Add a backend to your extension
Your extension can ship a backend part with which the frontend can interact with. This page provides information on why and how to add a backend.
Before you start, make sure you have installed the latest version of Docker Desktop.
Tip
Check the Quickstart guide and
docker extension init <my-extension>
. They provide a better base for your extension as it's more up-to-date and related to your install of Docker Desktop.
Why add a backend?
Thanks to the Docker Extensions SDK, most of the time you should be able to do what you need from the Docker CLI directly from the frontend.
Nonetheless, there are some cases where you might need to add a backend to your extension. So far, extension builders have used the backend to:
- Store data in a local database and serve them back with a REST API.
- Store the extension state, for example when a button starts a long-running process, so that if you navigate away from the extension user interface and comes back, the frontend can pick up where it left off.
For more information about extension backends, see Architecture.
Add a backend to the extension
If you created your extension using the docker extension init
command, you already have a backend setup. Otherwise, you have to first create a vm
directory that contains the code and updates the Dockerfile to
containerize it.
Here is the extension folder structure with a backend:
.
├── Dockerfile # (1)
├── Makefile
├── metadata.json
├── ui
└── index.html
└── vm # (2)
├── go.mod
└── main.go
- Contains everything required to build the backend and copy it in the extension's container filesystem.
- The source folder that contains the backend code of the extension.
Although you can start from an empty directory or from the vm-ui extension
sample,
it is highly recommended that you start from the docker extension init
command and change it to suit your needs.
Tip
The
docker extension init
generates a Go backend. But you can still use it as a starting point for your own extension and use any other language like Node.js, Python, Java, .Net, or any other language and framework.
In this tutorial, the backend service simply exposes one route that returns a JSON payload that says "Hello".
{ "Message": "Hello" }
Important
We recommend that, the frontend and the backend communicate through sockets, and named pipes on Windows, instead of HTTP. This prevents port collision with any other running application or container running on the host. Also, some Docker Desktop users are running in constrained environments where they can't open ports on their machines. When choosing the language and framework for your backend, make sure it supports sockets connection.
package main
import (
"flag"
"log"
"net"
"net/http"
"os"
"github.com/labstack/echo"
"github.com/sirupsen/logrus"
)
func main() {
var socketPath string
flag.StringVar(&socketPath, "socket", "/run/guest/volumes-service.sock", "Unix domain socket to listen on")
flag.Parse()
os.RemoveAll(socketPath)
logrus.New().Infof("Starting listening on %s\n", socketPath)
router := echo.New()
router.HideBanner = true
startURL := ""
ln, err := listen(socketPath)
if err != nil {
log.Fatal(err)
}
router.Listener = ln
router.GET("/hello", hello)
log.Fatal(router.Start(startURL))
}
func listen(path string) (net.Listener, error) {
return net.Listen("unix", path)
}
func hello(ctx echo.Context) error {
return ctx.JSON(http.StatusOK, HTTPMessageBody{Message: "hello world"})
}
type HTTPMessageBody struct {
Message string
}
Important
We don't have a working example for Node yet. Fill out the form and let us know if you'd like a sample for Node.
Important
We don't have a working example for Python yet. Fill out the form and let us know if you'd like a sample for Python.
Important
We don't have a working example for Java yet. Fill out the form and let us know if you'd like a sample for Java.
Important
We don't have a working example for .NET. Fill out the form and let us know if you'd like a sample for .NET.
Adapt the Dockerfile
Note
When using the
docker extension init
, it creates aDockerfile
that already contains what is needed for a Go backend.
To deploy your Go backend when installing the extension, you need first to configure the Dockerfile
, so that it:
- Builds the backend application
- Copies the binary in the extension's container filesystem
- Starts the binary when the container starts listening on the extension socket
Tip
To ease version management, you can reuse the same image to build the frontend, build the backend service, and package the extension.
# syntax=docker/dockerfile:1
FROM node:17.7-alpine3.14 AS client-builder
# ... build frontend application
# Build the Go backend
FROM golang:1.17-alpine AS builder
ENV CGO_ENABLED=0
WORKDIR /backend
COPY vm/go.* .
RUN --mount=type=cache,target=/go/pkg/mod \
--mount=type=cache,target=/root/.cache/go-build \
go mod download
COPY vm/. .
RUN --mount=type=cache,target=/go/pkg/mod \
--mount=type=cache,target=/root/.cache/go-build \
go build -trimpath -ldflags="-s -w" -o bin/service
FROM alpine:3.15
# ... add labels and copy the frontend application
COPY --from=builder /backend/bin/service /
CMD /service -socket /run/guest-services/extension-allthethings-extension.sock
Important
We don't have a working Dockerfile for Node yet. Fill out the form and let us know if you'd like a Dockerfile for Node.
Important
We don't have a working Dockerfile for Python yet. Fill out the form and let us know if you'd like a Dockerfile for Python.
Important
We don't have a working Dockerfile for Java yet. Fill out the form and let us know if you'd like a Dockerfile for Java.
Important
We don't have a working Dockerfile for .Net. Fill out the form and let us know if you'd like a Dockerfile for .Net.
Configure the metadata file
To start the backend service of your extension inside the VM of Docker Desktop, you have to configure the image name
in the vm
section of the metadata.json
file.
{
"vm": {
"image": "${DESKTOP_PLUGIN_IMAGE}"
},
"icon": "docker.svg",
"ui": {
...
}
}
For more information on the vm
section of the metadata.json
, see
Metadata.
Warning
Do not replace the
${DESKTOP_PLUGIN_IMAGE}
placeholder in themetadata.json
file. The placeholder is replaced automatically with the correct image name when the extension is installed.
Invoke the extension backend from your frontend
Using the advanced frontend extension example, we can invoke our extension backend.
Use the Docker Desktop Client object and then invoke the /hello
route from the backend service with ddClient. extension.vm.service.get
that returns the body of the response.
Replace the ui/src/App.tsx
file with the following code:
// ui/src/App.tsx
import React, { useEffect } from 'react';
import { createDockerDesktopClient } from "@docker/extension-api-client";
//obtain docker desktop extension client
const ddClient = createDockerDesktopClient();
export function App() {
const ddClient = createDockerDesktopClient();
const [hello, setHello] = useState<string>();
useEffect(() => {
const getHello = async () => {
const result = await ddClient.extension.vm?.service?.get('/hello');
setHello(JSON.stringify(result));
}
getHello()
}, []);
return (
<Typography>{hello}</Typography>
);
}
Important
We don't have an example for Vue yet. Fill out the form and let us know if you'd like a sample with Vue.
Important
We don't have an example for Angular yet. Fill out the form and let us know if you'd like a sample with Angular.
Important
We don't have an example for Svelte yet. Fill out the form and let us know if you'd like a sample with Svelte.
Re-build the extension and update it
Since you have modified the configuration of the extension and added a stage in the Dockerfile, you must re-build the extension.
docker build --tag=awesome-inc/my-extension:latest .
Once built, you need to update it, or install it if you haven't already done so.
docker extension update awesome-inc/my-extension:latest
Now you can see the backend service running in the Containers view of the Docker Desktop Dashboard and watch the logs when you need to debug it.
Tip
You may need to turn on the Show system containers option in Settings to see the backend container running. See Show extension containers for more information.
Open the Docker Desktop Dashboard and select the Containers tab. You should see the response from the backend service call displayed.
What's next?
- Learn how to share and publish your extension.
- Learn more about extensions architecture.