12-ducp.md

Lab 12: Docker Universal Control Plane

##NOTE This lab is for participants of the Docker Universal Control Plane beta program. It will not work unless you have access to the UCP binaries.

Difficulty: Intermediate

Time: 30 minutes

Tasks:

This lab will guide you through 5 tasks with Docker Universal Control Plane:

• Overview
• Prerequisites
• Task 1: Install Docker Universal Control Plane on ducp-0
• Task 2: Add ducp-1 as a second managed node
• Task 3: Create an Nginx container
• Task 4: Download the Client bundle to ducp-2
• Task 5: Use Docker Compose from ducp-2 to stand up an app

Overview

Docker Universal Control Plane is a an on-premises solution to help you manage your Docker-based applications.

Prerequisites

• This lab assumes you have 3 Docker hosts running Docker Engine <= 1.9 and Linux kernel version 3.16 or higher. These hosts should be named ducp-0, ducp-1, and ducp-2.

• You will need fully qualified domain names (FQDN) and IP addresses for each host.

• You will need a Docker hub account that has been given permission to access the Docker Universal Control Plane beta program.

• The following ports need to be accessible on all of you hosts for UCP to function:

  • 443: UCP controller
  • 2376: Swarm manager
  • 12376: Engine proxy
  • 12379, 12380: Key Value store
  • 12381: Swarm CA service
  • 12382: UCD CA service

• The following additional ports are used as part of this lab:

  • 8005: Nginx container created in Task 3
  • 5000, 5001: Voting containers deployed in Task 5

Task 1: Installing Docker Universal Control Plane

In this task we're going to install the Docker Universal Control Plane (UCP) server onto ducp-0. This is done by running a bootstrap container, and providing a few pieces of information.

Note*: Some dialogs/logs may say "Orca" this is the internal code name for UCP*

1. ssh into ducp-0

    $ ssh -i <indentity file> ubuntu@<ducp-0 public ip>
   

   Note*: You may be prompted to accept the RSA key. If so, enter* yes

2. Run the UCP installer

    docker run --rm -it -v /var/run/docker.sock:/var/run/docker.sock --name ucp dockerorca/ucp install -i
   

3. Provide the following inputs:

   • Initial DUCP password: <password of your choosing>

   • Your Docker hub credentials

   • Additional Aliases: <ducp-0 FQDN> <ducp-0 IP>

     Note*: Do not use the private IP. Use the one labled "IP"*

   The UCP installer should finish something similar to:

    INFO[0160] Installing UCP with host address 172.31.42.38
    INFO[0002] Generating Swarm Root CA
    INFO[0013] Generating UCP Root CA
    INFO[0022] Deploying UCP Containers
    INFO[0027] UCP instance ID: JJOB:SQP3:PERQ:UPP3:54UP:K7B6:ZWL6:GLES:CN7M:5KLO
    INFO[0027] UCP Server SSL: SHA1 Fingerprint=48:22:4F:6B:36:6D:
    INFO[0027] Login as "admin"/(your admin password) to UCP at https://<ducp-0 IP>:443
   

Note: If you are deploying to an EC2 istance, the installer may indicate the server is reachable on the private IP, but instead use your web browser navigate to the UCP server via ducp-0's public IP.

4. Point your browser to https://:443 and login to the UCP server with the username admin (case sensitive) and the password you chose earlier.

Note: You may be warned that your connection is not private. That is because we are not using publicly signed certificates for the SSL connnection to the website. To bypass this in Chrome click advanced and then proceed to . . . . link. Safari may not let you past this warning.

You'll be logged into the UCP dashboard. Notice you have several containers and images, as well as 1 node, and 0 applications running. These images and containers are what power the UCP server.

##Task 2: Deploy a Second Docker Host In this step we'll add a 2nd node (ducp-1) to our cluster.

1. In a new terminal session ssh into ducp-1

    $ ssh -i <identity file> ubuntu@<ducp-1 public ip>
   

   Note*: You may be prompted to accept the RSA key. If so, enter* yes

2. Run the UCP bootstrap with the join option

    docker run --rm -it  -v /var/run/docker.sock:/var/run/docker.sock --name ucp dockerorca/ucp join -i
   

3. Provide the following inputs:

   • URL to the Orca server: https://<ducp-0-public-IP>
   • Proceed with the join: y
   • Admin username: admin
   • Admin password: <password>
   • Docker Hub credentials
   • Additional Aliases: <ducp-1 FQDN> <ducp-1 IP>

   The Installer should finish with something similar to:

    INFO[0000] This engine will join UCP and advertise itself with host address 10.0.11.13
    INFO[0000] Verifying your system is compatible with UCP
    INFO[0012] Starting local swarm containers
   

4. Go back to your web browser, and refresh the dashboard. You should now see you have 2 nodes running.

5. Click Nodes

   Here you can see details on both of your running nodes

##Task 3: Create a Container In this section we'll deploy an Nginx container using UCP

1. In the UCP UI click the Containers icon in the navigation bar on the left side of the screen

   If you do not have a navigation bar on the left of the screen, click the menu button shown below to expand the navigation bar and click the Containers icon.

   

2. On the Containers page, click + Deploy Container

3. Provide the following inputs:

   • Image Name: nginx:latest
   • Container Name: mynginx
   • Under Network section, set port 80 to redirect to 8005 and click the + button

   

   Feel free to examine the other settings, but leave them at their defaults.

4. Click Run Container

   

5. Click the slider icon next to your container

   

   Scroll down to the node section of the page to find out which node your webserver is running on (ducp-0 or ducp-1). Depending on the lab config, this might show as an IP address that you have to manually map back to a particular lab node.

   In your web browser navigate to the IP address (and port 8005) of the node (ducp-0 or ducp-1) where Nginx is running.

   

   For example: http://52.23.41.23:8005

   You should see the Nginx welcome screen.

Task 4: Using UCP from the Command Line

One of the great things about UCP is that it doesn't stop you from using the Docker command line tools you're used to. In this task we're going to install the UCP client bundle onto your local machine.

1. Navigate to your UCP server in your web browser

2. In the upper right corner click admin and choose Profile

3. Click Create a Client Bundle

4. Navigate to where the bundle was downloaded, and unzip the client bundle

    $ unzip bundle.zip
    Archive:  bundle.zip
    extracting: ca.pem
    extracting: cert.pem
    extracting: key.pem
    extracting: cert.pub
    extracting: env.sh
   

5. Change into the directory that was created when the bundle was unzipped

6. Execute the env.sh script to set the appropriate environment variables for your UCP deployment

    $ source env.sh
   

7. Run docker info to examine the configuration of your Docker Swarm. Your output should show that you are managing the swarm vs. a single node.

    $ docker info
    Containers: 10
    Images: 15
    Role: primary
    Strategy: spread
    Filters: health, port, dependency, affinity, constraint
    Nodes: 2
     orca-ducp-0: 10.0.10.47:12376
      └ Containers: 7
      └ Reserved CPUs: 0 / 1
      └ Reserved Memory: 0 B / 3.859 GiB
      └ Labels: executiondriver=native-0.2, kernelversion=3.19.0-26-generic, operatingsystem=Ubuntu 14.04.3 LTS, storagedriver=aufs
     orca-ducp-1: 10.0.11.13:12376
      └ Containers: 3
      └ Reserved CPUs: 0 / 1
      └ Reserved Memory: 0 B / 3.859 GiB
      └ Labels: executiondriver=native-0.2, kernelversion=3.19.0-26-generic, operatingsystem=Ubuntu 14.04.3 LTS, storagedriver=aufs
    CPUs: 2
    Total Memory: 7.718 GiB
    Name: orca-ducp-0
    ID: F4Q3:NJRJ:GZ3M:6TKK:KUEE:TRUO:AEFG:ET7U:RAP4:3RHW:HYOH:I2TK
    Labels:
    swarm_master=tcp://10.0.10.47:2376
   

##Task 5: Use Docker Compose

In this task we'll use Docker Compose to stand up a multi-tier voting application.

1. Use the editor of your choice to create a file named docker-compose.yml, and copy the following text to the file.

    voting-app:
      image: dockercond2/dockercon-voting-app
      links:
        - redis:voteapps_redis_1
      ports:
       	 - "5000:80"
   
    redis:
      image: redis
      ports: ["6379"]
   
    worker:
      image: dockercond2/dockercon-worker
      links:
       	- redis:voteapps_redis_1
        - db:voteapps_db_1
   
    db:
      image: postgres:9.4
   
    result-app:
      image: dockercond2/dockercon-result-app
      links:
        - db:voteapps_db_1
      ports:
       	 - "5001:80"
   

2. Deploy the application. The compose file will start several different containers that comprise a voting app

    $ docker-compose up -d
   

   It will take several minutes for the compose to complete, and several lines of text will scroll by. It should finish similar to this.

    Creating ubuntu_voting-app_1
    Pulling result-app (dockercond2/dockercon-result-app:latest)...
    orca-ducp-1: Pulling dockercond2/dockercon-result-app:latest... : downloaded
    orca-ducp-0: Pulling dockercond2/dockercon-result-app:latest... : downloaded
    Creating ubuntu_result-app_1
   

3. In your web browser navigate back to the UCP server dashboard (https://<ducp-0 public IP>)

   Notice the dashboard now shows 1 application running.

4. Click on the menu icon and select applications from the navigation bar on the left.

   

5. List out all the running containers by clicking Show on the line listing the Ubuntu application

   

   From here you can also see which nodes the voting app and results are running (ducp-0 or ducp-1).

   

   In your web browser navigate to the IP address (and port 5000) of the node (ducp-0 or ducp-1) where the voting app is running.

   For example: http://52.23.41.23:5000

   On another browser tab, navigate to the IP address (and port 5001) of the node (ducp-0 or ducp-1) where the results app is running.

   For example: http://52.23.41.23:5001

6. Switch back to the UCP web UI. You should still be in the App view. Click the slider icon to the right of any of the contaienrs.

   This shows us the details of the running container. We can control container state here. Aditionally we can scale out a given container.

   Finally, the menu bar allows you to see performance stats, logs, and even open a console window into the container.

   Feel free to explore these options.

   

Conclusion

Congrats!! You've learned the basics of working with Docker Universal Control Plane.