lgtm.

lgtm is a LGTM (Looks Good To Me) image collection application. Users can browse images and copy the URL of the image they like to the clipboard with a single click.

We need some fun in the world of software development. Let's make the world a more pleasant and friendly place with LGTM images!

📚 Technologies

Terraform
Cloudfront CDN with S3 (Origin Access Identity)
API Gateway + Lambda (Serverless)
EventBridge (Scheduled Event)
DynamoDB (NoSQL)
ImageMagick (Image Processing)

🔍 Structure

This repository contains the following directories.

• terraform contains the Terraform configuration files for the infrastructure.
  • We use API Gateway + Lambda for the serverless backend.
  • We use Cloudfront CDN with S3 for distributing the GIF images.
  • We use DynamoDB for storing the metadata of the GIF images and allowing the search feature.
  • We use EventBridge to schedule the Lambda function in lambda-jobs to fetch the latest LGTM images from the Giphy API every 5 minutes.
• lambda-jobs contains the Lambda function source code for the scheduled job that runs every 5 minutes to fetch the latest LGTM images from the Giphy API and then do some image processing to generate the thumbnail and preview images.

🛠 Local development

1. To test the lambda-jobs Lambda function locally, build the Docker image and run the container with the following commands.

$ docker build -t lgtm:latest .
$ docker run -p 9000:8080 -e AWS_ACCESS_KEY_ID=xxx -e AWS_SECRET_ACCESS_KEY=xxx -e GIPHY_API=xxx lgtm:latest

2. Create a DynamoDB table and a S3 bucket called lgtm-tonystrawberry-codes. Make sure the DynamoDB has the same structure as the one defined in terraform/modules/storage/main.tf.

3. Call the local Lambda function with the following command.

$ curl -XPOST "http://localhost:9000/2015-03-31/functions/function/invocations" -d '{"source": "giphy", "keyword": "lgtm"}'

4. Check the DynamoDB table and the S3 bucket to see if the data is correctly stored.

⚙️ Deployment

The application is deployed to AWS with Terraform via Github Actions. The Github Actions workflow is triggered on every push to the main branch.

As a pre-requisite, the following variables need to be set in Terraform.

project = "lgtm-tonystrawberry-codes"
region  = "ap-northeast-1"
giphy_api_key = "xxxx"
opensearch_enabled = false # true if you want to enable the search feature with OpenSearch

1. First, we deploy to ECR the Docker image for the lambda-jobs Lambda function (the one that is used to fetch the latest LGTM images from the Giphy API and then do some image processing to generate the thumbnail and preview images).
2. Then, we deploy the infrastructure with Terraform and use the Docker image from ECR as the Lambda function source code. It will also deploy the Cloudfront CDN with S3, API Gateway + Lambda, DynamoDB, and EventBridge.

🧐 Notes & Related Links

• https://stackoverflow.com/questions/77446420/how-to-give-permission-to-a-local-running-docker-container-lambda-function-acces?noredirect=1#comment136533460_77446420
• OpenSearch was too costly to use, so I decided to implement it for study purposes but not to enable it. You can check the configuration in terraform/modules/opensearch/main.tf.