Creating a Docker Swarm Cluster in AWS

This procedure describes how to create a Docker Swarm cluster using the CloudFormation template provided by Docker.

Log into the AWS console.
Create an SSH key pair in the AWS portal and save the *.pem locally:
SSH key pairs are used to log into Docker manager node. You can also use any existing AWS key pair.
1. Change permissions: chmod 400 <*.pem>
2. Add the ssh key: ssh-add *.pem
For more information, see https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/ec2-key-pairs.html#having-ec2-create-your-key-pair.
Go to the Docker for AWS setup & prerequisites website. and click the Deploy Docker Community Edition (CE) for AWS (stable) button.
In the Docker Community Edition (CE) for AWS section, click:
- the first button (Deploy Docker Community Edition (CE) for AWS (stable)) if you want to create the cluster in a new VPC.
- the second button (Deploy Docker Community Edition (CE) for AWS (stable) uses your existing VPC) if you want to use an existing VPC.
After logging into AWS, dialog shown in the next step is displayed.
Click Next to create a Stack.
Specify the stack name, number of swarm managers, worker nodes, instance type, and so on.
Use the key pair name that was created in step 1 for 'Which SSH key to use".

Note: Ensure you specify the right swarm manager and worker instance type.
Accept the acknowledgment, then click Next to create the cluster.

When the cluster is created, you will see a screen similar to this:
Find the IP address of the Docker manager Node from your AWS Cluster Dashboard and ssh to it.
```
ssh -A docker@<manager_node_IP>
```
Copy docker-deploy.tgz (which was downloaded in step 2 in the Customizing the Deployment Manifest) into the Manager node:
```
scp docker-deploy.tgz docker@<manager_node_IP>:/home/docker
```
Note: Always log in and copy the docker-docker folder to the Docker Manager Node ONLY.

Untar the docker-deploy folder.

~ $ cd /home/docker/

~ $ ls -l

total 164

-rw-r--r--    1 docker   docker      164211 Sep 12 17:07 docker-deploy.tgz

~ $ tar -xzvf docker-deploy.tgz

Create a dockerLogin file.
Add the Docker login command generated in step 3 of Customizing the Deployment Manifest.
Change the access permissions as follows:
```
chmod 755 dockerlogin.sh
```

Execute the file from inside Docker Manager.

~/docker-deploy/aws/swarm $ ./dockerlogin.sh
 
WARNING! Using --password via the CLI is insecure. Use --password-stdin.
 
WARNING! Your password will be stored unencrypted in /home/docker/.docker/config.json.
 
Configure a credential helper to remove this warning. See
 
https://docs.docker.com/engine/reference/commandline/login/#credentials-store
 
 
Login Succeeded

Upon successful login, the ~/.docker/config.json Docker configuration file is also created.

List the docker nodes.

$ docker node ls
 
ID                            HOSTNAME                                      STATUS              AVAILABILITY        MANAGER STATUS      ENGINE VERSION
 
f9wmpa5657ayarpyj6tcykxzo     ip-172-31-2-219.us-east-2.compute.internal    Ready               Active                                  18.09.2
 
axydf7bg383f4nfz1syjiavvk     ip-172-31-10-185.us-east-2.compute.internal   Ready               Active              Reachable           18.09.2
 
q54yvmfmz4v7sd5ps3azgo2sq     ip-172-31-16-174.us-east-2.compute.internal   Ready               Active                                  18.09.2
 
wuexbeu4otat81jm3v2o75x86     ip-172-31-18-54.us-east-2.compute.internal    Ready               Active                                  18.09.2
 
14mw5sbkfbi94zdv3pjflemfx     ip-172-31-21-236.us-east-2.compute.internal   Ready               Active              Reachable           18.09.2
 
p9r9mya9f29yg08ifpg5ab9ut *   ip-172-31-32-254.us-east-2.compute.internal   Ready               Active              Leader              18.09.2
 
fdkjxs3eeau2czyarbp7yj8mf     ip-172-31-33-4.us-east-2.compute.internal     Ready               Active                                  18.09.2
 
letp6ywkcnxf68yc92j7dbbze     ip-172-31-40-182.us-east-2.compute.internal   Ready               Active                                  18.09.2

From the available Docker nodes, choose the hostnames where you want the stateful set deployed, then update the hostname locations in the stateful set yml files.

~/docker-deploy/aws/swarm/manifest-aws-swarm $ vi tmgc-nosql.yml 

~/docker-deploy/aws/swarm/manifest-aws-swarm $ vi tmgc-sql.yml 

~/docker-deploy/aws/swarm/manifest-aws-swarm $ vi tmgc-cache.yml 

~/docker-deploy/aws/swarm/manifest-aws-swarm $ vi tmgc-log.yml

For example:

node.hostname == ip-172-31-2-219.us-east-2.compute.internal in tmgc-log.yml

tmgc-log.yml
version: "3.3"
 
services:
   log:
     image: <docker-registry-host>/tml/tml-log:v5.5.0.GA.1
     deploy:
       replicas: 1
       restart_policy:
         condition: on-failure
       placement:
         constraints:
           - node.hostname == ip-172-31-2-219.us-east-2.compute.internal
     env_file:
       - tmgc-log.env
     secrets:
       - source: log_config
         target: /opt/mashery/containeragent/resources/properties/tml_log_properties.json
       - source: cluster-property-vol
         target: /opt/mashery/containeragent/resources/properties/tml_cluster_properties.json
       - source: zones-property-vol
         target: /opt/mashery/containeragent/resources/properties/tml_zones_properties.json
     volumes:
       - logvol:/mnt
     networks:
        - ml5
  
volumes:
   logvol:
     external:
        name: '{{.Service.Name}}-{{.Task.Slot}}-vol'
 
secrets:
   log_config:
     file: ./tml_log_properties.json
   cluster-property-vol:
     file: ./tml_cluster_properties.json
   zones-property-vol:
     file: ./tml_zones_properties.json
 
networks:
   ml5:
    external: true

If the Swarm cluster is created successfully, create an Overlay network "ml5" for the containers networking by executing the following on the manager node.
```
docker network create -d overlay --attachable ml5
```

Contents

Index

Search Results

Creating a Docker Swarm Cluster in AWS