haproxy 的安装

1.基础安装

yum install pcre pcre-devel openssl  openssl-develtar xf  haproxy-1.7.1.tar.gzcd haproxy-1.7.1make TARGET=linux2628 USE_PCRE=1 USE_OPENSSL=1 USE_ZLIB=1  PREFIX=/usr/local/haproxy make install PREFIX=/usr/local/haproxycd examples/cp  haproxy.init   /etc/init.d/haproxy chmod +x  /etc/init.d/haproxy

2.本机IP：192.168.56.15

[root@bogon examples]# cat /etc/haproxy/haproxy.cfg globalmaxconn 100000chroot /usr/local/haproxyuid 99gid 99daemonnbproc 1pidfile /usr/local/haproxy/logs/haproxy.pidlog 127.0.0.1 local0 infodefaultsoption http-keep-alivemaxconn 100000mode httptimeout connect 300000mstimeout client  300000mstimeout server  300000mslisten stats mode http bind 0.0.0.0:9999 stats enable log global stats uri     /haproxy-status stats auth    haadmin:123456listen  web_port   bind 0.0.0.0:80 mode tcp log global[root@bogon examples]# [root@bogon examples]# /etc/init.d/haproxy  restart Restarting haproxy (via systemctl):                        [  OK  ][root@bogon examples]#

http://192.168.56.15:9999/haproxy-status

3.对后端代理

[root@bogon examples]# cat /etc/haproxy/haproxy.cfg globalmaxconn 100000chroot /usr/local/haproxyuid 99gid 99daemonnbproc 1pidfile /usr/local/haproxy/logs/haproxy.pidlog 127.0.0.1 local0 infodefaultsoption http-keep-alivemaxconn 100000mode httptimeout connect 300000mstimeout client  300000mstimeout server  300000mslisten stats mode http bind 0.0.0.0:9999 stats enable log global stats uri     /haproxy-status stats auth    haadmin:123456listen  web_port   bind 0.0.0.0:80 mode tcp  #tcp的方式 log global server web1  192.168.56.12:80  check inter 2000 fall 15 rise 10 server web2  192.168.56.16:80    check inter 2000 fall 15 rise 10[root@bogon examples]#  /etc/init.d/haproxy  restart  #inter是监控检查时间间隔，即每间隔2秒进行一次检查， rise是连续检查10次失败后将服务器从负载删除， fall是连续15次监控检查成功后重新添加至负载，一般fall大于rise几次，

[root@bogon html]# curl 192.168.56.15192.168.56.12 web1[root@bogon html]# curl 192.168.56.15192.168.56.16 web2[root@bogon html]# curl 192.168.56.15192.168.56.12 web1[root@bogon html]# curl 192.168.56.15192.168.56.16 web2[root@bogon html]# curl 192.168.56.15192.168.56.12 web1[root@bogon html]# curl 192.168.56.15192.168.56.16 web2[root@bogon html]# curl 192.168.56.15192.168.56.12 web1[root@bogon html]# curl 192.168.56.15192.168.56.16 web2[root@bogon html]# curl 192.168.56.15192.168.56.12 web1[root@bogon html]# curl 192.168.56.15192.168.56.16 web2[root@bogon html]#

查看负载状况

################################################

haproxy调度算法

HAProxy的算法有如下8种：roundrobin，表示简单的轮询，static-rr，表示根据权重，建议关注；leastconn，表示最少连接者先处理，建议关注；source，表示根据请求源IP，建议关注；uri，表示根据请求的URI；url_param，表示根据请求的URl参数'balanceurl_param' requires an URL parameter namehdr(name)，表示根据HTTP请求头来锁定每一次HTTP请求；rdp-cookie(name)，表示根据据cookie(name)来锁定并哈希每一次TCP请求。

haproxy的摘除节点的 yum安装的

globalmaxconn 100000chroot /var/lib/haproxyuser haproxygroup haproxydaemonnbproc 1 pidfile /usr/local/haproxy/logs/haproxy.pid stats socket /var/lib/haproxy.sock mode 600 level adminlog 127.0.0.1 local10 infodefaultsoption http-keep-alivemaxconn 100000mode httptimeout connect 5000mstimeout client  50000mstimeout server 50000mslisten statsmode httpbind 0.0.0.0:8888stats enablestats uri     /haproxy-status stats auth    haproxy:saltstackfrontend frontend_www_example_combind 192.168.56.21:80mode httpoption httploglog global    default_backend backend_www_example_combackend backend_www_example_comoption forwardfor header X-REAL-IPoption httpchk HEAD / HTTP/1.0balance   roundrobin server web-node1  192.168.56.11:8080 check inter 2000 rise 30 fall 15server web-node2  192.168.56.12:8080 check inter 2000 rise 30 fall 15[root@web-node1 ~]#

这样就能吧haproxy的节点删除或者增加

echo "disable server backend_www_example_com/web-node1" | socat /var/lib/haproxy.sock stdioecho "enable server backend_www_example_com/web-node1" | socat /var/lib/haproxy.sock stdio

haproxy增加日志的功能

mkdir /var/log/haproxychmod a+w /var/log/haproxy/etc/rsyslog.conf打开选项：$ModLoad imudp$UDPServerRun 514# Save haproxy loglocal0.*                       /var/log/haproxy/haproxy.logvim  /etc/sysconfig/rsyslog[root@web-node1 nginx]# cat /etc/sysconfig/rsyslog # Options for rsyslogd# Syslogd options are deprecated since rsyslog v3.# If you want to use them, switch to compatibility mode 2 by "-c 2"  See rsyslogd(8) for more detailsSYSLOGD_OPTIONS="-r -m 0 -c 2"   #修改此处[root@web-node1 nginx]#

haproxy日志里面的localhost 和配置文件有关 这个IP可以修改

log 127.0.0.1 local10 info

3.集群配置

192.168.56.11 nginx 反向代理 keepalived haproxy

192.168.56.12 nginx 反向代理 keepalived haproxy

192.168.56.13 nginx 模拟tomcat

192.168.56.14 nginx 模拟tomcat

keepalived的VIP为192.168.56.21

问题：haproxy 不会绑定不存在的VIP 所以启动不起来 so,这样可以绑定不存在的VIP

echo 1 > /proc/sys/net/ipv4/ip_nonlocal_bind

vi /etc/sysctl.conf sysctl -p

net.ipv4.ip_nonlocal_bind = 1
开启允许绑定非本机的IP

查看keepalived的配置文件192.168.56.11（主）

[root@web-node1 nginx]# cat  /etc/keepalived/keepalived.conf ! Configuration File for keepalivedglobal_defs {   notification_email {     saltstack@example.com   }   notification_email_from keepalived@example.com   smtp_server 127.0.0.1   smtp_connect_timeout 30   router_id haproxy_ha}vrrp_instance haproxy_ha {state MASTERinterface eth0    virtual_router_id 36priority 150    advert_int 1authentication {auth_type PASS        auth_pass 1111    }    virtual_ipaddress {       192.168.56.21    }}[root@web-node1 nginx]#

查看keepalived的配置文件192.168.56.12（备）

[root@web-node2 ~]# cat /etc/keepalived/keepalived.conf ! Configuration File for keepalivedglobal_defs {   notification_email {     saltstack@example.com   }   notification_email_from keepalived@example.com   smtp_server 127.0.0.1   smtp_connect_timeout 30   router_id haproxy_ha}vrrp_instance haproxy_ha {state BACKUPinterface eth0    virtual_router_id 36priority 100    advert_int 1authentication {auth_type PASS        auth_pass 1111    }    virtual_ipaddress {       192.168.56.21    }}[root@web-node2 ~]#

查看192.168.56.11 haproxy的配置文件

[root@web-node2 ~]# cat /etc/haproxy/haproxy.cfgglobalmaxconn 100000chroot /var/lib/haproxyuser haproxygroup haproxydaemonnbproc 1 pidfile /usr/local/haproxy/logs/haproxy.pid stats socket /var/lib/haproxy.sock mode 600 level adminlog 127.0.0.1 local3 infodefaultsoption http-keep-alivemaxconn 100000mode httptimeout connect 5000mstimeout client  50000mstimeout server 50000mslisten statsmode httpbind 0.0.0.0:8888stats enablestats uri     /haproxy-status stats auth    haproxy:saltstackfrontend frontend_www_example_combind 192.168.56.21:80mode tcpoption httploglog global    default_backend backend_www_example_combackend backend_www_example_comoption forwardfor header X-REAL-IPoption httpchk HEAD / HTTP/1.0balance   roundrobin server web-node1  192.168.56.11:8080 check inter 2000 rise 30 fall 15server web-node2  192.168.56.12:8080 check inter 2000 rise 30 fall 15[root@web-node2 ~]#

查看192.168.56.12 haproxy的配置文件

[root@web-node1 nginx]# cat /etc/haproxy/haproxy.cfg globalmaxconn 100000chroot /var/lib/haproxyuser haproxygroup haproxydaemonnbproc 1 pidfile /usr/local/haproxy/logs/haproxy.pid stats socket /var/lib/haproxy.sock mode 600 level admin#log 127.0.0.1 local3 infolog  192.168.56.21 local0 debugdefaultsoption http-keep-alivemaxconn 100000mode httptimeout connect 5000mstimeout client  50000mstimeout server 50000mslisten statsmode httpbind 0.0.0.0:8888stats enablestats uri     /haproxy-status stats auth    haproxy:saltstackfrontend frontend_www_example_combind 192.168.56.21:80mode httpoption httploglog global    default_backend backend_www_example_combackend backend_www_example_comoption forwardfor header X-REAL-IPoption httpchk HEAD / HTTP/1.0balance   roundrobin server web-node1  192.168.56.11:8080 check inter 2000 rise 30 fall 15server web-node2  192.168.56.12:8080 check inter 2000 rise 30 fall 15[root@web-node1 nginx]#

查看192.168.56.11的nginx反向代理的配置文件(最简单的配置 demo)

[root@web-node1 nginx]# cat  /etc/nginx/nginx.confworker_processes  1;events {    worker_connections  1024;}http {    include       mime.types;    default_type  application/octet-stream;    sendfile        on;    keepalive_timeout  65;     upstream server_pools {         server 192.168.56.13:8080  weight=1;         server 192.168.56.14:8080  weight=1;    }     server {        listen       192.168.56.11:8080;        server_name  blog.liuhaixiao.com;        location / {            proxy_pass http://server_pools;            proxy_set_header Host  $host;            proxy_set_header X-Forwarded-For $remote_addr;        }    }} [root@web-node1 nginx]#

查看192.168.56.12的nginx反向代理的配置文件(最简单的配置 demo)

[root@web-node2 ~]# cat /etc/nginx/nginx.confworker_processes  1;events {    worker_connections  1024;}http {    include       mime.types;    default_type  application/octet-stream;    sendfile        on;    keepalive_timeout  65;     upstream server_pools {         server 192.168.56.13:8080  weight=1;         server 192.168.56.14:8080  weight=1;    }     server {        listen       192.168.56.12:8080;        server_name blog.liuhaixiao.com;        location / {            proxy_pass http://server_pools;            proxy_set_header Host  $host;            proxy_set_header X-Forwarded-For $remote_addr;        }    }}[root@web-node2 ~]#

查看192.168.56.13的nginx web容器[root@web-node3 html]# cat /usr/share/nginx/html/index.html node-2查看192.168.56.14的nginx web容器[root@web-node4 html]# cat /usr/share/nginx/html/index.html node-1

检查VIP zai 192.168.56.11上面

[root@web-node1 nginx]# ip addr1: lo:  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 forever2: eth0:  mtu 1500 qdisc pfifo_fast state UP qlen 1000    link/ether 00:0c:29:10:66:98 brd ff:ff:ff:ff:ff:ff    inet 192.168.56.11/24 brd 192.168.56.255 scope global eth0       valid_lft forever preferred_lft forever    inet 192.168.56.21/32 scope global eth0       valid_lft forever preferred_lft forever    inet6 fe80::20c:29ff:fe10:6698/64 scope link        valid_lft forever preferred_lft forever[root@web-node1 nginx]#

现在把本地的电脑blog.liuhaixiao.com 解析到192.168.56.21上面去

####################################################

新添加：192.168.56.15 增加一个nginx 域名

haproxy 监听80端口 nginx 监听的端口千万不能和后端tomcat一样 否则回包回不来

[root@web-node1 ~]# cat  /etc/nginx/nginx.confworker_processes  1;events {    worker_connections  1024;}http {    include       mime.types;    default_type  application/octet-stream;    sendfile        on;    keepalive_timeout  65;     upstream server_pools {         server 192.168.56.13:8080  weight=1;         server 192.168.56.14:8080  weight=1;    }      upstream server_tt {         server 192.168.56.15:8080  weight=1;    }    server {        listen       81;        server_name  blog.liuhaixiao.com;        location / {            proxy_pass http://server_pools;            proxy_set_header Host  $host;            proxy_set_header X-Forwarded-For $remote_addr;        }    }       server {        listen      81;        server_name  blog.sanlang.com;        location / {            proxy_pass http://server_tt;            proxy_set_header Host  $host;            proxy_set_header X-Forwarded-For $remote_addr;        }    }} [root@web-node1 ~]#

3是 另外一台服务器的192.168.56.15 tomcat/nginx 的页面

haproxy具体参数请参考：

http://www.cnblogs.com/zhang789/p/6057402.html

haproxy:多域名参考

基于域名负载均衡的Haproxy配置

http://blog.csdn.net/youyudehexie/article/details/7606504

#####################