haproxy + keepalived “一键安装包” 配置与测试HOWTO
本文为 “haproxy + keepalived 一键安装包”的 第四篇目的:帮助快速设置haproxy 与 keepalived 配置文档
第一篇 haproxy 安装与配置
第二篇 keepalived 安装与配置
第三篇 haproxy + keepalived 一键安装包(keepalived 为haproxy 做 HA 高可用）
第四篇 haproxy + keepalived “一键安装包”配置与测试HOWTO
一 haproxy.sh 与 keepalived.sh 服务脚本
[*]# ls /usr/local/sbin/*.ah
[*]haproxy.shkeepalived.sh
二 设置自动启动 haproxy.sh 与 keepalived.sh
[*]# cat /etc/rc.local
[*]#!/bin/sh
[*]#
[*]# This script will be executed *after* all the other init scripts.
[*]# You can put your own initialization stuff in here if you don't
[*]# want to do the full Sys V style init stuff.
[*]
[*]touch /var/lock/subsys/local
[*]ulimit -SHn 65535
[*]
[*]/usr/local/sbin/haproxy.sh start
[*]/usr/local/sbin/keepalived.sh start
三 haproxy 相关配置
[*]3.1 haroxy 程序目录
[*]#ls /usr/local/haproxy
[*]dochaproxy.cfghtmllogssbinshare
[*]
[*]3.2 haproxy 配置文档
[*]# file /usr/local/haproxy/haproxy.cfg
[*]/usr/local/haproxy/haproxy.cfg: UTF-8 Unicode text
[*]
[*]3.3 编辑haproxy 配置文档
[*]# vim /usr/local/haproxy/haproxy.cfg
[*]第37 行 定义健康检查html文档，所以你需要在你的web服务上建立check.html文件用于haproxy
[*]健康检查，也可以跟具需要自定义检查文档。
[*]optionhttpchk HEAD /check.html #用来做健康检查html文档
[*]第38 39 行修10.0.1.252/253 替换成你的web 服务器ip地址
[*]server server1 10.0.1.252:80 cookie server1 check inter 2000 rise 3 fall 3 weight 3
[*]server server2 10.0.1.253:80 cookie server2 check inter 2000 rise 3 fall 3 maxconn 120 weight 3
[*]
[*]3.4 启动haproxy 输入如下命令
[*]# haproxy.sh start
[*]
[*]3.5 检查是否有haproxy 进程
[*]# ps aux | grep haproxy | grep -v 'grep'
[*]nobody 42160.00.2 74361108 ? Ss 20:35 0:00 /usr/local/haproxy/sbin/haproxy -f /usr/local/haproxy/haproxy.cfg
[*]
[*]3.6 检查haproxy 是否监听80 端口
[*]# lsof -i :80
[*]COMMANDPID USER FD TYPE DEVICE SIZE NODE NAME
[*]haproxy 4216 nobody 4uIPv4 8229 TCP *:http (LISTEN)
[*]
[*]3.7 简单测试 haproxy 看看否返回 ’200 OK‘
[*]# curl -IL http://10.0.1.254 (换成你的ip )
[*]HTTP/1.1 200 OK
[*]Server: nginx/1.0.5
[*]Date: Wed, 28 Dec 2011 12:42:30 GMT
[*]Content-Type: text/html
[*]Content-Length: 159
[*]Last-Modified: Sat, 03 Dec 2011 11:36:10 GMT
[*]Connection: close
[*]Accept-Ranges: bytes
四 keepalived 相关配置
[*]4.1 keepalived 程序目录
[*]# ls /usr/local/keepalived
[*]binetchaproxy_check.shkeepalived.confsbinshare
[*]
[*]4.2 haproxy_check.sh shell脚本监控程序,用于检查 80端口，来判断haproxy 是否可用
[*]# file /usr/local/keepalived/haproxy_check.sh
[*]/usr/local/keepalived/haproxy_check.sh: Bourne-Again shell script text executable
[*]
[*]4.3 keepalived 配置文档
[*]# file /usr/local/keepalived/keepalived.conf
[*]/usr/local/keepalived/keepalived.conf: ASCII text
[*]
[*]4.4 编辑keepalived 配置文档
[*]# vim /usr/local/keepalived/keepalived.conf
[*]第16 行 interface 默认值 eth1 ,改成你的haproxy 外网卡(公网ip网卡)
[*]interface eth0
[*]
[*]第 34 行 默认值 10.0.1.250/24 dev eth1 ，10.0.1.250 替换成你virtual ip 地址,dev eth1 改成你的haproxy 外网卡(与第16行一致)
[*]10.0.1.250/24 dev eth0 scope global
[*]
[*]4.5 启动keepalived 输入如下命令
[*]# keepalived.sh start
[*]# 4242 #此处返回的是keepalived 进程id，敲下回车键即可
[*]
[*]4.6 检查是否有keepalived 进程
[*]# ps aux | grep keepalived | grep -v 'grep'
[*]root 42380.00.063844 464 pts/0 S 20:52 0:00 /bin/bash /usr/local/sbin/keepalived.sh start
[*]root 42400.00.135776 608 ? Ss 20:52 0:00 /usr/local/keepalived/sbin/keepalived -f /usr/local/keepalived/keepalived.conf
[*]root 42410.00.1399361048 ? S 20:52 0:00 /usr/local/keepalived/sbin/keepalived -f /usr/local/keepalived/keepalived.conf
[*]root 42420.00.2638481152 pts/0 S 20:52 0:00 /bin/bash /usr/local/keepalived/haproxy_check.sh start
[*]
[*]4.7 查看keepalived 输出日志
[*]# tail /var/log/messages
[*]Dec 28 20:52:28 four Keepalived_vrrp: Registering Kernel netlink reflector
[*]Dec 28 20:52:28 four Keepalived_vrrp: Registering Kernel netlink command channel
[*]Dec 28 20:52:28 four Keepalived_vrrp: Registering gratutious ARP shared channel
[*]Dec 28 20:52:28 four Keepalived: Starting VRRP child process, pid=4241
[*]Dec 28 20:53:08 four Keepalived_vrrp: Opening file '/usr/local/keepalived/keepalived.conf'.
[*]Dec 28 20:53:08 four Keepalived_vrrp: Configuration is using : 63577 Bytes
[*]Dec 28 20:53:08 four Keepalived_vrrp: Using LinkWatch kernel netlink reflector...
[*]Dec 28 20:53:08 four Keepalived_vrrp: VRRP_Instance(VI_1) Entering BACKUP STATE
[*]Dec 28 20:53:12 four Keepalived_vrrp: VRRP_Instance(VI_1) Transition to MASTER STATE
[*]Dec 28 20:53:13 four Keepalived_vrrp: VRRP_Instance(VI_1) Entering MASTER STATE #当前keepalived 状态为master
五 haproxy 与 keepalived 整体测试
[*]5.1 查看 haproxy 与 keepalived 进程树
[*]# pstree
[*]init─┬─crond
[*] ├─events/0
[*] ├─events/1
[*] ├─haproxy #haproxy 进程
[*] ├─keepalived───keepalived #keepalived 进程(2个)
[*] ├─keepalived.sh───haproxy_check.s───sleep #keepalived.sh 调用 haproxy_check.sh (监控脚本) 调用 sleep （用于控检查间隔时间,2s）
[*]
[*]
[*]5.2 模拟网络故障 haproxy 切换测试
[*]haproxy 1 日志
[*]# tail /var/log/messages -f
[*]Dec 28 21:09:43 one kernel: NFSD: starting 90-second grace period
[*]Dec 28 21:09:44 one Keepalived: Starting Keepalived v1.2.2 (12/25,2011)
[*]Dec 28 21:09:44 one Keepalived_vrrp: Registering Kernel netlink reflector
[*]Dec 28 21:09:44 one Keepalived_vrrp: Registering Kernel netlink command channel
[*]Dec 28 21:09:44 one Keepalived_vrrp: Registering gratutious ARP shared channel
[*]Dec 28 21:09:44 one Keepalived: Starting VRRP child process, pid=1993
[*]Dec 28 21:10:24 one Keepalived_vrrp: Opening file '/usr/local/keepalived/keepalived.conf'.
[*]Dec 28 21:10:24 one Keepalived_vrrp: Configuration is using : 63597 Bytes
[*]Dec 28 21:10:24 one Keepalived_vrrp: Using LinkWatch kernel netlink reflector...
[*]Dec 28 21:10:24 one Keepalived_vrrp: VRRP_Instance(VI_1) Entering BACKUP STATE
[*]
[*]haproxy 2 日志
[*]# ifdown eth0
[*]# tail /var/log/messages
[*]Dec 28 20:52:28 four Keepalived: Starting VRRP child process, pid=4241
[*]Dec 28 20:53:08 four Keepalived_vrrp: Opening file '/usr/local/keepalived/keepalived.conf'.
[*]Dec 28 20:53:08 four Keepalived_vrrp: Configuration is using : 63577 Bytes
[*]Dec 28 20:53:08 four Keepalived_vrrp: Using LinkWatch kernel netlink reflector...
[*]Dec 28 20:53:08 four Keepalived_vrrp: VRRP_Instance(VI_1) Entering BACKUP STATE
[*]Dec 28 20:53:12 four Keepalived_vrrp: VRRP_Instance(VI_1) Transition to MASTER STATE
[*]Dec 28 20:53:13 four Keepalived_vrrp: VRRP_Instance(VI_1) Entering MASTER STATE
[*]Dec 28 21:12:07 four Keepalived_vrrp: Kernel is reporting: interface eth0 DOWN
[*]Dec 28 21:12:07 four Keepalived_vrrp: VRRP_Instance(VI_1) Entering FAULT STATE
[*]Dec 28 21:12:07 four Keepalived_vrrp: VRRP_Instance(VI_1) Now in FAULT state #keepalived fault 状态
[*]
[*]haproxy 1 日志
[*]# tail /var/log/messages -f
[*]Dec 28 21:09:43 one kernel: NFSD: starting 90-second grace period
[*]Dec 28 21:09:44 one Keepalived: Starting Keepalived v1.2.2 (12/25,2011)
[*]Dec 28 21:09:44 one Keepalived_vrrp: Registering Kernel netlink reflector
[*]Dec 28 21:09:44 one Keepalived_vrrp: Registering Kernel netlink command channel
[*]Dec 28 21:09:44 one Keepalived_vrrp: Registering gratutious ARP shared channel
[*]Dec 28 21:09:44 one Keepalived: Starting VRRP child process, pid=1993
[*]Dec 28 21:10:24 one Keepalived_vrrp: Opening file '/usr/local/keepalived/keepalived.conf'.
[*]Dec 28 21:10:24 one Keepalived_vrrp: Configuration is using : 63597 Bytes
[*]Dec 28 21:10:24 one Keepalived_vrrp: Using LinkWatch kernel netlink reflector...
[*]Dec 28 21:10:24 one Keepalived_vrrp: VRRP_Instance(VI_1) Entering BACKUP STATE
[*]Dec 28 21:12:08 one Keepalived_vrrp: VRRP_Instance(VI_1) Transition to MASTER STATE
[*]Dec 28 21:12:09 one Keepalived_vrrp: VRRP_Instance(VI_1) Entering MASTER STATE #成功切换为 master
[*]
[*]haproxy 1 检查vip
[*]# ip add
[*]1: lo:mtu 16436 qdisc noqueue
[*] 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:mtu 1500 qdisc pfifo_fast qlen 1000
[*] link/ether 08:00:27:74:3c:05 brd ff:ff:ff:ff:ff:ff
[*] inet 10.0.2.251/24 brd 10.0.2.255 scope global eth0
[*] inet6 fe80::a00:27ff:fe74:3c05/64 scope link
[*] valid_lft forever preferred_lft forever
[*]3: eth1:mtu 1500 qdisc pfifo_fast qlen 1000
[*] link/ether 08:00:27:3e:9a:2b brd ff:ff:ff:ff:ff:ff
[*] inet 10.0.1.251/24 brd 10.0.1.255 scope global eth1
[*] inet 10.0.1.250/24 scope global secondary eth1 #virtual ip
[*] inet6 fe80::a00:27ff:fe3e:9a2b/64 scope link
[*] valid_lft forever preferred_lft forever
[*]4: sit0:mtu 1480 qdisc noop
[*] link/sit 0.0.0.0 brd 0.0.0.0
[*]
[*]5.3 模拟haproxy 80端口故障 keepalived 切换测试
[*]haproxy 2 恢复
[*]# ifup eth0
[*]haproxy 2 日志
[*]# tail /var/log/messages -f
[*]Dec 28 20:53:08 four Keepalived_vrrp: Configuration is using : 63577 Bytes
[*]Dec 28 20:53:08 four Keepalived_vrrp: Using LinkWatch kernel netlink reflector...
[*]Dec 28 20:53:08 four Keepalived_vrrp: VRRP_Instance(VI_1) Entering BACKUP STATE
[*]Dec 28 20:53:12 four Keepalived_vrrp: VRRP_Instance(VI_1) Transition to MASTER STATE
[*]Dec 28 20:53:13 four Keepalived_vrrp: VRRP_Instance(VI_1) Entering MASTER STATE
[*]Dec 28 21:12:07 four Keepalived_vrrp: Kernel is reporting: interface eth0 DOWN
[*]Dec 28 21:12:07 four Keepalived_vrrp: VRRP_Instance(VI_1) Entering FAULT STATE
[*]Dec 28 21:12:07 four Keepalived_vrrp: VRRP_Instance(VI_1) Now in FAULT state
[*]Dec 28 21:15:34 four kernel: e1000: eth0 NIC Link is Up 1000 Mbps Full Duplex, Flow Control: RX
[*]Dec 28 21:15:35 four Keepalived_vrrp: VRRP_Instance(VI_1) Entering BACKUP STATE
[*]
[*]
[*]haproxy 1 进程
[*]# pstree
[*]init─┬─crond
[*] ├─dbus-daemon
[*] ├─events/0
[*] ├─haproxy
[*] ├─keepalived───keepalived
[*] ├─keepalived.sh───haproxy_check.s───sleep
[*]
[*]停止 haproxy 进程
[*]# killall haproxy #停止haproxy 进程
[*]# tail /usr/local/keepalived/fault.log #fault log
[*] 2011-12-28_21:17:50 http://10.0.1.251:80 1 fault
[*] 2011-12-28_21:17:51 http://10.0.1.251:80 2 fault
[*] 2011-12-28_21:17:52 http://10.0.1.251:80 3 fault
[*] 2011-12-28_21:17:53 http://10.0.1.251:80 4 fault
[*] 2011-12-28_21:17:54 http://10.0.1.251:80 5 fault
[*] 2011-12-28_21:17:55 http://10.0.1.251:80 6 fault
[*] 2011-12-28_21:17:56 http://10.0.1.251:80 7 fault
[*] 2011-12-28_21:17:57 http://10.0.1.251:80 8 fault
[*]
[*]haproxy 1 日志
[*]# tail /var/log/messages
[*]Dec 28 21:12:09 one Keepalived_vrrp: VRRP_Instance(VI_1) Entering MASTER STATE
[*]Dec 28 21:17:58 one Keepalived: Terminating on signal
[*]Dec 28 21:17:58 one Keepalived: Stopping Keepalived v1.2.2 (12/25,2011) #keepalived 被haproxy_check.sh 监控脚本关闭
[*]Dec 28 21:17:58 one Keepalived_vrrp: Terminating VRRP child process on signal
[*]Dec 28 21:17:58 one kernel: ADDRCONF(NETDEV_UP): eth0: link is not ready
[*]Dec 28 21:17:58 one kernel: e1000: eth0 NIC Link is Up 1000 Mbps Full Duplex, Flow Control: RX
[*]Dec 28 21:17:58 one kernel: ADDRCONF(NETDEV_CHANGE): eth0: link becomes ready
[*]Dec 28 21:18:00 one kernel: ADDRCONF(NETDEV_UP): eth1: link is not ready
[*]Dec 28 21:18:00 one kernel: e1000: eth1 NIC Link is Up 1000 Mbps Full Duplex, Flow Control: RX
[*]Dec 28 21:18:00 one kernel: ADDRCONF(NETDEV_CHANGE): eth1: link becomes ready
[*]
[*]
[*]haproxy 2 日志
[*]# tail /var/log/messages -f
[*]Dec 28 20:53:08 four Keepalived_vrrp: Configuration is using : 63577 Bytes
[*]Dec 28 20:53:08 four Keepalived_vrrp: Using LinkWatch kernel netlink reflector...
[*]Dec 28 20:53:08 four Keepalived_vrrp: VRRP_Instance(VI_1) Entering BACKUP STATE
[*]Dec 28 20:53:12 four Keepalived_vrrp: VRRP_Instance(VI_1) Transition to MASTER STATE
[*]Dec 28 20:53:13 four Keepalived_vrrp: VRRP_Instance(VI_1) Entering MASTER STATE
[*]Dec 28 21:12:07 four Keepalived_vrrp: Kernel is reporting: interface eth0 DOWN
[*]Dec 28 21:12:07 four Keepalived_vrrp: VRRP_Instance(VI_1) Entering FAULT STATE
[*]Dec 28 21:12:07 four Keepalived_vrrp: VRRP_Instance(VI_1) Now in FAULT state
[*]Dec 28 21:15:34 four kernel: e1000: eth0 NIC Link is Up 1000 Mbps Full Duplex, Flow Control: RX
[*]Dec 28 21:15:35 four Keepalived_vrrp: VRRP_Instance(VI_1) Entering BACKUP STATE
[*]Dec 28 21:18:02 four Keepalived_vrrp: VRRP_Instance(VI_1) Transition to MASTER STATE
[*]Dec 28 21:18:03 four Keepalived_vrrp: VRRP_Instance(VI_1) Entering MASTER STATE #keepalived 状态为 master
[*]
[*]# ip add
[*]1: lo:mtu 16436 qdisc noqueue
[*] 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:mtu 1500 qdisc pfifo_fast qlen 1000
[*] link/ether 08:00:27:27:ef:a6 brd ff:ff:ff:ff:ff:ff
[*] inet 10.0.2.254/24 brd 10.0.2.255 scope global eth0
[*] inet6 fe80::a00:27ff:fe27:efa6/64 scope link
[*] valid_lft forever preferred_lft forever
[*]3: eth1:mtu 1500 qdisc pfifo_fast qlen 1000
[*] link/ether 08:00:27:cd:f3:db brd ff:ff:ff:ff:ff:ff
[*] inet 10.0.1.254/24 brd 10.0.1.255 scope global eth1
[*] inet 10.0.1.250/24 scope global secondary eth1 #virtual ip
[*] inet6 fe80::a00:27ff:fecd:f3db/64 scope link
[*] valid_lft forever preferred_lft forever
[*]4: sit0:mtu 1480 qdisc noop
[*] link/sit 0.0.0.0 brd 0.0.0.0
六 haproxy_check.sh 监控脚本
[*]脚本默认监控 eth1 网卡，所以你需要根据自己的需要调整，如果你haproxy virtual ip 在eth0 网卡，那么你需要做如下更改:
[*]6.1 使用"#"号注释掉 第14行 第16行，取消掉 第15行 第17行
[*]#eth1_ip=$(/sbin/ifconfig eth1| sed -n '/inet addr:/ s/inet addr://pg' | awk -F" " '{print $1}')
[*]eth0_ip=$(/sbin/ifconfig eth0| sed -n '/inet addr:/ s/inet addr://pg' | awk -F" " '{print $1}')
[*]#eth1_url="http://${eth1_ip}:80"
[*]eth0_url="http://${eth0_ip}:80"
[*]
[*]6.2 将第35行 $eth1_url,替换成 $eth0_url
[*]if curl -IL -A "haproxy_check" "$eth1_url" > /dev/null 2>&1 ;then
[*]6.3 将第48行 $eth1_url，替换成 $eth0_url
[*]if ! curl -IL -A "haproxy_check" "$eth1_url" > /dev/null 2>&1 ;then
七 其它
[*]5.2 模拟网络故障 haproxy 切换测试 与 5.3 模拟haproxy 80端口故障 keepalived 切换测试
[*]五 haproxy 与 keepalived 整体测试，反过来测试也是没有问题的，不过要保证 haproxy.sh keepalived.sh 这两个服务脚本已经启动了，可以使用pstree 查看
[*]如：# pstree
[*]init─┬─crond
[*] ├─events/0
[*] ├─events/1
[*] ├─haproxy
[*] ├─keepalived───keepalived
[*] ├─keepalived.sh───haproxy_check.s───sleep
页:
[1]