COM04-AWS400.md

COM04-AWS400 - Create a cluster of virtual machines using docker swarm

Cloud Service Provider

• Amazon Web Services

Difficulty

• Level 400 (Expert)

Project's Author(s)

Karan Gauswami

Objectives

You need to complete the following:

• Create an EC2 Instance with user data script to install Docker
• Setup that EC2 Instance as a manager node
• Create multiple instances (2 to 3) with user data script to create worker nodes and join them with manager node
• Start any service like nginx to test our Swarm Cluster
• Scale that service to multiple nodes
• Delete all the resources you created after completion

You need to answer the following:

• How to check status of Docker nodes ?
• What behaviour is expected if one of the worker or manager nodes went down ?
• How to get state of all the containers running on different nodes ?
• How to get token to add more workers and managers ?

References

• Running commands on your Linux instance at launch
• Install Docker
• Get Started with Docker Swarm

Costs

• Included in the Free Tier
• $1-2 per hour running

Estimated time to complete

• 4-5 hours

Tips

• Use t2.micro instances for creating nodes to avoid unexpected charges.
• To add more manager or worker nodes, it is not required to SSH into the new instance. Utilize the "user data" functionality.
• Only allow incoming SSH traffic in security groups for the instances that require it.