सारंग's Notes

☑️ Implement autoscaling on memory utilization

2 min read

Scenario:

  • Implement autoscaling on memory utilization
  • SNS will notification each time instance scale up and scale down

In short: Implementing Autoscaling on Memory Utilization with SNS Notifications Understanding the Components:

Services Required:

  1. AWS Auto Scaling Group (ASG): Manages a group of EC2 instances, automatically scaling them up or down based on predefined policies.
  2. Amazon CloudWatch: Monitors various metrics, including memory utilization, from your EC2 instances.
  3. Amazon SNS: A messaging service to send notifications to subscribed endpoints (e.g., email, SMS, or other AWS services).

Step 1: Create an SNS Topic

  • Navigate to the SNS console.
  • Create a new topic.
  • Give it a descriptive name (e.g., “AutoScalingNotifications”).

Step 2: Create an Auto Scaling Group

  • Navigate to the EC2 console and select “Auto Scaling Groups”.
  • Create a new Auto Scaling group with the desired configuration:
    • Launch Configuration: Specify the AMI, instance type, security groups, etc.
    • Scaling Policies:
      • Scaling Adjustment Policy: Define the scaling adjustment (e.g., add or remove instances) based on specific conditions.
      • Target Tracking Scaling Policy: Set a target value for a specific metric (e.g., memory utilization) and let the ASG automatically adjust the number of instances to maintain that target.

Created Launch Template with this User data

#!/bin/bash
yum install nginx unzip -y
systemctl enable --now nginx
curl -O [https://www.free-css.com/assets/files/free-css-templates/download/page296/oxer.zip](https://www.free-css.com/assets/files/free-css-templates/download/page296/oxer.zip)
unzip oxer.zip
rm oxer.zip
mv oxer* /usr/share/nginx/html
systemctl restart nginx

enable sns notification and select previously created sns topic

Step 3: Configure CloudWatch Alarms

  • Navigate to the CloudWatch console. - Create an alarm for the “MemoryUtilization” metric:

    • Metric Name: MemoryUtilization
    • Namespace: AWS/EC2
    • Statistic: Average
    • Period: 5 Minutes (or as needed)
    • Threshold: Set the desired threshold (e.g., 80%)
    • Comparison Operator: Greater Than or Equal To
    • Alarm Actions: Select the SNS topic created in step 1.

  • Associate the Alarm with the Auto Scaling Group:

    • In the Auto Scaling group settings, under “Notifications,” add the CloudWatch alarm created in step 3.

Create Alarm

select memory utilization metric

Graph View

Recent Articles