Deploy application infrastructure

You are viewing docs for legacy standalone Swarm. These topics describe standalone Docker Swarm. In Docker 1.12 and higher, Swarm mode is integrated with Docker Engine. Most users should use integrated Swarm mode — a good place to start is Getting started with swarm mode, Swarm mode CLI commands, and the Get started with Docker walkthrough). Standalone Docker Swarm is not integrated into the Docker Engine API and CLI commands.

Estimated reading time: 13 minutes

In this step, you create several Docker hosts to run your application stack on. Before you continue, make sure you have taken the time to learn the application architecture.

About these instructions

This example assumes you are running on a Mac or Windows system and enabling Docker Engine docker commands by provisioning local VirtualBox virtual machines thru Docker Machine. For this evaluation installation, you’ll need 6 (six) VirtualBox VMs.

While this example uses Docker Machine, this is only one example of an infrastructure you can use. You can create the environment design on whatever infrastructure you wish. For example, you could place the application on another public cloud platform such as Azure or DigitalOcean, on premises in your data center, or even in a test environment on your laptop.

Finally, these instructions use some common bash command substitution techniques to resolve some values, for example:

$ eval $(docker-machine env keystore)

In a Windows environment, these substitution fail. If you are running in Windows, replace the substitution $(docker-machine env keystore) with the actual value.

Task 1. Create the keystore server

To enable a Docker container network and Swarm discovery, you must deploy (or supply) a key-value store. As a discovery backend, the key-value store maintains an up-to-date list of cluster members and shares that list with the Swarm manager. The Swarm manager uses this list to assign tasks to the nodes.

An overlay network requires a key-value store. The key-value store holds information about the network state which includes discovery, networks, endpoints, IP addresses, and more.

Several different backends are supported. This example uses Consul container.

  1. Create a “machine” named keystore.

    $ docker-machine create -d virtualbox --virtualbox-memory "2000" \
    --engine-opt="label=com.function=consul"  keystore
    

    You can set options for the Engine daemon with the --engine-opt flag. You’ll use it to label this Engine instance.

  2. Set your local shell to the keystore Docker host.

    $ eval $(docker-machine env keystore)
    
  3. Run the consul container.

    $ docker run --restart=unless-stopped -d -p 8500:8500 -h consul progrium/consul -server -bootstrap
    

    The -p flag publishes port 8500 on the container which is where the Consul server listens. The server also has several other ports exposed which you can see by running docker ps.

    $ docker ps
    CONTAINER ID        IMAGE               ...       PORTS                                                                            NAMES
    372ffcbc96ed        progrium/consul     ...       53/tcp, 53/udp, 8300-8302/tcp, 8400/tcp, 8301-8302/udp, 0.0.0.0:8500->8500/tcp   dreamy_ptolemy
    
  4. Use a curl command test the server by listing the nodes.

    $ curl $(docker-machine ip keystore):8500/v1/catalog/nodes
    [{"Node":"consul","Address":"172.17.0.2"}]
    

Task 2. Create the Swarm manager

In this step, you create the Swarm manager and connect it to the keystore instance. The Swarm manager container is the heart of your Swarm cluster. It is responsible for receiving all Docker commands sent to the cluster, and for scheduling resources against the cluster. In a real-world production deployment, you should configure additional replica Swarm managers as secondaries for high availability (HA).

You’ll use the --eng-opt flag to set the cluster-store and cluster-advertise options to refer to the keystore server. These options support the container network you’ll create later.

  1. Create the manager host.

    $ docker-machine create -d virtualbox --virtualbox-memory "2000" \
    --engine-opt="label=com.function=manager" \
    --engine-opt="cluster-store=consul://$(docker-machine ip keystore):8500" \
    --engine-opt="cluster-advertise=eth1:2376" manager
    

    You also give the daemon a manager label.

  2. Set your local shell to the manager Docker host.

    $ eval $(docker-machine env manager)
    
  3. Start the Swarm manager process.

    $ docker run --restart=unless-stopped -d -p 3376:2375 \
    -v /var/lib/boot2docker:/certs:ro \
    swarm manage --tlsverify \
    --tlscacert=/certs/ca.pem \
    --tlscert=/certs/server.pem \
    --tlskey=/certs/server-key.pem \
    consul://$(docker-machine ip keystore):8500
    

    This command uses the TLS certificates created for the boot2docker.iso or the manager. This is key for the manager when it connects to other machines in the cluster.

  4. Test your work by using displaying the Docker daemon logs from the host.

    $ docker-machine ssh manager
    <-- output snipped -->
    docker@manager:~$ tail /var/lib/boot2docker/docker.log
    time="2016-04-06T23:11:56.481947896Z" level=debug msg="Calling GET /v1.15/version"
    time="2016-04-06T23:11:56.481984742Z" level=debug msg="GET /v1.15/version"
    time="2016-04-06T23:12:13.070231761Z" level=debug msg="Watch triggered with 1 nodes" discovery=consul
    time="2016-04-06T23:12:33.069387215Z" level=debug msg="Watch triggered with 1 nodes" discovery=consul
    time="2016-04-06T23:12:53.069471308Z" level=debug msg="Watch triggered with 1 nodes" discovery=consul
    time="2016-04-06T23:13:13.069512320Z" level=debug msg="Watch triggered with 1 nodes" discovery=consul
    time="2016-04-06T23:13:33.070021418Z" level=debug msg="Watch triggered with 1 nodes" discovery=consul
    time="2016-04-06T23:13:53.069395005Z" level=debug msg="Watch triggered with 1 nodes" discovery=consul
    time="2016-04-06T23:14:13.071417551Z" level=debug msg="Watch triggered with 1 nodes" discovery=consul
    time="2016-04-06T23:14:33.069843647Z" level=debug msg="Watch triggered with 1 nodes" discovery=consul
    

    The output indicates that the consul and the manager are communicating correctly.

  5. Exit the Docker host.

    docker@manager:~$ exit
    

Task 3. Add the load balancer

The application uses Interlock and Nginx as a loadbalancer. Before you build the load balancer host, you’ll create the configuration you’ll use for Nginx.

  1. On your local host, create a config directory.

  2. Change directories to the config directory.

    $ cd config
    
  3. Get the IP address of the Swarm manager host.

    For example:

    $ docker-machine ip manager
    192.168.99.101
    
  4. Use your favorite editor to create a config.toml file and add this content to the file:

    ListenAddr = ":8080"
    DockerURL = "tcp://SWARM_MANAGER_IP:3376"
    TLSCACert = "/var/lib/boot2docker/ca.pem"
    TLSCert = "/var/lib/boot2docker/server.pem"
    TLSKey = "/var/lib/boot2docker/server-key.pem"
    
    [[Extensions]]
    Name = "nginx"
    ConfigPath = "/etc/nginx/nginx.conf"
    PidPath = "/var/run/nginx.pid"
    MaxConn = 1024
    Port = 80
    
  5. In the configuration, replace the SWARM_MANAGER_IP with the manager IP you got in Step 4.

    You use this value because the load balancer listens on the manager’s event stream.

  6. Save and close the config.toml file.

  7. Create a machine for the load balancer.

    $ docker-machine create -d virtualbox --virtualbox-memory "2000" \
    --engine-opt="label=com.function=interlock" loadbalancer
    
  8. Switch the environment to the loadbalancer.

    $ eval $(docker-machine env loadbalancer)
    
  9. Start an interlock container running.

    $ docker run \
        -P \
        -d \
        -ti \
        -v nginx:/etc/conf \
        -v /var/lib/boot2docker:/var/lib/boot2docker:ro \
        -v /var/run/docker.sock:/var/run/docker.sock \
        -v $(pwd)/config.toml:/etc/config.toml \
        --name interlock \
        ehazlett/interlock:1.0.1 \
        -D run -c /etc/config.toml
    

    This command relies on the config.toml file being in the current directory. After running the command, confirm the image is runing:

    $ docker ps
    CONTAINER ID        IMAGE                      COMMAND                  CREATED             STATUS              PORTS                     NAMES
    d846b801a978        ehazlett/interlock:1.0.1   "/bin/interlock -D ru"   2 minutes ago       Up 2 minutes        0.0.0.0:32770->8080/tcp   interlock
    

    If you don’t see the image runing, use docker ps -a to list all images to make sure the system attempted to start the image. Then, get the logs to see why the container failed to start.

    $ docker logs interlock
    INFO[0000] interlock 1.0.1 (000291d)
    DEBU[0000] loading config from: /etc/config.toml
    FATA[0000] read /etc/config.toml: is a directory
    

    This error usually means you weren’t starting the docker run from the same config directory where the config.toml file is. If you run the command and get a Conflict error such as:

    docker: Error response from daemon: Conflict. The name "/interlock" is already in use by container d846b801a978c76979d46a839bb05c26d2ab949ff9f4f740b06b5e2564bae958. You have to remove (or rename) that container to be able to reuse that name.
    

    Remove the interlock container with the docker rm interlock and try again.

  10. Start an nginx container on the load balancer.

    $ docker run -ti -d \
      -p 80:80 \
      --label interlock.ext.name=nginx \
      --link=interlock:interlock \
      -v nginx:/etc/conf \
      --name nginx \
      nginx nginx -g "daemon off;" -c /etc/conf/nginx.conf
    

Task 4. Create the other Swarm nodes

A host in a Swarm cluster is called a node. You’ve already created the manager node. Here, the task is to create each virtual host for each node. There are three commands required:

  • create the host with Docker Machine
  • point the local environment to the new host
  • join the host to the Swarm cluster

If you were building this in a non-Mac/Windows environment, you’d only need to run the join command to add a node to the Swarm cluster and register it with the Consul discovery service. When you create a node, you also give it a label, for example:

--engine-opt="label=com.function=frontend01"

You’ll use these labels later when starting application containers. In the commands below, notice the label you are applying to each node.

  1. Create the frontend01 host and add it to the Swarm cluster.

    $ docker-machine create -d virtualbox --virtualbox-memory "2000" \
    --engine-opt="label=com.function=frontend01" \
    --engine-opt="cluster-store=consul://$(docker-machine ip keystore):8500" \
    --engine-opt="cluster-advertise=eth1:2376" frontend01
    $ eval $(docker-machine env frontend01)
    $ docker run -d swarm join --addr=$(docker-machine ip frontend01):2376 consul://$(docker-machine ip keystore):8500
    
  2. Create the frontend02 VM.

    $ docker-machine create -d virtualbox --virtualbox-memory "2000" \
    --engine-opt="label=com.function=frontend02" \
    --engine-opt="cluster-store=consul://$(docker-machine ip keystore):8500" \
    --engine-opt="cluster-advertise=eth1:2376" frontend02
    $ eval $(docker-machine env frontend02)
    $ docker run -d swarm join --addr=$(docker-machine ip frontend02):2376 consul://$(docker-machine ip keystore):8500
    
  3. Create the worker01 VM.

    $ docker-machine create -d virtualbox --virtualbox-memory "2000" \
    --engine-opt="label=com.function=worker01" \
    --engine-opt="cluster-store=consul://$(docker-machine ip keystore):8500" \
    --engine-opt="cluster-advertise=eth1:2376" worker01
    $ eval $(docker-machine env worker01)
    $ docker run -d swarm join --addr=$(docker-machine ip worker01):2376 consul://$(docker-machine ip keystore):8500
    
  4. Create the dbstore VM.

    $ docker-machine create -d virtualbox --virtualbox-memory "2000" \
    --engine-opt="label=com.function=dbstore" \
    --engine-opt="cluster-store=consul://$(docker-machine ip keystore):8500" \
    --engine-opt="cluster-advertise=eth1:2376" dbstore
    $ eval $(docker-machine env dbstore)
    $ docker run -d swarm join --addr=$(docker-machine ip dbstore):2376 consul://$(docker-machine ip keystore):8500
    
  5. Check your work.

    At this point, you have deployed on the infrastructure you need to run the application. Test this now by listing the running machines:

    $ docker-machine ls
    NAME           ACTIVE   DRIVER       STATE     URL                         SWARM   DOCKER    ERRORS
    dbstore        -        virtualbox   Running   tcp://192.168.99.111:2376           v1.10.3
    frontend01     -        virtualbox   Running   tcp://192.168.99.108:2376           v1.10.3
    frontend02     -        virtualbox   Running   tcp://192.168.99.109:2376           v1.10.3
    keystore       -        virtualbox   Running   tcp://192.168.99.100:2376           v1.10.3
    loadbalancer   -        virtualbox   Running   tcp://192.168.99.107:2376           v1.10.3
    manager        -        virtualbox   Running   tcp://192.168.99.101:2376           v1.10.3
    worker01       *        virtualbox   Running   tcp://192.168.99.110:2376           v1.10.3
    
  6. Make sure the Swarm manager sees all your nodes.

    $ docker -H $(docker-machine ip manager):3376 info
    Containers: 4
     Running: 4
     Paused: 0
     Stopped: 0
    Images: 3
    Server Version: swarm/1.1.3
    Role: primary
    Strategy: spread
    Filters: health, port, dependency, affinity, constraint
    Nodes: 4
     dbstore: 192.168.99.111:2376
      └ Status: Healthy
      └ Containers: 1
      └ Reserved CPUs: 0 / 1
      └ Reserved Memory: 0 B / 2.004 GiB
      └ Labels: com.function=dbstore, executiondriver=native-0.2, kernelversion=4.1.19-boot2docker, operatingsystem=Boot2Docker 1.10.3 (TCL 6.4.1); master : 625117e - Thu Mar 10 22:09:02 UTC 2016, provider=virtualbox, storagedriver=aufs
      └ Error: (none)
      └ UpdatedAt: 2016-04-07T18:25:37Z
     frontend01: 192.168.99.108:2376
      └ Status: Healthy
      └ Containers: 1
      └ Reserved CPUs: 0 / 1
      └ Reserved Memory: 0 B / 2.004 GiB
      └ Labels: com.function=frontend01, executiondriver=native-0.2, kernelversion=4.1.19-boot2docker, operatingsystem=Boot2Docker 1.10.3 (TCL 6.4.1); master : 625117e - Thu Mar 10 22:09:02 UTC 2016, provider=virtualbox, storagedriver=aufs
      └ Error: (none)
      └ UpdatedAt: 2016-04-07T18:26:10Z
     frontend02: 192.168.99.109:2376
      └ Status: Healthy
      └ Containers: 1
      └ Reserved CPUs: 0 / 1
      └ Reserved Memory: 0 B / 2.004 GiB
      └ Labels: com.function=frontend02, executiondriver=native-0.2, kernelversion=4.1.19-boot2docker, operatingsystem=Boot2Docker 1.10.3 (TCL 6.4.1); master : 625117e - Thu Mar 10 22:09:02 UTC 2016, provider=virtualbox, storagedriver=aufs
      └ Error: (none)
      └ UpdatedAt: 2016-04-07T18:25:43Z
     worker01: 192.168.99.110:2376
      └ Status: Healthy
      └ Containers: 1
      └ Reserved CPUs: 0 / 1
      └ Reserved Memory: 0 B / 2.004 GiB
      └ Labels: com.function=worker01, executiondriver=native-0.2, kernelversion=4.1.19-boot2docker, operatingsystem=Boot2Docker 1.10.3 (TCL 6.4.1); master : 625117e - Thu Mar 10 22:09:02 UTC 2016, provider=virtualbox, storagedriver=aufs
      └ Error: (none)
      └ UpdatedAt: 2016-04-07T18:25:56Z
    Plugins:
     Volume:
     Network:
    Kernel Version: 4.1.19-boot2docker
    Operating System: linux
    Architecture: amd64
    CPUs: 4
    Total Memory: 8.017 GiB
    Name: bb13b7cf80e8
    

    The command is acting on the Swarm port, so it returns information about the entire cluster. You have a manager and no nodes.

Next Step

Your key-value store, load balancer, and Swarm cluster infrastructure are up. You are ready to build and run the voting application on it.

docker, swarm, scale, voting, application, certificates