Go 言語イメージのビルド
Overview
In this section you're going to build a container image. The image includes everything you need to run your application – the compiled application binary file, the runtime, the libraries, and all other resources required by your application.
Required software
To complete this tutorial, you need the following:
- Docker running locally. Follow the instructions to download and install Docker.
- An IDE or a text editor to edit files. Visual Studio Code is a free and popular choice but you can use anything you feel comfortable with.
- A Git client. This guide uses a command-line based
gitclient, but you are free to use whatever works for you. - A command-line terminal application. The examples shown in this module are from the Linux shell, but they should work in PowerShell, Windows Command Prompt, or OS X Terminal with minimal, if any, modifications.
Meet the example application
The example application is a caricature of a microservice. It is purposefully trivial to keep focus on learning the basics of containerization for Go applications.
The application offers two HTTP endpoints:
- It responds with a string containing a heart symbol (
<3) to requests to/. - It responds with
{"Status" : "OK"}JSON to a request to/health.
It responds with HTTP error 404 to any other request.
The application listens on a TCP port defined by the value of environment variable PORT. The default value is 8080.
The application is stateless.
The complete source code for the application is on GitHub: github.com/docker/docker-gs-ping. You are encouraged to fork it and experiment with it as much as you like.
To continue, clone the application repository to your local machine:
$ git clone https://github.com/docker/docker-gs-ping
The application's main.go file is straightforward, if you are familiar with Go:
package main
import (
"net/http"
"os"
"github.com/labstack/echo/v4"
"github.com/labstack/echo/v4/middleware"
)
func main() {
e := echo.New()
e.Use(middleware.Logger())
e.Use(middleware.Recover())
e.GET("/", func(c echo.Context) error {
return c.HTML(http.StatusOK, "Hello, Docker! <3")
})
e.GET("/health", func(c echo.Context) error {
return c.JSON(http.StatusOK, struct{ Status string }{Status: "OK"})
})
httpPort := os.Getenv("PORT")
if httpPort == "" {
httpPort = "8080"
}
e.Logger.Fatal(e.Start(":" + httpPort))
}
// Simple implementation of an integer minimum
// Adapted from: https://gobyexample.com/testing-and-benchmarking
func IntMin(a, b int) int {
if a < b {
return a
}
return b
}Create a Dockerfile for the application
To build a container image with Docker, a Dockerfile with build instructions is required.
Begin your Dockerfile with the (optional) parser directive line that instructs BuildKit to
interpret your file according to the grammar rules for the specified version of the syntax.
You then tell Docker what base image you would like to use for your application:
# syntax=docker/dockerfile:1
FROM golang:1.19Docker images can be inherited from other images. Therefore, instead of creating your own base image from scratch, you can use the official Go image that already has all necessary tools and libraries to compile and run a Go application.
メモ
If you are curious about creating your own base images, you can check out the following section of this guide: creating base images. Note, however, that this isn't necessary to continue with your task at hand.
Now that you have defined the base image for your upcoming container image, you can begin building on top of it.
To make things easier when running the rest of your commands, create a directory
inside the image that you're building. This also instructs Docker to use this
directory as the default destination for all subsequent commands. This way you
don't have to type out full file paths in the Dockerfile, the relative paths
will be based on this directory.
WORKDIR /appUsually the very first thing you do once you’ve downloaded a project written in Go is to install the modules necessary to compile it. Note, that the base image has the toolchain already, but your source code isn't in it yet.
So before you can run go mod download inside your image, you need to get your
go.mod and go.sum files copied into it. Use the COPY command to do this.
In its simplest form, the COPY command takes two parameters. The first
parameter tells Docker what files you want to copy into the image. The last
parameter tells Docker where you want that file to be copied to.
Copy the go.mod and go.sum file into your project directory /app which,
owing to your use of WORKDIR, is the current directory (./) inside the
image. Unlike some modern shells that appear to be indifferent to the use of
trailing slash (/), and can figure out what the user meant (most of the time),
Docker's COPY command is quite sensitive in its interpretation of the trailing
slash.
COPY go.mod go.sum ./メモ
If you'd like to familiarize yourself with the trailing slash treatment by the
COPYcommand, see Dockerfile reference. This trailing slash can cause issues in more ways than you can imagine.
Now that you have the module files inside the Docker image that you are
building, you can use the RUN command to run the command go mod download
there as well. This works exactly the same as if you were running go locally
on your machine, but this time these Go modules will be installed into a
directory inside the image.
RUN go mod downloadAt this point, you have a Go toolchain version 1.19.x and all your Go dependencies installed inside the image.
The next thing you need to do is to copy your source code into the image. You’ll
use the COPY command just like you did with your module files before.
COPY *.go ./This COPY command uses a wildcard to copy all files with .go extension
located in the current directory on the host (the directory where the Dockerfile
is located) into the current directory inside the image.
Now, to compile your application, use the familiar RUN command:
RUN CGO_ENABLED=0 GOOS=linux go build -o /docker-gs-pingThis should be familiar. The result of that command will be a static application
binary named docker-gs-ping and located in the root of the filesystem of the
image that you are building. You could have put the binary into any other place
you desire inside that image, the root directory has no special meaning in this
regard. It's just convenient to use it to keep the file paths short for improved
readability.
Now, all that is left to do is to tell Docker what command to run when your image is used to start a container.
You do this with the CMD command:
CMD ["/docker-gs-ping"]Here's the complete Dockerfile:
# syntax=docker/dockerfile:1
FROM golang:1.19
# Set destination for COPY
WORKDIR /app
# Download Go modules
COPY go.mod go.sum ./
RUN go mod download
# Copy the source code. Note the slash at the end, as explained in
# https://docs.docker.com/reference/dockerfile/#copy
COPY *.go ./
# Build
RUN CGO_ENABLED=0 GOOS=linux go build -o /docker-gs-ping
# Optional:
# To bind to a TCP port, runtime parameters must be supplied to the docker command.
# But we can document in the Dockerfile what ports
# the application is going to listen on by default.
# https://docs.docker.com/reference/dockerfile/#expose
EXPOSE 8080
# Run
CMD ["/docker-gs-ping"]The Dockerfile may also contain comments. They always begin with a # symbol,
and must be at the beginning of a line. Comments are there for your convenience
to allow documenting your Dockerfile.
There is also a concept of Dockerfile directives, such as the syntax directive
you added. The directives must always be at the very top of the Dockerfile, so
when adding comments, make sure that the comments follow after any directives
that you may have used:
# syntax=docker/dockerfile:1
# A sample microservice in Go packaged into a container image.
FROM golang:1.19
# ...Build the image
Now that you've created your Dockerfile, build an image from it. The docker build command creates Docker images from the Dockerfile and a context. A
build context is the set of files located in the specified path or URL. The
Docker build process can access any of the files located in the context.
The build command optionally takes a --tag flag. This flag is used to label
the image with a string value, which is easy for humans to read and recognize.
If you don't pass a --tag, Docker will use latest as the default value.
Build your first Docker image.
$ docker build --tag docker-gs-ping .
The build process will print some diagnostic messages as it goes through the build steps. The following is just an example of what these messages may look like.
[+] Building 2.2s (15/15) FINISHED
=> [internal] load build definition from Dockerfile 0.0s
=> => transferring dockerfile: 701B 0.0s
=> [internal] load .dockerignore 0.0s
=> => transferring context: 2B 0.0s
=> resolve image config for docker.io/docker/dockerfile:1 1.1s
=> CACHED docker-image://docker.io/docker/dockerfile:1@sha256:39b85bbfa7536a5feceb7372a0817649ecb2724562a38360f4d6a7782a409b14 0.0s
=> [internal] load build definition from Dockerfile 0.0s
=> [internal] load .dockerignore 0.0s
=> [internal] load metadata for docker.io/library/golang:1.19 0.7s
=> [1/6] FROM docker.io/library/golang:1.19@sha256:5d947843dde82ba1df5ac1b2ebb70b203d106f0423bf5183df3dc96f6bc5a705 0.0s
=> [internal] load build context 0.0s
=> => transferring context: 6.08kB 0.0s
=> CACHED [2/6] WORKDIR /app 0.0s
=> CACHED [3/6] COPY go.mod go.sum ./ 0.0s
=> CACHED [4/6] RUN go mod download 0.0s
=> CACHED [5/6] COPY *.go ./ 0.0s
=> CACHED [6/6] RUN CGO_ENABLED=0 GOOS=linux go build -o /docker-gs-ping 0.0s
=> exporting to image 0.0s
=> => exporting layers 0.0s
=> => writing image sha256:ede8ff889a0d9bc33f7a8da0673763c887a258eb53837dd52445cdca7b7df7e3 0.0s
=> => naming to docker.io/library/docker-gs-ping 0.0s
Your exact output will vary, but provided there aren't any errors, you should
see the word FINISHED in the first line of output. This means Docker has
successfully built your image named docker-gs-ping.
View local images
To see the list of images you have on your local machine, you have two options. One is to use the CLI and the other is to use Docker Desktop. Since you're currently working in the terminal, take a look at listing images with the CLI.
To list images, run the docker image lscommand (or the docker images shorthand):
$ docker image ls
REPOSITORY TAG IMAGE ID CREATED SIZE
docker-gs-ping latest 7f153fbcc0a8 2 minutes ago 1.11GB
...
Your exact output may vary, but you should see the docker-gs-ping image with
the latest tag. Because you didn't specify a custom tag when you built your
image, Docker assumed that the tag would be latest, which is a special value.
Tag images
An image name is made up of slash-separated name components. Name components may contain lowercase letters, digits, and separators. A separator is defined as a period, one or two underscores, or one or more dashes. A name component may not start or end with a separator.
An image is made up of a manifest and a list of layers. In simple terms, a tag points to a combination of these artifacts. You can have multiple tags for the image and, in fact, most images have multiple tags. Create a second tag for the image you built and take a look at its layers.
Use the docker image tag (or docker tag shorthand) command to create a new
tag for your image. This command takes two arguments; the first argument is the
source image, and the second is the new tag to create. The following command
creates a new docker-gs-ping:v1.0 tag for the docker-gs-ping:latest you
built:
$ docker image tag docker-gs-ping:latest docker-gs-ping:v1.0
The Docker tag command creates a new tag for the image. It doesn't create a
new image. The tag points to the same image and is just another way to reference
the image.
Now run the docker image ls command again to see the updated list of local
images:
$ docker image ls
REPOSITORY TAG IMAGE ID CREATED SIZE
docker-gs-ping latest 7f153fbcc0a8 6 minutes ago 1.11GB
docker-gs-ping v1.0 7f153fbcc0a8 6 minutes ago 1.11GB
...
You can see that you have two images that start with docker-gs-ping. You know
they're the same image because if you look at the IMAGE ID column, you can
see that the values are the same for the two images. This value is a unique
identifier Docker uses internally to identify the image.
Remove the tag that you just created. To do this, you’ll use the
docker image rm command, or the shorthand docker rmi (which stands for
"remove image"):
$ docker image rm docker-gs-ping:v1.0
Untagged: docker-gs-ping:v1.0
Notice that the response from Docker tells you that the image hasn't been removed but only untagged.
Verify this by running the following command:
$ docker image ls
You will see that the tag v1.0 is no longer in the list of images kept by your Docker instance.
REPOSITORY TAG IMAGE ID CREATED SIZE
docker-gs-ping latest 7f153fbcc0a8 7 minutes ago 1.11GB
...The tag v1.0 has been removed but you still have the docker-gs-ping:latest
tag available on your machine, so the image is there.
Multi-stage builds
You may have noticed that your docker-gs-ping image weighs in at over a
gigabyte, which is a lot for a tiny compiled Go application. You may also be
wondering what happened to the full suite of Go tools, including the compiler,
after you had built your image.
The answer is that the full toolchain is still there, in the container image. Not only this is inconvenient because of the large file size, but it may also present a security risk when the container is deployed.
These two issues can be solved by using multi-stage builds.
In a nutshell, a multi-stage build can carry over the artifacts from one build stage into another, and every build stage can be instantiated from a different base image.
Thus, in the following example, you are going to use a full-scale official Go image to build your application. Then you'll copy the application binary into another image whose base is very lean and doesn't include the Go toolchain or other optional components.
The Dockerfile.multistage in the sample application's repository has the
following content:
# syntax=docker/dockerfile:1
# Build the application from source
FROM golang:1.19 AS build-stage
WORKDIR /app
COPY go.mod go.sum ./
RUN go mod download
COPY *.go ./
RUN CGO_ENABLED=0 GOOS=linux go build -o /docker-gs-ping
# Run the tests in the container
FROM build-stage AS run-test-stage
RUN go test -v ./...
# Deploy the application binary into a lean image
FROM gcr.io/distroless/base-debian11 AS build-release-stage
WORKDIR /
COPY --from=build-stage /docker-gs-ping /docker-gs-ping
EXPOSE 8080
USER nonroot:nonroot
ENTRYPOINT ["/docker-gs-ping"]Since you have two Dockerfiles now, you have to tell Docker what Dockerfile
you'd like to use to build the image. Tag the new image with multistage. This
tag (like any other, apart from latest) has no special meaning for Docker,
it's just something you chose.
$ docker build -t docker-gs-ping:multistage -f Dockerfile.multistage .
Comparing the sizes of docker-gs-ping:multistage and docker-gs-ping:latest
you see a few orders-of-magnitude difference.
$ docker image ls
REPOSITORY TAG IMAGE ID CREATED SIZE
docker-gs-ping multistage e3fdde09f172 About a minute ago 28.1MB
docker-gs-ping latest 336a3f164d0f About an hour ago 1.11GB
This is so because the "distroless" base image that you have used in the second stage of the build is very barebones and is designed for lean deployments of static binaries.
There's much more to multi-stage builds, including the possibility of multi-architecture builds, so feel free to check out multi-stage builds. This is, however, not essential for your progress here.
Next steps
In this module, you met your example application and built and container image for it.
In the next module, you’ll take a look at how to run your image as a container.