Get Started, Part 6: Deploy your app

Estimated reading time: 13 minutes

Prerequisites

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 go through some options for running your Dockerized application.

Choose an option

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 do not link 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.

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?

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.

    get swarm connect command from Cloud UI

    Or …

    get swarm connect command from Cloud UI

    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:

Service Type Protocol Port
web HTTP TCP 80
visualizer HTTP TCP 8080
redis TCP TCP 6379

Methods for doing this vary depending on your cloud provider.

We 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.

    See the 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 is 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.

    open web service port

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

    security groups rules

  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.

    running instances

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

    Hello World in browser on cloud server

    Visualizer on cloud server

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 continue to run on cloud servers regardless of whether you shut down your local host.

Customers of Docker Enterprise Edition run a stable, commercially-supported version of Docker Engine, and as an add-on they get our first-class management software, Docker Datacenter. You can manage every aspect of your application via UI using Universal Control Plane, run a private image registry with Docker Trusted Registry, integrate with your LDAP provider, sign production images with Docker Content Trust, and many other features.

Take a tour of Docker Enterprise Edition

The bad news is: the only cloud providers with official Docker Enterprise editions are Amazon Web Services and Microsoft Azure.

The good news is: there are one-click templates to quickly deploy Docker Enterprise on each of these providers:

Note: Having trouble with these? View our setup guide for AWS. You can also view the WIP guide for Microsoft Azure.

Once you’re all set up and Datacenter is running, you can deploy your Compose file from directly within the UI.

Deploy an app on DDC

After that, you can see it running, and can change any aspect of the application you choose, or even edit the Compose file itself.

Managing app on DDC

Customers of Docker Enterprise Edition run a stable, commercially-supported version of Docker Engine, and as an add-on they get our first-class management software, Docker Datacenter. You can manage every aspect of your application via UI using Universal Control Plane, run a private image registry with Docker Trusted Registry, integrate with your LDAP provider, sign production images with Docker Content Trust, and many other features.

Take a tour of Docker Enterprise Edition

Bringing your own server to Docker Enterprise and setting up Docker Datacenter essentially involves two steps:

  1. Get Docker Enterprise Edition for your server’s OS from Docker Store.
  2. Follow the instructions to install Datacenter on your own host.

Note: Running Windows containers? View our Windows Server setup guide.

Once you’re all set up and Datacenter is running, you can deploy your Compose file from directly within the UI.

Deploy an app on DDC

After that, you can see it running, and can change any aspect of the application you choose, or even edit the Compose file itself.

Managing app on DDC

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