This exercise should be performed locally using Terraform and Docker. Using Terraform and any docker provider, create the following cluster:

1. Run X web-servers that serves a single static page with a message: Hello from web-server (1..X)

2. Run a load-balancer in front of the web-servers that performs round-robin load balancing

3. On both web-servers and load-balancer, add a health endpoint returning the name of the component (web-server-(1..X) / load-balancer).

4. Create support for setting different versions for the web-server and load-balancer

5. Write a shell script for install(using git pull from GitHub)/start/stop/status of the cluster

Prepare:

# make sure you have curl and docker installed
# install terraform (you can use simple-terraform-ubuntu-install.sh script for Ubuntu)
chmod +x manager.sh # one time
terraform init -upgrade

Working with the script:

./manager.sh start_cluster
./manager.sh stop_cluster
./manager.sh status_cluster

Custom parameters on start:

terraform apply -var "webservers_count=3"
terraform apply -var "nginx_docker_tag=latest" -var "js-webserver_docker_tag=latest"

Testing:

curl localhost:7777
curl localhost:7777/health
curl localhost:8000/health # Where 0 - is X (webserver)

Notes:

• Nginx as round-robin load-balancer