The config examples of keepalived, e.g. master-backup, master-master.

• In master-backup mode, if the nginx on master node failed to work, then the backup node will take it over.
• In master-master mode, if the nginx on any node failed to work, then the other node will take it over.

Tested on:

• Ubuntu 14.04
• Keepalived 1.2.7
• Nginx 1.4.6

sudo apt-get install keepalived nginx
sudo cp check_nginx.sh /etc/keepalived/

The IPs:

master ip: 192.168.33.10
backup ip: 192.168.33.11
virtual ip: 192.168.33.50

On master node:

sudo cp master-backup/master/keepalived.conf /etc/keepalived/
sudo service keepalived start

On backup node:

sudo cp master-backup/backup/keepalived.conf /etc/keepalived/
sudo service keepalived start

Check:

1. Check logging messages in /var/log/syslog on each node.
2. Check the virtual ip status with the following command on each node:

ip a

3. Check priority changes on each node:

• on master: sudo tcpdump -vvv -n -i eth1 host 192.168.33.10
• on backup: sudo tcpdump -vvv -n -i eth1 host 192.168.33.11

4. Stop nginx on master and check 1, 2, 3 again:

sudo service nginx stop

5. Then start nginx on master and check 1, 2, 3 again:

sudo service nginx start

The IPs:

master ip: 192.168.33.10
backup ip: 192.168.33.11
virtual ip 1: 192.168.33.51
virtual ip 2: 192.168.33.52

And the other steps are similar to master-backup mode, except that you can stop the nginx on any node and the other node will take it over.

• man keepalived.conf
• Keepalived 1.2.7 Configuration Synopsis
• See comments in master-backup keepalived.conf