Which solution will meet these requirements?
Set up a geolocation routing policy. Send the traffic that is near us-west-1 to the on-premises data center. Send the traffic that is near eu-central-1 to eu-central-1.
Set up a simple routing policy that routes all traffic that is near eu-central-1 to eu-central-1 and routes all traffic that is near the on-premises datacenter to the on-premises data center.
Set up a latency routing policy. Associate the policy with us-west-1.
Set up a weighted routing policy. Split the traffic evenly between eu-central-1 and the on-premises data center.
Explanations:
A geolocation routing policy can send traffic based on the geographic location of the requester. This allows traffic near us-west-1 to be directed to the on-premises data center, and traffic near eu-central-1 to be directed to the AWS-hosted website, minimizing load time for users by routing them to the nearest server.
A simple routing policy does not account for geographic location or latency, so it cannot optimize load time based on proximity to the user. It would route all traffic to a specific region without considering geographic proximity.
A latency routing policy helps route traffic based on the lowest latency, but the question specifically asks about minimizing load time, which could be better achieved using a geolocation routing policy. The policy should take geographical location into account rather than just latency.
A weighted routing policy splits traffic between two locations, but it does not optimize for geographic location or minimize load times based on proximity. This policy could lead to suboptimal performance for users in specific regions.
I evaluate that the answer is:
Set up a geolocation routing policy. Send the traffic that is near us-west-1 to the on-premises data center. Send the traffic that is near eu-central-1 to eu-central-1.