What should the DevOps team do to meet this requirement?
Create an Amazon Simple Notification Service (Amazon SNS) topic. Subscribe the email addresses of all operations team members to the SNS topic. Apply a notification configuration for the autoscaling:EC2_INSTANCE_LAUNCH notification type to all the existing Auto Scaling groups.
Create an Amazon Simple Queue Service (Amazon SQS) queue. Add an AWS Lambda function trigger to the SQS queue. Apply a notification configuration for the autoscaling:EC2_INSTANCE_LAUNCH notification type to all the existing Auto Scaling groups.
Create an Amazon Simple Notification Service (Amazon SNS) topic. Subscribe the email addresses of all operations team members to the SNS topic. Apply a notification configuration for the autoscaling:EC2_INSTANCE_TERMINATE notification type to all the existing Auto Scaling groups.
Create an Amazon Simple Queue Service (Amazon SQS) queue. Add an AWS Lambda function trigger to the SQS queue. Apply a notification configuration for the autoscaling:EC2_INSTANCE_TERMINATE notification type to all the existing Auto Scaling groups.
Explanations:
This option subscribes to theautoscaling:EC2_INSTANCE_LAUNCHnotification type, which notifies when instances are launched, not terminated. Therefore, it does not meet the requirement to notify the operations team when instances are marked unhealthy and terminated.
Similar to option A, this option applies a notification configuration forautoscaling:EC2_INSTANCE_LAUNCH. While it uses SQS and a Lambda function, it still does not address the requirement of notifying about terminated instances.
This option correctly subscribes the operations team to an SNS topic configured with theautoscaling:EC2_INSTANCE_TERMINATEnotification type. This ensures that the operations team receives notifications whenever EC2 instances are terminated due to health check failures, thereby meeting the requirement.
Although this option correctly applies theautoscaling:EC2_INSTANCE_TERMINATEnotification type, it uses SQS with a Lambda function. While this setup could work, it adds unnecessary complexity compared to a straightforward SNS notification. The requirement does not necessitate the use of Lambda or SQS for notifications, making this option less optimal than C.