Build high-availability load balancing-centos6.4+haproxy+keepalive

1. Experimental Topology as follows, this is done with NGINX Port-based virtual host, easy to test later

650) this.width=650; "src=" http://s3.51cto.com/wyfs02/M01/54/6B/wKioL1SBuUmgED6_AAEIceBESQw873.jpg "title=" t1.jpg "alt=" wkiol1sbuumged6_aaeicebesqw873.jpg "/>

2. Preparation before installation

Haproxy Domestic Download You know

keepalived www.keepalive.org

Sync time

#ntpdate pool.ntp.org

3.haproxy installation ( same as Main and standby node installation )

#tar-zxvf haproxy-1.5.9.tar.gz #cd haproxy-1.5.9# uname-alinux localhost.localdomain 2.6.32-358.el6.x86_64 #1 SMP Fri Fe b 00:31:26 UTC x86_64 x86_64 x86_64 gnu/linux#make target=linux26 prefix=/usr/local/haproxy install#cd/usr/local/ haproxy# mkdir conf Logs

4. Create configuration file information as follows

#vim  conf/haproxy.cfg global       log 127.0.0.1 local0 Log output configuration, all logs are recorded on this machine, output      maxconn  5120  //maximum number of connections via local0        chroot   /usr/local/haproxy  //Change Current working directory       uid      100  //the uid     of the owning user  gid      100  //the Gid     daemon of the owning user   //Run as background haproxy        nbproc   2   //Create 2 processes in Deamon mode      pidfile  /usr/local/haproxy/logs/haproxy.pid   //pid file   defaults       log      127.0.0.1  local3       mode    http   //http seven floor, TCP4 layer, health onlyreturn ok     option  httplog  //using httpd log format       option  dontlognull       option  redispatch        retries 3  //Three failures think Service Unavailable  maxconn 3000   //default maximum number of connections      contimeout 5000  //connection time-out       clitimeout 50000  //Client Timeout Time      srvtimeout 50000   //server time-out time      stats enable       stats Url     stats auth admin:admin of the  URI /ADMIN  //statistics page   //User name password for the statistics page      stats realm haproxy \ statistic   //Statistics page Tip Information  stats hide-version//Hide Statistics page version information   listen web_proxy *:80  server web1 192.168.1.128:80 check inter 5000 fall 1 rise 2   server web1  192.168.1.128:8080 check inter 5000 fall 1 rise 2//check inter  5000  is to detect the heartbeat frequency,rise 2  is 2 times correct to think that the service available,fall 1  is 1 times failed to think the service is unavailable, weight represents the weight

5,keepalive installation configuration ( Primary and standby installation, note the configuration file modification )

# yum-y Install OPENSSL-DEVEL#TAR-ZXVF keepalived-1.2.13.tar.gz#cd keepalived-1.2.13#./configure--prefix=/usr/local /keepalived#make && make install #cp/usr/local/keepalived/etc/rc.d/init.d/keepalived/etc/init.d/#ln-S/usr/ Local/keepalived/etc/sysconfig/keepalived/etc/sysconfig/keepalived#chmod +x/etc/init.d/keepalived #mkdir/etc/ Keepalived#ln-sf/usr/local/keepalived/etc/keepalived/keepalived.conf/etc/keepalived/keepalived.conf

6. Modify the configuration file

# cat /etc/keepalived/keepalived.conf ! configuration file for keepalived  global_defs {   notification_email {     [email  protected]   }   notification_email_from [email protected]    smtp_server 127.0.0.1   smtp_connect_timeout 30   router_id  LVS_DEVEL}vrrp_script chk_haproxy {   script  "/root/chk_haproxy.sh"//Detection script    interval 2   weight 2} vrrp_instance VI_1 {     state master //the initial state of the instance, the standby node is backup    interface eth0     virtual_router_id 51    priority 100//High Priority Select Master, at least high backup50     advert_int 1    authentication {         auth_type pass//authentication mode, support PASS and ah        auth_pass 1111// Certified Passwords     }    virtual_ipaddress {         192.168.1.200//vip    }         track_script {        chk_haproxy//Detection Command          }}

7.keepalived detection Script

# cat chk_haproxy.sh #!/bin/basha= ' ps-c haproxy--no-header | Wc-l ' if [$A = 0]; then/etc/init.d/keepalived STOPFI

8. Start keepalived

# service keepalived start error message starting keepalived:/bin/bash:keepalived:command not found[failed] Workaround # ln-s/usr/local /keepalived/sbin/keepalived/usr/sbin/keepalived

9. validation Process

First look at the nginx configuration

650) this.width=650; "src=" http://s3.51cto.com/wyfs02/M02/54/6B/wKioL1SBucbA6ItBAAAzqQE_bzc009.jpg "style=" float: none; "title=" n1.jpg "alt=" Wkiol1sbucba6itbaaazqqe_bzc009.jpg "/>

650) this.width=650; "src=" http://s3.51cto.com/wyfs02/M02/54/6C/wKiom1SBuTfwak8fAAA3WKZaigc720.jpg "style=" float: none; "title=" n2.jpg "alt=" Wkiom1sbutfwak8faaa3wkzaigc720.jpg "/>

Start master and SLAVE separately , and see the Primary and Standby node log as follows:

650) this.width=650; "src=" http://s3.51cto.com/wyfs02/M01/54/6B/wKioL1SBuerDM0_DAAbUFIbrmZY465.jpg "style=" float: none; "title=" k2.jpg "alt=" Wkiol1sbuerdm0_daabufibrmzy465.jpg "/>

650) this.width=650; "src=" http://s3.51cto.com/wyfs02/M01/54/6C/wKiom1SBuVuTVXm7AAWD8QeEkDM914.jpg "style=" float: none; "title=" k3.jpg "alt=" Wkiom1sbuvutvxm7aawd8qeekdm914.jpg "/>

VIP information can be seen on Master

650) this.width=650; "src=" http://s3.51cto.com/wyfs02/M02/54/6B/wKioL1SBugnBBbePAAHTydFGE3I380.jpg "title=" k1.jpg "alt=" wkiol1sbugnbbbepaahtydfge3i380.jpg "/>

after the test Haproxy Whether load balancing is correct

650) this.width=650; "src=" http://s3.51cto.com/wyfs02/M00/54/6B/wKioL1SBuiajdLtwAAAywJz0uGU276.jpg "style=" float: none; "title=" h1.jpg "alt=" Wkiol1sbuiajdltwaaaywjz0ugu276.jpg "/>

650) this.width=650; "src=" http://s3.51cto.com/wyfs02/M00/54/6C/wKiom1SBuZbB8uGkAAA2Hl_b_yQ987.jpg "style=" float: none; "title=" h2.jpg "alt=" Wkiom1sbuzbb8ugkaaa2hl_b_yq987.jpg "/>

Look at the automatic monitoring page and discover that polling is normal!

650) this.width=650; "src=" http://s3.51cto.com/wyfs02/M01/54/6C/wKiom1SBuaSTGL02AAU2LoofBZs748.jpg "title=" h3.jpg "alt=" wkiom1sbuastgl02aau2loofbzs748.jpg "/>

After the test Master haproxy hangs, whether to switch

#killall Haproxy

After looking at the keepalived log information, found that the master has been stopped according to the script, and the VIP began to drift

650) this.width=650; "src=" http://s3.51cto.com/wyfs02/M02/54/6B/wKioL1SBulaDv-RTAAEvuh5mw14842.jpg "style=" float: none; "title=" k4.jpg "alt=" Wkiol1sbuladv-rtaaevuh5mw14842.jpg "/>

Slave can see that it has risen to Master and has received a VIP address

650) this.width=650; "src=" http://s3.51cto.com/wyfs02/M02/54/6C/wKiom1SBuceDAnHdAAHOMu_gli8719.jpg "title=" k5.jpg "Style=" Float:none; "alt=" wkiom1sbucedanhdaahomu_gli8719.jpg "/>

650) this.width=650; "src=" http://s3.51cto.com/wyfs02/M00/54/6B/wKioL1SBulbitXtRAAG-qE5S4ng497.jpg "title=" k6.jpg "Style=" Float:none; "alt=" wkiol1sbulbitxtraag-qe5s4ng497.jpg "/>

Finally in test access, are normal!

650) this.width=650; "src=" http://s3.51cto.com/wyfs02/M00/54/6C/wKiom1SBug-w1e9kAAApyRuh3Gs501.jpg "style=" float: none; "title=" n3.jpg "alt=" Wkiom1sbug-w1e9kaaapyruh3gs501.jpg "/>

650) this.width=650; "src=" http://s3.51cto.com/wyfs02/M01/54/6B/wKioL1SBup_x8xZiAAAwvlfn9YQ627.jpg "style=" float: none; "title=" n4.jpg "alt=" Wkiol1sbup_x8xziaaawvlfn9yq627.jpg "/>

This article is from the "Bamboo Potential" blog, please be sure to keep this source http://peaceweb.blog.51cto.com/3226037/1586797

Build high-availability load balancing-centos6.4+haproxy+keepalive