A manufacturing company has EC2 instances running in AWS. The EC2 instances are configured with Auto Scaling. There are a lot of requests being lost because of too much load on the servers. The Auto Scaling is launching new EC2 instances to take the load accordingly yet, there are still some requests that are being lost.

Which of the following is the most cost-effective solution to avoid losing recently submitted requests?

Options are :

a. Use an SQS queue to decouple the application components
b. Keep one extra Spot EC2 instance always ready in case a spike occurs.
c. Use larger instances for your application
d. Pre-warm your Elastic Load Balancer