Deploy Scalable, High Available, and Secured Web App in AWS cloud

iwayqtech commented 3 years ago

AWS-App-Deployment

Goal: Goal of this POC is to prove that webservices can be deployed into private subnets and serve the traffic to public users via Internet facing Load Balancer.

Resources Used: VPC Subnets Route Tables Internet gateway NAT Gateway Security Groups Network ACL Elastic IP S3 Bucket SNS Topic Cloud Watch Alarms EC2 Instances Auto Scaling Group Network Load Balancer Route53 Hosted Zone IAM Policy IAM Role Session Manager

Configuration: You can choose any EC2 configuration (OS, Instance Type, Storage Size, Key Pair etc...) as the goal to deploy a webapp.

Deployment:

Deploy Scalable, High Available, and Secured Web App in AWS cloud as per the architecture provided above.
Create Golden AMI contains global configuration. Install Java8 Install httpd web server Install epel repo
Deploy Auto Scaling group with the Golden AMI created in the previous step.
Configure notification on each Auto Scaling Group event change.
Configure Scaling Policy to scale out when CPU utilization breaches the threshold 80% utilization.
Configure Scaling Policy to scale In when CPU utilization below the threshold 80% utilization.
Keep the index.html file in S3 bucket and configure Launch Template with user data to GET the index.html to /var/www/html location during EC2 launch.
Assume Security Group and Network ACL are integrated in this architecture to allow ports 22 and 80.

Verification: Login to EC2 instances using Session Manager and run stress command to verify the Scale Out activity Browse the web service using a browser from the public internet .

Destroy: Once POC is completed, destroy the resources created in this project.

bikash2241 commented 3 years ago

I had completed this activity. But one doubt If CPU utilization reached to 80 % on only one instance then then why it create 2 more instance and reach to max? It must create only one instance rt?

AntoDevops commented 3 years ago

Great

iwayqtech commented 3 years ago

I had completed this activity. But one doubt If CPU utilization reached to 80 % on only one instance then then why it create 2 more instance and reach to max? It must create only one instance rt?

Great!! Scaling Policy is executed for every 5 minutes interval. Each time the alert is triggered it adds 1 Instance. In your scenario might be the load is high for long time and hence it added 2 more servers.