Auto-scaling Implementation

Purpose

• Adding an auto-scaling Mechanism for a web server

Steps

1. Create an ELB (how to), but no need to register instances as balancing targets at this moment.

2. Launch Configuration

   • On EC2 Dashboard page -> left side "AUTO SCALING" -> "Launch Configurations" -> "Create launch configuration". (Most of the steps are similar to creating EC2.)

   • On "Configure details" page:

     • Give configuration name

     • Grant IAM Role with privileges if needed. (In this example it's "AmazonS3FullAccess".)

     • Create a web server. (Refer to 3.(3) of this article)

   • Following constructions to complete settings and click "Create Launch Configuration".

3. Create Auto Scaling Group

   • On EC2 Dashboard page -> left side "AUTO SCALING" -> "Auto Scaling Group" -> "Create Auto Scaling Group":

     • Give a group name

     • Add subnet items (each subnet maps to an AZ).

     • "Advanced Details":

       • Tick "Receive traffic from one or more load balancers", and select target groups created in step 1.

       • Set "Health Check Grace Period" seconds with appropriate time for instances to initialize.

   • On "Configure scaling policies" page, choose to keep a fixed number of instances or criteria to increase / decrease group size.

   • On "Configure Notifications" page, set up notification criteria.

   • Follow instructions to complete auto scaling group setting, and go back to EC2 dashboard or browse to ELB DNS name to see the result.