Learn, Execute, Innovate

Setup High Availability Nginx using Keepalived

Written by: Bagus Facsi Aginsa
Published at: 28 Sep 2021

In most enterprises, it is a common use case to have an active-passive server setup for extra redundancy. If the active server goes down, the passive server can replace it and the system will have almost no downtime. This is possible with VRRP (Virtual Router Redundancy Protocol). The Software that we will use to make this possible is Keepalived.

For this tutorial, we will use the Nginx as Load Balancer to be our use case. Together, Nginx and Keepalived will make a High Availability Load Balancer. You can still follow this tutorial if you have a different use case because our focus is the keepalived configuration, not the load balancer.

Now, let’s begin to set up high availability Nginx using keepalived.

Check Prerequisite

  1. 2 Ubuntu Server, my setup is using ubuntu server. We will install the Nginx on these servers.
  2. Virtual IP, basically it is unused IP on the subnet, you can contact your network/cloud provider if it is available. Note that the Virtual IP address is must be on the subnet with the ubuntu server.
  3. Sudo privileges, make sure you have root access in your VM. The firewall must be set by the root user. So, before we start, make sure you execute sudo su.

Use Case

For this tutorial, we will use this use case

                                                 ____________
                                                |            |
                        ___________      -----> |    App1    |
                       |           |    |       |____________|
                   --> |    LB M   |----|        ____________
                  |    |___________|    |       |            |
     user --------                      |-----> |    App2    |
                  |     ___________     |       |____________|
                  |    |           |    |        ____________
                   --> |    LB B   |----|       |            |
                       |___________|     -----> |    App3    |
                                                |____________|

LB M is Master Load Balancer, while LB B is Backup Load Balancer. The LB M will serve the traffic from the user, but if the LB Mis down, the LB B will immediately replace the LB M using the virtual IP address. This will be the spec of each node:

  1. LB M:
    1. ip: 10.205.199.174/24
    2. interface: eth0
  2. LB B:
    1. ip: 10.205.199.175/24
    2. interface: eth0
  3. Virtual IP: 10.205.199.200

Install Nginx

We will install the Nginx as a load balancer on both Ubuntu servers. First, install Nginx from the package manager

apt install nginx

And the configure the nginx to be a load balancer. You can follow my tutorial on How to Configure Nginx as Load Balancer.

Note that you have to install the Nginx on 2 servers, LB M and LB B. Also, the configuration should be identical for both the Nginx at this point.

Install Keepalived

Keepalived is already available in the Ubuntu package manager. Install the keepalived on both servers by running this command:

apt install keepalived

Keepalived configuration

After installing the keepalived, create a keepalived configuration file on both load balancer

touch /etc/keepalived/keepalived.conf

Master Load Balancer Configuration

Edit the keepalived configuration file on the LB M server.

nano /etc/keepalived/keepalived.conf

Copy & paste this script to the Master Load Balancer:

vrrp_script check_nginx {
    script "killall -0 nginx"
    interval 2
}

vrrp_instance VI_1 {
    interface eth0
    state MASTER
    priority 200

    virtual_router_id 33
    virtual_ipaddress {
        10.205.199.200
    }

    authentication {
        auth_type PASS
        auth_pass password
    }

    track_script {
        check_nginx
    }

}

Edit the password to your own password. This password is used for other keepalived to join the vrrp group.

The check_nginx script will enable keepalived to check the Nginx every 2 seconds, if the Nginx is unavailable, keepalived will send a message to the Backup Load Balancer.

After that, start the keepalived service

service keepalived start

Backup Load Balancer Configuration

Edit the keepalived configuration file on the LB B server.

nano /etc/keepalived/keepalived.conf

Copy & paste this script :

vrrp_script check_nginx {
    script "killall -0 nginx"
    interval 2
}

vrrp_instance VI_1 {
    interface eth0
    state BACKUP
    priority 100

    virtual_router_id 33
    virtual_ipaddress {
        10.205.199.200
    }

    authentication {
        auth_type PASS
        auth_pass password
    }

    track_script {
        check_nginx
    }

}

Edit the password to your own password. This password is used for other keepalived to join the vrrp group. The master & backup load balancer must have the same password.

Same with the Master, the backup also uses script check_nginx to always check the Nginx every 2 seconds.

After that, start the keepalived service

service keepalived start

Checking The Virtual IP

To check the keepalived is working, you can check the IP address of each load balancer using the ip addr command. There will be an additional IP (The virtual IP) on the interface on the active Load Balancer.

ip addr on LB M:

root@loadbalancer-1:/etc/keepalived# ip addr
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
    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 forever
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP group default qlen 1000
    link/ether 00:0c:29:7a:56:e5 brd ff:ff:ff:ff:ff:ff
    inet 10.205.199.174/24 brd 10.205.199.255 scope global dynamic noprefixroute eth0
       valid_lft 5197537sec preferred_lft 5197537sec
    inet 10.205.199.200/32 scope global eth0
       valid_lft forever preferred_lft forever
    inet6 fe80::5871:8c8f:b957:2cce/64 scope link noprefixroute
       valid_lft forever preferred_lft forever

ip addr on the LB B

root@loadbalancer-2:/etc/keepalived# ip addr
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
    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 forever
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP group default qlen 1000
    link/ether 00:0c:29:f2:37:67 brd ff:ff:ff:ff:ff:ff
    inet 10.205.199.175/24 brd 10.205.199.255 scope global dynamic noprefixroute eth0
       valid_lft 7614991sec preferred_lft 7614991sec
    inet6 fe80::ba88:c7e8:9b7c:dd3a/64 scope link noprefixroute
       valid_lft forever preferred_lft forever

You will see the Master Load Balancer get extra ip on the eth0 automatically. And when you stop the nginx service on the LB M or shut down the machine, the ip will move to the Backup Load Balancer, hence the virtual ip (10.205.199.200) will “always” be available. Let’s check if it is true or not.

Now, try to stop the nginx service on the LB M by using this command:

service nginx stop

And then try to check the IP address on both servers again.

ip addr on LB M:

root@loadbalancer-1:/etc/keepalived# ip addr
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
    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 forever
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP group default qlen 1000
    link/ether 00:0c:29:7a:56:e5 brd ff:ff:ff:ff:ff:ff
    inet 10.205.199.174/24 brd 10.205.199.255 scope global dynamic noprefixroute eth0
       valid_lft 5197537sec preferred_lft 5197537sec
    inet6 fe80::5871:8c8f:b957:2cce/64 scope link noprefixroute
       valid_lft forever preferred_lft forever

ip addr on the LB B

root@loadbalancer-2:/etc/keepalived# ip addr
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
    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 forever
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP group default qlen 1000
    link/ether 00:0c:29:f2:37:67 brd ff:ff:ff:ff:ff:ff
    inet 10.205.199.175/24 brd 10.205.199.255 scope global dynamic noprefixroute eth0
       valid_lft 7614991sec preferred_lft 7614991sec
    inet 10.205.199.200/32 scope global eth0
       valid_lft forever preferred_lft forever
    inet6 fe80::ba88:c7e8:9b7c:dd3a/64 scope link noprefixroute
       valid_lft forever preferred_lft forever

You can see that the Virtual IP (10.205.199.200) is moving to the LB B server. So, use this IP to expose your service, because this IP will “always” be available.

That’s it! Congratulation now you can set up your own high availability Nginx using keepalived.

Related Tutorials:



Latest Tutorials:



Learn, Execute, Innovate

This website was developed with love by Bagus Facsi Aginsa, Illustration by Pixeltrue from Ouch!, and hosted on Cloudflare Pages. Please read through our Privacy Policy to ensure you that we value your privacy.