Use Dagger with Multi-stage Container Builds

Multi-stage builds are a common practice when building containers with Docker.

• First, your application is compiled in a context which has tools that are required for building the application, but not necessarily required for running it.
• Next, to reduce the number of dependencies and hence the size of the image, the compiled application is copied to a different base image which only has the required components to run the application.

tip

Learn more about multi-stage builds in the Docker documentation.

This guide explains how to perform multi-stage builds with the Dagger SDKs.

Requirements

This guide assumes that:

• You have a Go, Python or Node.js development environment. If not, install Go, Python or Node.js.
• You have a Dagger SDK installed for one of the above languages. If not, follow the installation instructions for the Dagger Go, Python or Node.js SDK.
• You have Docker installed and running on the host system. If not, install Docker.
• You have an application that you wish to build. This guide assumes a Go application, but you can use an application of your choice.

Example

The following code snippet demonstrates a multi-stage build.

Go

package main

import (
	"context"
	"fmt"
	"os"

	"dagger.io/dagger"
)

func main() {
	// create dagger client
	ctx := context.Background()
	client, err := dagger.Connect(ctx, dagger.WithLogOutput(os.Stderr))
	if err != nil {
		panic(err)
	}
	defer client.Close()

	// get host directory
	project := client.Host().Directory(".")

	// build app
	builder := client.Container().
		From("golang:latest").
		WithDirectory("/src", project).
		WithWorkdir("/src").
		WithEnvVariable("CGO_ENABLED", "0").
		WithExec([]string{"go", "build", "-o", "myapp"})

	// publish binary on alpine base
	prodImage := client.Container().
		From("alpine").
		WithFile("/bin/myapp", builder.File("/src/myapp")).
		WithEntrypoint([]string{"/bin/myapp"})

	addr, err := prodImage.Publish(ctx, "localhost:5000/multistage")
	if err != nil {
		panic(err)
	}

	fmt.Println(addr)
}

Node.js

import { connect, Client } from "@dagger.io/dagger"

connect(
  async (client: Client) => {
    // get host directory
    const project = client.host().directory(".")

    // build app
    const builder = client
      .container()
      .from("golang:latest")
      .withDirectory("/src", project)
      .withWorkdir("/src")
      .withEnvVariable("CGO_ENABLED", "0")
      .withExec(["go", "build", "-o", "myapp"])

    // publish binary on alpine base
    const prod = client
      .container()
      .from("alpine")
      .withFile("/bin/myapp", builder.file("/src/myapp"))
      .withEntrypoint(["/bin/myapp"])
    const addr = await prod.publish("localhost:5000/multistage")

    console.log(addr)
  },
  { LogOutput: process.stderr },
)

Python

import sys

import anyio

import dagger


async def main():
    # create Dagger client
    async with dagger.Connection(dagger.Config(log_output=sys.stderr)) as client:
        # get host directory
        project = client.host().directory(".")

        # build app
        builder = (
            client.container()
            .from_("golang:latest")
            .with_directory("/src", project)
            .with_workdir("/src")
            .with_env_variable("CGO_ENABLED", "0")
            .with_exec(["go", "build", "-o", "myapp"])
        )

        # publish binary on alpine base
        prod = (
            client.container()
            .from_("alpine")
            .with_file("/bin/myapp", builder.file("/src/myapp"))
            .with_entrypoint(["/bin/myapp"])
        )
        addr = await prod.publish("localhost:5000/multistage")

    print(addr)


anyio.run(main)

This code listing starts by creating a Dagger client and loading the project to be built. It obtains a reference to the project and then builds the application by using the golang:latest image to mount the source directory, sets CGO_ENABLED= since the binary will be published on alpine, and executes go build.

Next, the multi-stage build is achieved by transferring the build artifact from the builder image to a runtime image based on alpine. The steps are:

• Create a new container image which will be used as the runtime image.
• Transfer the build artifact from the builder image to the new container image.
• Set the container entrypoint to the application so that it is executed by default when the container runs.

The final optimized image can now be pushed to a registry and deployed!

Conclusion

This tutorial walked you through the process of performing a multi-stage build with Docker.

Use the API Key Concepts page and the Go, Node.js and Python SDK References to learn more about Dagger.