Acesso a instância EC2 via HTTP e HTTPS lentos para IPs específicos

The slow HTTP and HTTPS connectivity you're experiencing from specific IP addresses to your t3.xlarge EC2 instance could be due to several factors, even though your iperf tests show good bandwidth (~700Mbps).

This issue where only specific IP addresses are affected while others work fine could be related to:

1. Security Groups or Network ACLs: Check if there are any restrictive rules specifically affecting those IP addresses or ranges.

2. Network Configuration: Verify your VPC and subnet configuration, ensuring that DNS hostnames and DNS resolution are properly enabled.

3. Instance-specific issues: While your t3.xlarge should have good network performance, check if there are any specific configurations or applications on that instance causing the slowdown.

4. NAT Gateway limitations: If your instance is using a NAT Gateway for outbound connections, there might be connection or bandwidth limitations affecting specific traffic patterns.

5. External routing or ISP issues: The problem might be in the network path between those specific IPs and AWS, rather than within AWS itself.

AWS doesn't typically implement traffic shaping for specific customer IPs. To troubleshoot further:

• Use VPC Flow Logs and CloudWatch metrics to analyze traffic patterns
• Test connections with different protocols to determine if the issue is specific to HTTP/HTTPS
• Check MTU settings on your instance
• Verify that the instance isn't under heavy load from other processes
• Consider contacting AWS Support as they can provide more detailed analysis of your specific network configuration and potential issues

The significant difference between your iperf results (~700Mbps) and HTTP/HTTPS performance (~150KBps) suggests there might be application-layer issues or specific network configuration problems affecting only web protocols.
Sources
Internet Speed Very Slow | AWS re:Post