ImGo is an image service written in Go.

Introduction

ImGo might be considered like an ambiguous abbreviation for “Images in Go” or “I'm going to learn GoLang” (your choice). The major purpose of the service is to be fast for getting stored images rather fast uploading.

Features

• Upload, download and convert images
• Supported: PostgreSQL, MySQL, Redis, Memcached
• Based on: ImageMagick and Go standard library

OS and Compiller

The project was built with Go 1.18 on Ubuntu 20.04. Hopefully, the project will be able to build within other OS and compiler version.

Dependencies

• ImageMagick

Build

git clone https://github.com/tdv/imgo.git  
cd imgo
go build .

Note If you have a build issue on Ubuntu related on ImageMagick, you can try to install the package with a command like below

sudo apt-get install libmagickwand-dev

Usage

Note
Ones the service was built you need to have the installed PostgreSQL and Redis on your workstation. Nevertheless, that is possible to try all in one out of the box having used a self-contained solution based on docker and docker-compose (see the folder 'docker'). Definitely recommended!

Upload images
For testing image uploads, you can use curl.

curl -is -XPUT "http://localhost:55555/put?format=JPG" --data-binary @./images/1.jpg

To upload an image with a custom size, you can add extra parameters

curl -is -XPUT "http://localhost:55555/put?format=JPG&width=600&height=300" --data-binary @./images/1.jpg

Note
You need to specify the format parameter for the converter (ImageMagick library).

Response example:
HTTP/1.1 200 OK
Content-Type: text/plain
Date: Mon, 26 Aug 2019 21:02:44 GMT
Content-Length: 40

5966f327301f3922fce598f0574fa518d492f808

The returned identifier is intended to obtain the image, as shown in the example below.

Getting images
After the image was uploaded, you can get the one by a link similar to
http://localhost:55555/get/5966f327301f3922fce598f0574fa518d492f808

Tests

The ab utility might be used to assess the speed.

ab -c 800 -n 1000000 -r -k "http://localhost:55555/get/645b92e65c697d9f97a61f81b9a8739dd18e5a1f"


Server Software:        
Server Hostname:        localhost
Server Port:            55555

Document Path:          /get/645b92e65c697d9f97a61f81b9a8739dd18e5a1f
Document Length:        86416 bytes

Concurrency Level:      800
Time taken for tests:   111.834 seconds
Complete requests:      1000000
Failed requests:        0
Keep-Alive requests:    0
Total transferred:      86497000000 bytes
HTML transferred:       86416000000 bytes
Requests per second:    8941.79 [#/sec] (mean)
Time per request:       89.468 [ms] (mean)
Time per request:       0.112 [ms] (mean, across all concurrent requests)
Transfer rate:          755310.84 [Kbytes/sec] received

Connection Times (ms)
              min  mean[+/-sd] median   max
Connect:        0   12  39.0     10    1013
Processing:     1   78   7.4     78     138
Waiting:        0   13   3.8     13     131
Total:          1   89  39.4     88    1112

Let's look at the line from the results

Requests per second:    8941.79 [#/sec] (mean)

I think, that is a good result. If your service really used with loads close to 10k rps, you'll have a lot of interesting tasks...