Deploy Docker Swarm cluster

In this quick start, a Docker Swarm cluster is deployed using the Azure CLI. A multi-container application consisting of web front end and a Redis instance is then deployed and run on the cluster. Once completed, the application is accessible over the internet.

If you don't have an Azure subscription, create a free account before you begin.

This quickstart requires that you are running the Azure CLI version 2.0.4 or later. Run az --version to find the version. If you need to install or upgrade, see Install Azure CLI 2.0.

Create a resource group

Create a resource group with the az group create command. An Azure resource group is a logical group in which Azure resources are deployed and managed.

The following example creates a resource group named myResourceGroup in the westus location.

az group create --name myResourceGroup --location westus


  "id": "/subscriptions/00000000-0000-0000-0000-000000000000/resourceGroups/myResourceGroup",
  "location": "westcentralus",
  "managedBy": null,
  "name": "myResourceGroup",
  "properties": {
    "provisioningState": "Succeeded"
  "tags": null

Create Docker Swarm cluster

Create a Docker Swarm cluster in Azure Container Service with the az acs create command.

The following example creates a cluster named mySwarmCluster with one Linux master node and three Linux agent nodes.

az acs create --name mySwarmCluster --orchestrator-type Swarm --resource-group myResourceGroup --generate-ssh-keys

In some cases, such as with a limited trial, an Azure subscription has limited access to Azure resources. If the deployment fails due to limited available cores, reduce the default agent count by adding --agent-count 1 to the az acs create command.

After several minutes, the command completes and returns json formatted information about the cluster.

Connect to the cluster

Throughout this quick start, you need the IP address of both the Docker Swarm master and the Docker agent pool. Run the following command to return both IP addresses.

az network public-ip list --resource-group myResourceGroup --query "[*].{Name:name,IPAddress:ipAddress}" -o table


Name                                                                 IPAddress
-------------------------------------------------------------------  -------------

Create an SSH tunnel to the Swarm master. Replace IPAddress with the IP address of the Swarm master.

ssh -p 2200 -fNL 2375:localhost:2375 azureuser@IPAddress

Set the DOCKER_HOST environment variable. This allows you to run docker commands against the Docker Swarm without having to specify the name of the host.

export DOCKER_HOST=:2375

You are now ready to run Docker services on the Docker Swarm.

Run the application

Create a file named docker-compose.yaml and copy the following content into it.

version: '3'
    image: redis
    container_name: azure-vote-back
        - "6379:6379"

    image: microsoft/azure-vote-front:v1
    container_name: azure-vote-front
      REDIS: azure-vote-back
        - "80:80"

Run the following command to create the Azure Vote service.

docker-compose up -d


Creating network "user_default" with the default driver
Pulling azure-vote-front (microsoft/azure-vote-front:v1)...
swarm-agent-EE873B23000005: Pulling microsoft/azure-vote-front:v1...
swarm-agent-EE873B23000004: Pulling microsoft/azure-vote-front:v1... : downloaded
Pulling azure-vote-back (redis:latest)...
swarm-agent-EE873B23000004: Pulling redis:latest... : downloaded
Creating azure-vote-front ... 
Creating azure-vote-back ... 
Creating azure-vote-front
Creating azure-vote-back ...

Test the application

Browse to the IP address of the Swarm agent pool to test out the Azure Vote application.

Image of browsing to Azure Vote

Delete cluster

When the cluster is no longer needed, you can use the az group delete command to remove the resource group, container service, and all related resources.

az group delete --name myResourceGroup --yes --no-wait

Get the code

In this quick start, pre-created container images have been used to create a Docker service. The related application code, Dockerfile, and Compose file are available on GitHub.

Next steps

In this quick start, you deployed a Docker Swarm cluster and deployed a multi-container application to it.

To learn about integrating Docker warm with Visual Studio Team Services, continue to the CI/CD with Docker Swarm and VSTS.