Prerequisites

• Install Docker version 1.13 or higher .
• Get Docker Compose as described in Part 3 prerequisites .
• Get Docker Machine as described in Part 4 prerequisites .
• Read the orientation in Part 1 .
• Learn how to create containers in Part 2 .
• Make sure you have published the friendlyhello image you created by pushing it to a registry . We’ll be using that shared image here.
• Be sure your image works as a deployed container. Run this command, slotting in your info for username, repo, and tag: docker run -p 80:80 username/repo:tag, then visit http://localhost/.
• Have the final version of docker-compose.yml from Part 5 handy.

Introduction

You’ve been editing the same Compose file for this entire tutorial. Well, we have good news. That Compose file works just as well in production as it does on your machine. Here, we’ll go through some options for running your Dockerized application.

Choose an option

• Docker CE (Cloud provider)
• Enterprise (Cloud provider)
• Enterprise (On-premise)

If you’re okay with using Docker Community Edition in production, you can use Docker Cloud to help manage your app on popular service providers such as Amazon Web Services, DigitalOcean, and Microsoft Azure.

To set up and deploy:

• Connect Docker Cloud with your preferred provider, granting Docker Cloud permission to automatically provision and “Dockerize” VMs for you.
• Use Docker Cloud to create your computing resources and create your swarm.
• Deploy your app.

Note: We will be linking into the Docker Cloud documentation here; be sure to come back to this page after completing each step.

Connect Docker Cloud

You can run Docker Cloud in standard mode or in Swarm mode .

If you are running Docker Cloud in standard mode, follow instructions below to link your service provider to Docker Cloud.

• Amazon Web Services setup guide
• DigitalOcean setup guide
• Microsoft Azure setup guide
• Packet setup guide
• SoftLayer setup guide
• Use the Docker Cloud Agent to bring your own host

If you are running in Swarm mode (recommended for Amazon Web Services or Microsoft Azure), then skip to the next section on how to create your swarm .

Create your swarm

Ready to create a swarm?

• If you’re on Amazon Web Services (AWS) you can automatically create a swarm on AWS .
• If you are on Microsoft Azure, you can automatically create a swarm on Azure .
• Otherwise, create your nodes in the Docker Cloud UI, and run the docker swarm init and docker swarm join commands you learned in part 4 over SSH via Docker Cloud . Finally, enable Swarm Mode by clicking the toggle at the top of the screen, and register the  swarm you just created.

Note: If you are Using the Docker Cloud Agent to Bring your Own Host , this provider does not support swarm mode. You can register your own existing swarms with Docker Cloud.

Deploy your app on a cloud provider

1. Connect to your swarm via Docker Cloud . There are a couple of different ways to connect:
• From the Docker Cloud web interface in Swarm mode, select Swarms at the top of the page, click the swarm you want to connect to, and copy-paste the given command into a command line terminal.



Or …

• On Docker for Mac or Docker for Windows, you can connect to your swarms directly through the desktop app menus .



Either way, this opens a terminal whose context is your local machine, but whose Docker commands are routed up to the swarm running on your cloud service provider. You directly access both your local file system and your remote swarm, enabling pure docker commands.

2. Run docker stack deploy -c docker-compose.yml getstartedlab to deploy the app on the cloud hosted swarm.

docker stack deploy -c docker-compose.yml getstartedlab

Creating network getstartedlab_webnet

Creating service getstartedlab_web

Creating service getstartedlab_visualizer

Creating service getstartedlab_redis

Your app is now running on your cloud provider.

Run some swarm commands to verify the deployment

You can use the swarm command line, as you’ve done already, to browse and manage the swarm. Here are some examples that should look familiar by now:

• Use docker node ls to list the nodes.

[getstartedlab] ~ $ docker node ls

ID HOSTNAME STATUS AVAILABILITY MANAGER STATUS

9442yi1zie2l34lj01frj3lsn ip-172-31-5-208.us-west-1.compute.internal Ready Active

jr02vg153pfx6jr0j66624e8a ip-172-31-6-237.us-west-1.compute.internal Ready Active

thpgwmoz3qefdvfzp7d9wzfvi ip-172-31-18-121.us-west-1.compute.internal Ready Active

n2bsny0r2b8fey6013kwnom3m * ip-172-31-20-217.us-west-1.compute.internal Ready Active Leader

• Use docker service ls to list services.

[getstartedlab] ~/sandbox/getstart $ docker service ls

ID NAME MODE REPLICAS IMAGE PORTS

x3jyx6uukog9 dockercloud-server-proxy global 1/1 dockercloud/server-proxy *:2376->2376/tcp

ioipby1vcxzm getstartedlab_redis replicated 0/1 redis:latest *:6379->6379/tcp

u5cxv7ppv5o0 getstartedlab_visualizer replicated 0/1 dockersamples/visualizer:stable *:8080->8080/tcp

vy7n2piyqrtr getstartedlab_web replicated 5/5 sam/getstarted:part6 *:80->80/tcp

• Use docker service ps <service> to view tasks for a service.

[getstartedlab] ~/sandbox/getstart $ docker service ps vy7n2piyqrtr

ID NAME IMAGE NODE DESIRED STATE CURRENT STATE ERROR PORTS

qrcd4a9lvjel getstartedlab_web.1 sam/getstarted:part6 ip-172-31-5-208.us-west-1.compute.internal Running Running 20 seconds ago

sknya8t4m51u getstartedlab_web.2 sam/getstarted:part6 ip-172-31-6-237.us-west-1.compute.internal Running Running 17 seconds ago

ia730lfnrslg getstartedlab_web.3 sam/getstarted:part6 ip-172-31-20-217.us-west-1.compute.internal Running Running 21 seconds ago

1edaa97h9u4k getstartedlab_web.4 sam/getstarted:part6 ip-172-31-18-121.us-west-1.compute.internal Running Running 21 seconds ago

uh64ez6ahuew getstartedlab_web.5 sam/getstarted:part6 ip-172-31-18-121.us-west-1.compute.internal Running Running 22 seconds ago

Open ports to services on cloud provider machines

At this point, your app is deployed as a swarm on your cloud provider servers, as evidenced by the docker commands you just ran. But, you still need to open ports on your cloud servers in order to:

• allow communication between the redis service and web service on the worker nodes
• allow inbound traffic to the web service on the worker nodes so that Hello World and Visualizer are accessible from a web browser.
• allow inbound SSH traffic on the server that is running the manager (this may be already set on your cloud provider)

These are the ports you need to expose for each service:

Methods for doing this will vary depending on your cloud provider.

We’ll use Amazon Web Services (AWS) as an example.

What about the redis service to persist data?

To get the redis service working, you need to ssh into the cloud server where the manager is running, and make a data/ directory in /home/docker/ before you run docker stack deploy. Another option is to change the data path in the docker-stack.yml to a pre-existing path on the manager server. This example does not include this step, so the redis service is not up in the example output.

Example: AWS

1. Log in to the AWS Console , go to the EC2 Dashboard, and click into your Running Instances to view the nodes.
2. On the left menu, go to Network & Security > Security Groups.

You’ll see security groups related to your swarm for getstartedlab-Manager-<xxx>, getstartedlab-Nodes-<xxx>, and getstartedlab-SwarmWide-<xxx>.

3. Select the “Node” security group for the swarm. The group name will be something like this: getstartedlab-NodeVpcSG-9HV9SMHDZT8C.
4. Add Inbound rules for the web, visualizer, and redis services, setting the Type, Protocol and Port for each as shown in the table above , and click Save to apply the rules.



Tip: When you save the new rules, HTTP and TCP ports will be auto-created for both IPv4 and IPv6 style addresses.



5. Go to the list of Running Instances, get the public DNS name for one of the workers, and paste it into the address bar of your web browser.

Just as in the previous parts of the tutorial, the Hello World app displays on port 80, and the Visualizer displays on port 8080.

Iteration and cleanup

From here you can do everything you learned about in previous parts of the tutorial.

• Scale the app by changing the docker-compose.yml file and redeploy on-the-fly with the docker stack deploy command.
• Change the app behavior by editing code, then rebuild, and push the new image. (To do this, follow the same steps you took earlier to build the app and publish the image ).
• You can tear down the stack with docker stack rm. For example:

docker stack rm getstartedlab

Unlike the scenario where you were running the swarm on local Docker machine VMs, your swarm and any apps deployed on it will continue to run on cloud servers regardless of whether you shut down your local host.

Congratulations!

You’ve taken a full-stack, dev-to-deploy tour of the entire Docker platform.

There is much more to the Docker platform than what was covered here, but you have a good idea of the basics of containers, images, services, swarms, stacks, scaling, load-balancing, volumes, and placement constraints.

Want to go deeper? Here are some resources we recommend:

• Samples : Our samples include multiple examples of popular software running in containers, and some good labs that teach best practices.
• User Guide : The user guide has several examples that explain networking and storage in greater depth than was covered here.
• Admin Guide : Covers how to manage a Dockerized production environment.
• Training : Official Docker courses that offer in-person instruction and virtual classroom environments.
• Blog : Covers what’s going on with Docker lately. deploy , production , datacenter , cloud , aws , azure , provider , admin , enterprise