.18.67.11 [root@inode1 ~]# curl http://172.18.67.88 RS2:172.18.67.12 [root@inode1 ~]# curl http://172.18.67.88 RS1:172.18.67.11
高可用负载集群成功搭建完成。
3.模拟一台web服务器宕机
将RS1的httpd服务停掉，并再此访问：
[root@inode4:~]# systemctl stop httpd [root@inode1 ~]# curl http://172.18.67.66 RS2:172.18.67.12 [root@inode1 ~]# curl http://172.18.67.66 RS2:172.18.67.12 [root@inode1 ~]# curl http://172.18.67.88 RS2:172.18.67.12 [root@inode1 ~]# curl http://172.18.67.88 RS2:172.18.67.12
发现照样可以访问服务器，实际环境中两台RS的内容应该是一模一样的，在这里我为了以示区别将两台内容编辑成不同。
4.模拟一台高可用负载集群宕机
将VS1的keepalived服务关闭并测试：
[root@inode2 ~]# systemctl stop keepalived.service [root@inode1 ~]# curl http://172.18.67.88 RS1:172.18.67.11 [root@inode1 ~]# curl http://172.18.67.88 RS2:172.18.67.12 [root@inode1 ~]# curl http://172.18.67.88 RS1:172.18.67.11 [root@inode1 ~]# curl http://172.18.67.88 RS2:172.18.67.12
我们发现即使高可用负载均衡集群中的某一个主机宕机了，我们任然可以通过其中的一个IP访问web服务器，体现出了高可用的实用性，并且在访问中lvs调度器将客户端请求按设置的权重分别向后端的服务器实现调度。