Swarm workshop

This is the high level plan :: for the swarm workshop

So what we are going to see here is the following flow:

1. Setup 1 Virtual Instances { via Virtualbox } or 1 Instance on AWS
2. Setup 3 Virtual Instances { via Virtualbox } or 3 Instances on AWS
3. Setup all the needed packadges, services etc on each of the vms/instances
4. Configure each node individually
5. Deply a demo application on top of the Swarm Cluster

This project uses Ansible to automate the entire process.

Once we are done for a kickstart you will need to:

1. Install perquisites: Vagrnat (& Virtualbox), Ansible, Git + Optional requirements: AWS account for the aws flow.
2. Invoke Ansible for infra services such as DNS, NTP etc like so: ansible-playbook playbooks/deploy-infra.yml - if you stop here you should have a clean vm named shelleg-infra01 with a Docker daemon you can play arround with Docker ...
3. Invoke Ansible for swarm cluster like so: ansible-playbook playbooks/deploy-swarm-cluster.yml
4. Deploy your application by running: ansible-playbook playbooks/deploy-app.yml [ still WIP ]
5. Test the app lets run zulip via compose -> https://github.com/galexrt/docker-zulip/blob/master/docker-compose.yml

To definitely be continued ...

QuickStart

• git clone this repo
• Download all dependent roles like so:

ansible-galaxy install -r requirements.yml

• Instantiate infra machine

ansible-playbook playbooks/deploy-infra.yml

• Instantiate swarm machine(s)

ansible-playbook playbooks/deploy-swarm-cluster.yml

Testing your swarm cluster

cd vagrant
vagrant ssh swarm-mgr01
docker node ls

Should yield something like:

ID                           HOSTNAME        STATUS  AVAILABILITY  MANAGER STATUS
5ljlrm1lp2o5zpm99056hcnhf *  swarm-mgr01     Ready   Active        Leader
5rm45jkau11ion4kcqtix1um4    swarm-worker01  Ready   Active
bfyad0lw29cf0dgonvgd5y279    swarm-worker02  Ready   Active

BTW, this same command will fail on a worker node ... execute check-swarm-cluster.yml playbook and see !