A company uses many application VPCs connected to on-premises via Site-to-Site VPN. The network team is migrating to a Transit Gateway and configured the transit gateway with ECMP. Two temporary test VPCs were attached and two new Site-to-Site VPNs to the transit gateway were created, but the team cannot reach 2.5 Gbps across the pair of VPNs. What steps should the team take to improve bandwidth and reduce congestion? (Choose three.)

To improve bandwidth and reduce congestion, creating new accelerated Site-to-Site VPN connections is crucial. AWS Site-to-Site VPN acceleration uses AWS Global Accelerator to route traffic over the AWS global network, significantly improving performance and reducing latency compared to standard VPNs. Advertising the on-premises prefix with the same BGP ASPATH attribute across all VPN connections enables Equal-Cost Multi-Path (ECMP) routing. ECMP allows the Transit Gateway to distribute traffic evenly across multiple paths, maximizing bandwidth utilization. Finally, verifying that the on-premises location is sending traffic using multiple flows is essential because ECMP distributes traffic based on flow hashes. If traffic consists of a single flow, it will only use one VPN tunnel, negating the benefits of ECMP. Incorrect options: Enabling acceleration for existing VPN connections is not an option; acceleration must be configured when the VPN connection is created. Advertising different BGP ASPATH attributes would prevent ECMP, as it would make the paths appear unequal, causing traffic to prefer one path over others. Confirming Transit Gateway attachment AZs is relevant for intra-VPC routing but not directly for improving VPN bandwidth to on-premises.