Scale Mesh

Deploy your web apps with ease.

Architecture

Setup and run locally.

Requirements

• Go version go1.23.2

• Install air package for live reload

  go install github.com/air-verse/air@latest
  

• Clone the repository.

1. Run the api-server (main backend server)

cd api-server

Install the dependencies

go mod download

Run the Postgre SQL database

docker run --name pg-docker -e POSTGRES_USER=admin -e POSTGRES_PASSWORD=secret -e POSTGRES_DB=mydb -p 5432:5432 -d postgres

Export the Database URI

export DBURI="host=localhost user=admin password=secret dbname=mydb port=5432 sslmode=disable TimeZone=UTC"

Export the Session secret

export SESSION_KEY=mysecret

Run the API

air --build.cmd "go build -o bin/api ./cmd/web/" --build.bin "./bin/api"

this API will start at port 9000, check the api at 127.0.0.1:9000/health.

2. Run the reverse-proxy API

cd reverse-proxy
go mod download
air --build.cmd "go build -o bin/api ./cmd/web/main.go" --build.bin "./bin/api"

3. Build the build-server container Go to the Build server directory

cd build-server

Build the image

docker build -t scale-mesh/build-server-container-image .

Run the container, it require 2 env variables.

GITHUB_REPO_URL projectID

Components

1. Build Server

2. Reverse Proxy API

3. Main API Server

4. Frontend Server

5. Log Collection Pipeline

Build Server

To build the code and push the artifacts to the S3 bucket.

How it works?

It is a custom Docker container which uses the GitHub repo url, clones it and then build it and push them to SS3 bucket

Structure

• Dockerfile
• entry.sh
• main.go

It is a Multistage Docker build, which builds the Go binary in first stage and then run the bianry in 2nd stage to build the user's web app code and push the build artifacts to the S3 bucket.

Build Server image is pushed to AWS ECR, and then a ECS cluster & Task defination are created to run a container from the ECR image, and after task completed, it'll destroy the container.

S3 buckets are uploaded in this path __output/{projectID}/

API Server

Main backend API, via which user interacts with the Web app.

Working Flow

• User Signup/login
• Save the user's web app project info(GitHub repo url, name) to deploy.
• Deploy the web app.

Structure

• cmd/web/ Contains all the API and business logic.

• pkg Contains the ancillary non-application-specfic code, which could potentially be reused.

• Contains the Database models and methods.

Endpoints of API server

• GET /health to check health of the API.
• POST /deploy to deploy the project.
• POST /project to save the info of the project.
• POST /user/signup to signup.
• POST /user/login to login.
• POST /user/logout to logout.

Reverse Proxy API

To serve the user Web App dynamically using the unique project id.

Frontend Server

Serve a basic HTML template for user to interact with the application.

Logging Pipeline

Build-Server container pushes the logs to the Redis using pub/sub feature and a web socker server is subscribing to the redis channel.

Application Deployment (DevOps)

CI Pipeline (GitLab CICD)

• To build and push the image to Docker Hub registry of API server.

Deployment on Kubernetes

• k8s directory contains the YAML files to deploy the web app on k8s.

• How to deploy

  • Create Namespace.
  • Create secret
  • Create DB
  • Deploy service.