Overview

This little project was built to answare our need for a simple image proxy solution. It is very light, using GD/iMagick through the image intervention library by Oliver Vogel.

Deployment

Performance

Response time

A 3MB file request:

• Check for cached version
• Get source CloudFront
• Fit: 350x350px
• Our CloudFront
• User the average results under 500ms (95% of which is getting the image from source)

CPU Load

With 3 t2.nano instances we utilize less then 10% cpu for 350K requests per day

Usage

 https://static.example.com/param_string/source_url

param_string

Function	String
Fit the original image to the given size	fit:{width}x{height}
Set width	w:{width}
Set width & keep aspect ratio	w:{width}c
Set height	h:{height}
Set height & keep aspect ratio	h:{height}c
Crop to given size (centered)	c:{width}x{height}
Set quality	q:{percent}

Examples

Best fit to a 250x250 box, 100% quality

https://static.example.com/fit:250x250,q:100/source_url

Crop to 250x250 box, 50% quality

https://static.example.com/c:250x250,q:50/source_url

Set width to 250px and keep the height as is

https://static.example.com/w:250/source_url

Set height to 250px and keep aspect ratio

https://static.example.com/h:250c/source_url

Installation

The installation is really straight forward, you can use chef / docker if you know how. or skip the overhead and deploy a plain installation with ec2 images and basic bash scripts

Ngnix configuration example:

server {
        listen 80 default_server;
        listen [::]:80 default_server;

        root /home/ubuntu/image-proxy/src;

        # Add index.php to the list if you are using PHP
        index index.php index.html index.htm index.nginx-debian.html;

        server_name _;

        location / {
                index index.php;
                rewrite ^/check$ / last;
                rewrite ^/(.*)/(.*\:.*) /?source_url=$2&params=$1 last;
        }
        location ~ \.php$ {
                include snippets/fastcgi-php.conf;
                fastcgi_pass unix:/run/php/php7.3-fpm.sock;
        }

}

Deploy From Comunity AMI:

ami-aff007b9

Create a redundant AWS auto scaling deployment

• Create VPC with at lease 2 availability zones (for redundancy)
• Create launch configurations for the AMI(ami-aff007b9) and VPC
• Create an Auto Scaling Group for the Launch Configuration
• Create Scaling policies
• Create Target Group under LOAD BALANCING in the EC2 tab (set health check to path: /check )
• Create a Load Balancer with the Target Group
• Set your DNS for static.yourdoamin.com to the Load Balancer CNAME record
• Test your setup go to: static.yourdomain.com/fit:500x500,q:90/https://static.pexels.com/photos/211929/pexels-photo-211929.jpeg
• OPTIONAL: You can add a CDN to this setup: create a CloudFront distribution connected to your load balancer, and set you DNS to the CloudFront distribution CNAME

TODO

• Create an installation script
• Create sample nginx & php conf files
• Add some security features to the code
• Add Redis cache adapter

Thanks

• Oliver Vogel, for his intervention/image library http://image.intervention.io/