ANS-C01 Question 374
Select 2You are troubleshooting a network performance issue involving intermittent packet loss between two Amazon EC2 instances in different AWS Regions. Which combination of logs and metrics would be most appropriate to assess the packet loss and diagnose the issue?
- A
VPC Flow Logs for both the source and destination instances
- B
Amazon CloudWatch metrics for the NetworkIn and NetworkOut of the source and destination instances
- C
Amazon CloudTrail logs for API calls related to the EC2 instances
- D
AWS Global Accelerator Flow Logs to check regional connectivity
- E
Enhanced Monitoring metrics for the EC2 instances
Show answer and explanation
Correct answers: A, B
Explanation
To diagnose packet loss between two EC2 instances, it is essential to use tools that provide network traffic insights and metrics. VPC Flow Logs help identify traffic acceptance or rejection, and CloudWatch NetworkIn and NetworkOut metrics highlight potential traffic anomalies. While other options like CloudTrail and Enhanced Monitoring provide valuable information, they are not directly related to diagnosing packet loss in this scenario.
- A. Correct.
VPC Flow Logs provide detailed information about traffic flow at the network interface level, including accepted and rejected packets, making them critical for identifying potential packet loss or network reachability issues.
- B. Correct.
Amazon CloudWatch metrics for NetworkIn and NetworkOut track the volume of incoming and outgoing traffic, helping you detect significant drops in traffic which might correlate with packet loss.
- C. Incorrect.
Amazon CloudTrail logs track API calls and events but do not provide insights into network performance or packet loss, making it less relevant for this scenario.
- D. Incorrect.
AWS Global Accelerator Flow Logs are useful for monitoring traffic through the Global Accelerator service, but they are not applicable here as the scenario does not involve Global Accelerator.
- E. Incorrect.
Enhanced Monitoring metrics for EC2 instances provide detailed instance-level performance metrics, but they do not directly indicate network-level issues like packet loss.