Skip to content
Docker extension for VS Code.
Branch: master
Clone or download
hacst and philliphoff Fix nuget fallback volume mount for dotnet debugging (#793)
Previously the NuGetFallbackFolder mounted into the container used for
debugging was hardcoded to /usr/local/share/dotnet/sdk/NuGetFallbackFolder
for all non-windows platforms. However this path is only the valid
default for OSX. On Linux /usr/share/dotnet/sdk/NuGetFallbackFolder is
the default installation location. This patch corrects that behavior.
Latest commit d45b1ec Mar 22, 2019
Type Name Latest commit message Commit time
Failed to load latest commit information.
.azure-pipelines Clean up gulp/pipelines and re-enable mac tests (#685) Dec 6, 2018
.github Add VS Code bot rules (#769) Jan 30, 2019
.vscode Fix run/run interactive (#843) Mar 12, 2019
build webpack (#645) Nov 27, 2018
commands Add more state icons (#834) Mar 13, 2019
debugging/coreclr Fix nuget fallback volume mount for dotnet debugging (#793) Mar 22, 2019
dockerfile minor fixes PR #656 Dec 6, 2018
images Add more state icons (#834) Mar 13, 2019
style/acr-logs theme color text fixes for view logs (#850) Mar 15, 2019
utils Use win-ca fallback logic since nAPI isn't currently compatible with … Jan 30, 2019
.gitignore Add package-lock.json (#670) Dec 3, 2018
.vscodeignore webpack (#645) Nov 27, 2018 Updating readme file Nov 14, 2015
extension.ts Group icons, fix onCommand, remove group by menu items (#841) Mar 13, 2019
extensionVariables.ts Add group by for images (#823) Mar 11, 2019
main.js Used shared webpack dev code (#770) Feb 14, 2019
parser.ts Copyright headers (#385) Aug 17, 2018
tsconfig.json Clean up gulp/pipelines and re-enable mac tests (#685) Dec 6, 2018
webpack.config.js Used shared webpack dev code (#770) Feb 14, 2019

Docker Support for Visual Studio Code

Version Installs Build Status

The Docker extension makes it easy to build, manage and deploy containerized applications from Visual Studio Code, for example:

  • Automatic Dockerfile, docker-compose.yml, and .dockerignore file generation (Press F1 and search for Docker: Add Docker files to Workspace)
  • Syntax highlighting, hover tips, IntelliSense (completions) for docker-compose.yml and Dockerfile files
  • Linting (errors and warnings) for Dockerfile files
  • Command Palette (F1) integration for the most common Docker commands (for example docker build, docker push, etc.)
  • Explorer integration for managing Images, running Containers, and Docker Hub registries
  • Deploy images from Docker Hub and Azure Container Registries directly to Azure App Service
  • Debug .NET Core applications running in Linux Docker containers
  • Working with docker will walk you through many of the features of this extension


To use much of the Docker extension functionality, you will need to install Docker on your machine and set up on the system path.


Since VS Code runs as a non-root user, you will also need to follow the steps in “Manage Docker as a non-root user” from Post-installation steps for Linux for the extension to be able to access docker.

Generating Docker Files

Press F1 and search for Docker: Add Docker Files to Workspace to generate Dockerfile, docker-compose.yml, docker-compose.debug.yml, and .dockerignore files for your workspace type:


Note: The docker-compose.yml and docker-compose.debug.yml files are not generated for .NET Core applications.


Rich IntelliSense (completions) for Dockerfile and docker-compose.yml files:

IntelliSense for DockerFiles

Docker commands

Many of the most common Docker and docker compose commands are built right into the Command Palette (F1).


Docker View

The Docker extension contributes a new Docker View to VS Code. Within the View, the Explorer lets you view and manage your Images, Containers, and browse your Docker Hub registry. If the Azure Account extension is installed, you can browse your Azure Container Registries as well.

The right click context menu provides quick access to the same rich set of commands found in the Command Palette (F1).

explorer integration

You can move the View up or down by dragging the Docker icon and you can hide the View by right clicking on the icon and choosing Hide. To bring it back, right click on the Activity Bar area and check the Docker item.

show and hide the view

The showExplorer configuration setting controls the visibility of the Docker View.

"docker.showExplorer": false

Docker Hub Login

The first time you expand the Docker Hub node you'll be prompted to log in to your Docker Hub account.

Docker Hub Login

Your user name and password are stored in your operating system credentials vault (for example macOS keychain, Windows Credential Store) so that you don't need to log in every time. You can log out of Docker Hub by right clicking on the Docker Hub label and choosing log out. This will delete the credentials from the OS store.

Deploying images to Azure App Service

With the Docker Explorer you can deploy images from Docker Hub Registries or Azure Container Registries directly to an Azure App Service instance, as detailed in this getting started guide. This functionality requires installing the Azure Account extension and an Azure Subscription. If you do not have an Azure subscription, sign up today for a free 30 day account and get $200 in Azure Credits to try out any combination of Azure services.

To log into Azure, press F1 and search for Azure Sign In. You will then sign into your account using the Device Login flow. Click on "Copy & Open" to open your default browser.

Azure Login

Paste in the access code and continue the sign in process.

Azure Login

You can now right click on an image in Docker Hub or an Azure Container Registry and choose "Deploy Image to Azure App Service".

Deploy to Azure

From here you will be prompted for a Resource Group, location, an App Service Plan, and a globally unique website name.

Azure CLI

Microsoft ships the latest Azure CLI as a Docker image. You can easily launch a container running the CLI from the Command Palette (press F1 and search for Docker: Azure CLI). The extension will then run an interactive terminal attached to the container.

After the container is started, you will be prompted to login to your Azure account. From there, set the subscription you want to work with using az account set (you can see all of your subscriptions with az account list). You do not need to login in every time you run the container because the extension volume mounts the local $HOME/.azure folder to the container's $HOME/.azure folder.

Private registries (Preview)

This build includes preview support for connecting to private registries (such as those described in Docker Hub documentation). At the moment, OAuth is not supported, only basic authentication. We hope to extend this support in the future.

Self-signed and corporate certificates

If you are using a self-signed or corporate CA certificate (e.g. for a private Docker registry) and have the certificate authority's certificate registered in the Windows or Mac certificate store, you will want to use the following setting:

"docker.importCertificates": true

This causes the extension automatically pick up system-wide certificates. Leaving it at the default false means the default Node.js list of trusted certificates will be used. You can fine-tune the values this way:

        "useCertificateStore": true,
        "certificatePaths": [

The exact folder to use for certificatePaths on Linux will depend on the distribution.

Debugging .NET Core (Preview)

Note that Windows containers are not currently supported, only Linux containers. However, both standard and Alpine .NET Core runtime base images are supported.


  1. (All users) Install the .NET Core SDK which includes support for attaching to the .NET Core debugger.

  2. (All users) Install the C# VS Code extension which includes support for attaching to the .NET Core debugger in VS Code.

  3. (Mac users) add /usr/local/share/dotnet/sdk/NuGetFallbackFolder as a shared folder in your Docker preferences.

Docker Shared Folders

Starting the Debugger

To debug a .NET Core application running in a Linux Docker container, add a Docker .NET Core launch configuration:

  1. Switch to the debugging tab.
  2. Select Add configuration...
  3. Select Docker: Launch .NET Core (Preview)
  4. Set a breakpoint.
  5. Start debugging.

Upon debugging, a Docker image will be built and a container will be run based on that image. The container will have volumes mapped to the locally-built application and the .NET Core debugger. If the Docker container exposes port 80, after the debugger is attached the browser will be launched and navigate to the application's initial page.

NOTE: you may see errors in the debug console when debugging ends (e.g. "Error from pipe program 'docker': ..."). This appears due to debugger issue #2439 and should not impact debugging.

Most properties of the configuration are optional and will be inferred from the project. If not, or if there are additional customizations to be made to the Docker image build or container run process, those can be added under the dockerBuild and dockerRun properties of the configuration, respectively.

    "configurations": [
            "name": "Docker: Launch .NET Core (Preview)",
            "type": "docker-coreclr",
            "request": "launch",
            "preLaunchTask": "build",
            "dockerBuild": {
                // Image customizations
            "dockerRun": {
                // Container customizations

Application Customizations

When possible, the location and output of the application will be inferred from the workspace folder opened in VS Code. When they cannot be inferred, these properties can be used to make them explicit:

Property Description Default
appFolder The root folder of the application The workspace folder
appProject The path to the project file The first .csproj or .fsproj found in the application folder
appOutput The application folder relative path to the output assembly The TargetPath MS Build property

You can specify either appFolder or appProject but should not specify both.

Docker Build Customizations

Customize the Docker image build process by adding properties under the dockerBuild configuration property.

Property Description Default
args Build arguments applied to the image. None
context The Docker context used during the build process. The workspace folder, if the same as the application folder; otherwise, the application's parent (i.e. solution) folder
dockerfile The path to the Dockerfile used to build the image. The file Dockerfile in the application folder
labels The set of labels added to the image. = visual-studio-code
tag The tag added to the image. <Application Name>:dev
target The target (stage) of the Dockerfile from which to build the image. base

Example build customizations:

    "configurations": [
            "name": "Launch .NET Core in Docker",
            "type": "docker-coreclr",
            "request": "launch",
            "preLaunchTask": "build",
            "dockerBuild": {
                "args": {
                    "arg1": "value1",
                    "arg2": "value2"
                "context": "${workspaceFolder}/src",
                "dockerfile": "${workspaceFolder}/src/Dockerfile",
                "labels": {
                    "label1": "value1",
                    "label2": "value2"
                "tag": "mytag",
                "target": "publish"

Docker Run Customization

Customize the Docker container run process by adding properties under the dockerRun configuration property.

Property Description Default
containerName The name of the container. <Application Name>-dev
env Environment variables applied to the container. None
envFiles Files of environment variables read in and applied to the container. Environment variables are specified one per line, in <name>=<value> format. None
extraHosts Hosts to be added to the container's hosts file for DNS resolution. None
labels The set of labels added to the container. = visual-studio-code
network The network to which the container will be connected. Use values as described in the Docker run documentation. bridge
ports Ports that are going to be mapped on the host. All ports exposed by the Dockerfile will be bound to a random port on the host machine
volumes Volumes that are going to be mapped to the container. None


Property Description Required Default
hostPort Port number to be bound on the host. No None
containerPort Port number of the container to be bound. Yes None
protocol Specific protocol for the binding (`tcp udp`). If no protocol is specified it will bind both. No


Property Description Required Default
localPath Path on local machine that will be mapped. The folder will be created if it does not exist. Path may use the ${workspaceFolder} variable when needed. Yes None
containerPath Path where the volume will be mapped within the container. The folder will be created if it does not exist. Yes None
permissions Permissions for the container for the mapped volume, rw for read-write or ro for read-only. Yes rw

Example run customization:

    "configurations": [
            "name": "Launch .NET Core in Docker",
            "type": "docker-coreclr",
            "request": "launch",
            "preLaunchTask": "build",
            "dockerRun": {
                "containerName": "my-container",
                "env": {
                    "var1": "value1",
                    "var2": "value2"
                "envFiles": [
                "labels": {
                    "label1": "value1",
                    "label2": "value2"
                "network": "host",
                "ports": [
                        "hostPort": 80,
                        "containerPort": 80
                        "containerPort": 443
                        "containerPort": 6029,
                        "protocol": "udp"
                        "containerPort": 6029,
                        "protocol": "tcp"
                        "hostPort": 4562,
                        "containerPort": 5837,
                        "protocol": "tcp"
                "extraHosts": [
                        "hostname": "some-hostname",
                        "ip": "some-ip"
                        "hostname": "some-other-hostname",
                        "ip": "some-other-ip"
                "volumes": [
                        "localPath": "path-on-host-machine",
                        "containerPath": "path-inside-container",
                        "permissions": "ro|rw"

Configuration Settings

The Docker extension comes with a number of useful configuration settings allowing you to customize your workflow.

Setting Description Default Value
docker.attachShellCommand.linuxContainer Attach command to use for Linux containers /bin/sh
docker.attachShellCommand.windowsContainer Attach command to use for Windows containers powershell
docker.dockerComposeBuild Run docker-compose with the --build argument, defaults to true true
docker.dockerComposeDetached Run docker-compose with the --d (detached) argument, defaults to true true
docker.defaultRegistryPath Default registry and path when tagging an image ""
docker.explorerRefreshInterval Explorer refresh interval, default is 1000ms 1000 Host to connect to (same as setting the DOCKER_HOST environment variable) ""
docker.imageBuildContextPath Build context PATH to pass to Docker build command ""
docker.languageserver.diagnostics.deprecatedMaintainer Controls the diagnostic severity for the deprecated MAINTAINER instruction warning
docker.languageserver.diagnostics.directiveCasing Controls the diagnostic severity for parser directives that are not written in lowercase warning
docker.languageserver.diagnostics.emptyContinuationLine Controls the diagnostic severity for flagging empty continuation lines found in instructions that span multiple lines warning
docker.languageserver.diagnostics.instructionCasing Controls the diagnostic severity for instructions that are not written in uppercase warning
docker.languageserver.diagnostics.instructionCmdMultiple Controls the diagnostic severity for flagging a Dockerfile with multiple CMD instructions warning
docker.languageserver.diagnostics.instructionEntrypointMultiple Controls the diagnostic severity for flagging a Dockerfile with multiple ENTRYPOINT instructions warning
docker.languageserver.diagnostics.instructionHealthcheckMultiple Controls the diagnostic severity for flagging a Dockerfile with multiple HEALTHCHECK instructions warning
docker.languageserver.diagnostics.instructionJSONInSingleQuotes Controls the diagnostic severity for JSON instructions that are written incorrectly with single quotes warning
docker.languageserver.diagnostics.instructionWorkdirRelative Controls the diagnostic severity for WORKDIR instructions that do not point to an absolute path warning
docker.promptOnSystemPrune Prompt for confirmation when running System Prune command true
docker.showExplorer Show or hide the Explorer true
docker.truncateLongRegistryPaths Truncate long Image and Container registry paths in the Explorer false
docker.truncateMaxLength Maximum number of characters for long registry paths in the Explorer, including ellipsis 10


In VS Code, open the Extension Viewlet, type in Docker, locate the extension and click on Install. Once the extension is installed you will be prompted to restart Visual Studio Code which will only take (literally) a couple of seconds.

Of course, you will need to have Docker installed on your computer in order to run commands from the Command Palette (F1, type in Docker).

Running commands on Linux

By default, Docker runs as the root user, requiring other users to access it with sudo. This extension does not assume root access, so you will need to create a Unix group called docker and add users to it. Instructions can be found here: Create a Docker group

Connecting to docker-machine

The default connection of the extension is to connect to the local docker daemon. You can connect to a docker-machine instance if you launch Visual Studio Code and have the DOCKER_HOST environment variable set to a valid host or if you set the configuration setting.

If the docker daemon is using TLS, the DOCKER_CERT_PATH environment variable must also be set (e.g. $HOME\.docker\machine\machines\default). See docker documentation for more information.


There are a couple of ways you can contribute to this repository:

  • Ideas, feature requests and bugs: We are open to all ideas and we want to get rid of bugs! Use the Issues section to either report a new issue, provide your ideas or contribute to existing threads
  • Documentation: Found a typo or strangely worded sentences? Submit a PR!
  • Code: Contribute bug fixes, features or design changes.


Before we can accept your pull request you will need to sign a Contribution License Agreement. All you need to do is to submit a pull request, then the PR will get appropriately labelled (for example cla-required, cla-norequired, cla-signed, cla-already-signed). If you already signed the agreement we will continue with reviewing the PR, otherwise system will tell you how you can sign the CLA. Once you sign the CLA all future PR's will be labeled as cla-signed.


I get "unauthorized: authentication required" in the terminal when executing some commands, such as "Docker: push".

Make sure you are signed in to the Docker Hub or Azure container registry from the docker CLI via docker login (using your username, not your e-mail address).

If you are using an Azure container registry, you will need to get the username and password from Azure by right-clicking on the Azure container registry in the extension and selecting "Browse in the Azure Portal", then selecting the "Access Keys" tab. Getting Azure username and password

Finally, execute docker login, for example:

docker login

and respond with the username and password specified by Azure.

I'm on Linux and get the error "Unable to connect to Docker, is the Docker daemon running?"

Please see Prerequisites and the specific section on Linux.


This extension collects telemetry data to help us build a better experience for building micro-service applications with Docker and VS Code. We only collect data on which commands are executed. We do not collect any information about image names, paths, etc. The extension respects the telemetry.enableTelemetry setting which you can learn more about in our FAQ.



You can’t perform that action at this time.