Autoscaling excercise

This project automates the build of 3 Linux based web servers. The scripts/stress_test.sh script generates traffic against the deployed ELB. Once the traffic reaches a high water threshold (more than 50% of average CPUUsage of the AS group), the number of web servers increases from 3 to 5. When traffic drops below a low water threshold (lower than 30% of average CPUUsage of the AS group) the web farm scales back down from 5 web servers to 3 Also, a monitoring dashboard is deployed showing:

• Active traffic requests
• Historic web server scaling

Deployment

Configure your AWS account credentials. Deploy the following stacks from the cloudformation folder with a custom name:

aws cloudformation deploy --template-file 01-network-layer.yml --stack-name web-server-stress-test-network --parameter-overrides Name=web-server-stress-test-network

aws cloudformation deploy --template-file 02-webservers.yml --stack-name web-server-stress-test --parameter-overrides Name=web-server-stress-test

Generate traffic

Run the scripts/stress_test.sh script against the ELB endpoint (simple copy the LoadBalancerUrl output from the second stack):

./scripts/stress_test.sh <LoadBalancerUrl>

Observe the monitoring dashboard and metrics:

Open the CloudWatch dashboard and observe the historic web server scaling and active request counts.