Nginx
Installation of Nginx load balancer
Nginx is used as both reverse proxy and load balancing for on-prem deployments.
Installation
Run this to install nginx.
Run this to delete default server.
Set
client_max_body_size
to50m
on/etc/nginx/nginx.conf
:Restart nginx.
This is only a one-time installation. Whenever you want to add new servers to this Nginx, follow the Install Servers to Nginx Section.
Install servers to Nginx
Prerequisites
Create wildcard TLS certificates (This certificate can be created each time for all the other servers you can configure later).
Installation
On AWS EC2, the number of network interfaces that can be created is limited depending on the node type. For example on t3a.small
node, the maximum number of network interfaces is 2. Refer to EC2 Network Specifications for more info.
Once nginx server is installed, it will create
sites-enabled
andsites-available
directories inside /etc/nginx directory.Navigate to
/etc/nginx/sites-available
directory and create a file called<sandbox name>.conf
(Example:prod-openg2p.conf
) by using kubernetes/nginx/sites.sample.conf file as a template.
Creation of the <sandbox name>.conf
file applies only to one server in the nginx node. Repeat this section for every server to be added.
Use a new Listen IP Address for every server. It is recommended to add a new network interface in the same VM which is part of the same network.
When configuring upstream servers, you need to configure the node port of the Istio IngressGateway. Therefore, it is important to understand the ports and determine which ports connect to which IngressGateway and for what purpose.
Run this to enable the server that is added now.
Test nginx conf for errors.
Restart nginx.
Post-installation
Map the hostnames to Nginx IPs on your DNS service, such as Route53 on AWS.
Last updated