A company wants to improve availability and performance for a hybrid application. The application has a stateful TCP workload on Amazon EC2 instances across multiple AWS Regions and a stateless UDP workload on-premises. Which combination of actions should a solutions architect take to improve availability and performance? (Choose two.)

AWS Global Accelerator improves availability and performance by directing user traffic to the closest healthy endpoint via the AWS global network, reducing latency and increasing fault tolerance for both TCP and UDP workloads. Adding load balancers as endpoints allows Global Accelerator to distribute traffic effectively. Network Load Balancers (NLBs) are ideal for both the stateful TCP workload on EC2 and the stateless UDP workload on-premises because they operate at Layer 4, preserving source IP addresses and handling high throughput with low latency, which is crucial for stateful TCP and efficient for UDP. Incorrect options: Amazon CloudFront is a CDN primarily for caching web content (HTTP/S) and is not suitable for stateful TCP or general UDP workloads. Configuring separate Application Load Balancers (ALBs) for EC2 and on-premises is less efficient and ALBs are Layer 7, not ideal for the given TCP/UDP requirements. Using an ALB for on-premises UDP traffic is incorrect as ALBs do not support UDP.