HAProxy

This guide will show how to install the HAProxy and Keepalived on two CentOS servers, using the packages from epel repositories.

This cluster will be assembled of three servers/nodes:

node #1
hostname: ha1
IP: 192.168.70.74

node #2
hostname: ha2
IP: 192.168.70.75

virtual ip
hostname: ha-vip
IP: 192.168.70.76

HAProxy Installation Guide

Prerequisite

On all MySQL server install and configure the mysql check with xinetd:

yum -y install xinetd
echo "mysqlchk 9200/tcp" >> /etc/services
chkconfig xinetd on
sed -i s'+/usr/bin/clustercheck+/usr/bin/clustercheck\n\tserver_args\t= clustercheckuser CLUSTERCHECK_PWD+' /etc/xinetd.d/mysqlchk
sed -i 's+log_type.*+log_type = FILE /var/log/xinetd.log+' /etc/xinetd.conf
sed -i 's+log_on_success.*+log_on_success  =+' /etc/xinetd.conf
service xinetd start

Change CLUSTERCHECK_PWD with your password

On your mysql cluster add the clustercheckuser user:

GRANT PROCESS ON *.* TO 'clustercheckuser'@'localhost' IDENTIFIED BY 'CLUSTERCHECK_PWD';
FLUSH PRIVILEGES;

Install HAProxy

On all two nodes execute:

yum -y install haproxy

Configure HAProxy

On both two nodes execute the following commands (substitute the hostnames percona#.domain and the IPs with those of your cluster):

mv /etc/haproxy/haproxy.cfg /etc/haproxy/haproxy.cfg.orig
cat << EOF >    /etc/haproxy/haproxy.cfg
global
        log 127.0.0.1   local0
        log 127.0.0.1   local1 notice
        maxconn 4096
        uid 188
        gid 188
        daemon
        #debug
        #quiet

defaults
        log     global
        mode    http
        option  tcplog
        option  dontlognull
        retries 3
        option redispatch
        maxconn 2000
        contimeout      5000
        clitimeout      50000
        srvtimeout      50000

listen mysql_cluster_roundrobin 0.0.0.0:5306
    mode    tcp
    balance roundrobin
    option  httpchk

    server percona1.domain 192.168.70.71:3306 check port 9200
    server percona2.domain 192.168.70.72:3306 check port 9200
    server percona3.domain 192.168.70.73:3306 check port 9200

listen mysql_cluster-failover 0.0.0.0:3306
    mode    tcp
    balance leastconn
    option  httpchk
    server  percona1.domain 192.168.70.71:3306 check port 9200
    server  percona2.domain 192.168.70.72:3306 check port 9200 backup
    server  percona3.domain 192.168.70.73:3306 check port 9200 backup

listen admin 0.0.0.0:80
    stats  uri /
EOF

Start HAProxy

Start haproxy service on each server:

service haproxy start
chkconfig haproxy on

Keepalived Installation Guide

Keepalived makes the two HAProxy nodes redundant and create a single affidable access point using an virtual IP. This virutal IP is assigned only to one for time of the two HAProxy server and chenge in case of failure.

Install Keepalived

Install from epel repository:

yum -y install keepalived

Configure Keepalived

Configure keeplalived service:

mv /etc/keepalived/keepalived.conf /etc/keepalived/keepalived.conf.orig
cat << EOF > /etc/keepalived/keepalived.conf
! Configuration File for keepalived

global_defs {
   notification_email {
     grid-operations@lists.cnaf.infn.it
   }
   notification_email_from noreply-keepalived-gridops@cnaf.infn.it
   smtp_server smpt.server
   smtp_connect_timeout 30
   router_id ha1
}

vrrp_script chk_haproxy {
   script "killall -0 haproxy"
   interval 1                     # check every second
   weight 2                       # add 2 points of prio if OK
}

vrrp_instance VI_1 {
   interface eth0
   state MASTER
   smtp_alert
   virtual_router_id 51
   priority 101                   # 100 on master, 101 on slaves
   advert_int 1
      authentication {
         auth_type PASS
         auth_pass 1111
      }
   virtual_ipaddress {
      192.168.70.76
   }

track_script {
   chk_haproxy
}

}
EOF

Configure sysctl on each server:

echo << EOF >> /etc/sysctl.conf
net.ipv4.ip_nonlocal_bind = 1
EOF
sysctl -p

Start Keepalived

On each server start service:

service keepalived start
chkconfig keepalived on

Check the virtual IP

Check the ip configuration managed by keepalived, on each server use the command and see which server is the master with the virtual IP assigned:

ip addr sh eth0

The master server must have output like this:

2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
    link/ether 52:54:00:f6:ed:35 brd ff:ff:ff:ff:ff:ff
    inet 192.168.70.74/24 brd 192.168.122.255 scope global eth0
    inet 192.168.70.76/32 scope global eth0
    inet6 fe80::5054:ff:fef6:ed35/64 scope link
       valid_lft forever preferred_lft forever