Simple Go serverless website on Amazon Web Services (AWS).
Demo: https://6epko5iya8.execute-api.ap-southeast-1.amazonaws.com/dev.
Website's frontend is written in VueJS
, which stored on a public AWS S3
bucket.
There're two simple AWS Lambdas
written in Go (deployed via apex
with NodeJS shim runtime):
less_crawler
: Crawls Github trending repositories, parses and persists data toDynamoDB
. This function is triggered by aCloudWatch
cron job.less_caterer
: Receives HTTP request from client and lookup onDynamoDB
for trending repositories by day.
API Gateway
serves two APIs, one to GET the index.html
page on S3 bucket (forward/proxy request to the URL of index.html
file on S3). The other allows client to lookup Gihub trending repositories by day (which calling the less_caterer
lambda).
This repo contains the code for Lambda
and S3
parts in the image below.
You have to install Go SDK, glide, AWS CLI, apex and configure the AWS credential to deploy your functions on AWS Lambda.
-
Clone the repository to your local
$GOPATH
and install Go dependencies:$ go get github.com/lnquy/less $ cd $GOPATH/src/github.com/lnquy/less $ glide install
-
The AWS region is
ap-southeast-1
, you have to take a look on thefunctions
code and change theawsRegion
value to deploy on another region.
-
Create a role for the Lambdas on AWS IAM which have access to
DynamoDB
service. -
Change the role to match your IAM Lambda role: project.json#L4.
-
Deploy to AWS:
$ apex deploy
-
Create a public S3 bucket for static web hosting, get the URL of the bucket.
-
Change the
publicPath
to your S3 bucket URL: webpack.config.js#L6 -
Change the
GetCaterer
to yourPOST /api/v1/trending/
API URL: main.js#L22 -
Build frontend:
$ cd frontend $ npm install // or yarn $ npm run build
-
Upload all
frontend/dist
files to the root of S3 bucket, make sure all files has public read permission. -
Note the HTTP URL to the
index.html
file.
- Create a
24_hours
interval scheduled job. - Apply that job as the trigger of the
less_crawler
lambda.
-
Create a table with name of
less-crawler-dev
or anything you want, just make sure to change thedynamoTable
value in Go code, too. -
Primary partition key:
date
(String). -
Primary sort key:
sort
(Number).
- Create a root
/
GET API to forward/proxy the HTTP request to the URL ofindex.html
file on S3 bucket. - Create a
/api/v1/trending
POST API which called theless_caterer
lambda. - You may have to allow the CORS permission on APIs, too.
- Deploy the APIs to a stage (E.g:
dev
).
Open browser and follow the link to the GET /
API, you now have a simple serverless website up and running on AWS. Congrats :)
This project is under the MIT License. See the LICENSE file for the full license text.