Skip to main content

EC2 Network & Metrics Hidden Gaps Diagnostic Troubleshooter

Use the interactive troubleshooter below to identify your EC2 network or metrics hidden gap by symptom, review the raw evidence, understand the root cause, and apply the recommended fix.

🚨 Step 1: What specific monitoring blind spot are you experiencing?

Please click the most accurate description:


Quick Reference Table

#ScenarioKey Error SignalRoot CauseThe Fix
1Network microbursts are invisible in CloudWatch instance metricsCloudWatch metrics are not granular enough to reflect these microsecond spikesCloudWatch metrics collected at 1-minute or 5-minute intervals average out microsecond-level network traffic spikes, hiding packet drops.N/A
2Baked AMIs publish metrics under the wrong Instance IDemit metrics using the instance ID of the original instanceThe CloudWatchClient.pm script caches instance metadata locally, causing cloned instances to silently report metrics under the original instance's ID until the TTL expires.rm /var/tmp/aws-mon/instance-id
3ENA Express silently falls back to standard UDP transmissioncommunication between these two instances over UDP uses standard ENA transmissionENA Express requires identical configuration on both the sending and receiving instances; a mismatch causes UDP traffic to silently bypass SRD technology.aws ec2 modify-network-interface-attribute --network-interface-id eni-0123f4567890a1b23 --ena-srd-specification 'EnaSrdEnabled=true,EnaSrdUdpSpecification={EnaSrdUdpEnabled=true}'
4ENA Express SRD eligibility metric miscounts packetsthe packet is still reflected as eligible in the counterPackets exceeding the MTU limit are falsely tracked as eligible by ena_srd_eligible_tx_pkts but cannot transmit via SRD and silently revert to standard ENA.sudo ip link set dev eth0 mtu 8900