A company currently uses a load balancer to distribute traffic to EC2 instances in a single Availability Zone. The company wants to re-architect the solution to meet these needs: inbound requests must be filtered for common web vulnerabilities, rejected requests must be forwarded to a third-party auditing application, and all resources must be highly available. Which architecture meets these requirements?

The correct option addresses all requirements. A Multi-AZ Auto Scaling group ensures high availability and scalability for the EC2 instances. An Application Load Balancer (ALB) distributes traffic and integrates with AWS WAF. AWS WAF filters web vulnerabilities, and associating it with the ALB allows it to inspect inbound requests. Enabling WAF logging with Kinesis Data Firehose as the destination directly forwards rejected requests (logged by WAF) to the third-party auditing application, meeting the auditing requirement efficiently. Subscribing to AWS Managed Rules provides pre-configured protection against common threats. The first incorrect option uses Amazon Inspector, which is for security assessment, not real-time filtering of web vulnerabilities, and relies on Lambda to push reports, which is less direct for rejected requests than Kinesis Data Firehose. The second incorrect option lacks high availability for the EC2 instances and uses CloudWatch Logs with Lambda for auditing, which is less efficient for direct forwarding than Kinesis Data Firehose. The third incorrect option also lacks high availability for the EC2 instances.