Services Covered

  • EC2 EC2

Lab description

Configure Auto Scaling to automatically launch EC2 instances using conditions described by CloudWatch alarms. In order to initiate new instances stress tool will cause a CPU utilization spike.

Learning Objectives

  • Create and configure a Network Load Balance
  • Utilize Auto Scaling and a Network Load Balancer
  • Build an elastic cluster by integrating Auto Scaling with an Elastic Load Balancer

Lab date

02-11-2021

Prerequisites

  • AWS account

Lab steps

  1. Create a Network Load Balancer. Choose: Internet-facing in all availability zones, Create target group in Listeners and routing. Target groups configuration: Instances, protocol TCP,port 80, Health checks Interval set to 10 seconds. Back in NLB creation tab choose the newly created target group

  2. When the load balancer is created select it and from Actions drop-down menu choose Edit attributes. Then enable Cross-Zone Load Balancing.
  3. Go to Target Groups, choose the earlier created group and change it’s Deregistration delay to 30 seconds under Attributes tab.
  4. Create a Security group. Allow inbound SSH and HTTP from anywhere.
  5. Create a Launch Template: Amazon Linux 2, t2.micro, add key pair and attach secuirty group from previous step. Enable Detailed CloudWatch monitoring. Add following User data 
    #!/bin/bash

# Enable the epel-release
sudo amazon-linux-extras install epel

# Install and start Apache web server
sudo yum install -y httpd php

# Start the httpd service
service httpd start

# Install CPU stress test tool
sudo yum install -y stress

    This bash script installs PHP, an Apache webserver (httpd), and a tool for stress testing called Stress.

  6. Create an Auto Scaling Group and use launch template from previous step. Under Load balancing choose Attach to an existing load balancer and use the one created earlier. Under Health check grace period, enter 80. Set Maximum capacity to 4. Scaling policies: choose Target tracking scaling policy. In the Instances need text-box, enter 80.
  7. Navigate to Load Balancer and open the DNS name in your browser. You should see Apache webserver sample page
  8. Terminate that running instance, the auto scaling group will detect the changes and initiate a new Instance. SSH to that instance and run following command to run stress 
    stress --cpu 2 --io 1 --vm 1 --vm-bytes 128M --timeout 5m

    This will cause a CPU utilization spike which will trigger the Auto Scaling Group to initiate additional instance. You can follow the CPU utilization in CloudWatch detailed monitoring.

  9. You will see two running instances, or if your auto-scaling group has already initiated a scale-in (because the stress command finished and CPU utilization dropped), you will see two terminated instances including the one you manually terminated, and one running instance.

Lab files

Acknowledgements

By : LU Tags: