Deploy Spring Boot

Spring Boot should be run on your own machine or server, then exposed with Mekong on its local port.

This is also the general rule for backend apps in MekongTunnel:

• Spring Boot
• Go servers
• FastAPI
• Flask
• PHP / Laravel
• ASP.NET
• Express / Fastify
• any other backend that listens on an HTTP port

Backend + frontend split

You can split the app into two parts:

• Frontend: upload a built static frontend through /dashboard/deploy
• Backend: run the backend locally or on your own VM, then expose its port with mekong <port>

Example split:

Frontend: React / Vue / static site
Backend: Spring Boot / Go / FastAPI / PHP / ASP.NET

In that setup, the frontend can stay hosted online while the backend is exposed from the running server where the API lives.

Spring Boot Maven workflow

Build the JAR:

./mvnw clean package

Run it locally:

java -jar target/app.jar

Expose the running app:

mekong 8080

Spring Boot Gradle workflow

Build the JAR:

./gradlew bootJar

Run it locally:

java -jar build/libs/app.jar

Expose the running app:

mekong 8080

Other backend examples

The same pattern works for many other stacks:

# Go
go run main.go
mekong 8080
 
# FastAPI
uvicorn main:app --host 0.0.0.0 --port 8000
mekong 8000
 
# PHP / Laravel
php artisan serve --host=127.0.0.1 --port=8000
mekong 8000
 
# ASP.NET
dotnet run
mekong 5000

The exact framework does not matter. The only important part is the local port your backend is listening on.

Optional one-time project setup

Save the default port and start command in .mekong.json:

mekong init --port 8080 --start "java -jar target/app.jar"

After that, plain mekong can reuse the saved settings.

When to use dashboard deploy instead

Use /dashboard/deploy when you already have a packaged frontend or app archive ready to upload.

Use a normal tunnel when:

• the app should run from your local machine
• the backend needs live code execution
• you want to expose your dev server directly
• you are testing APIs, webhooks, or local services