ENA Network Performance Troubleshooter
Use the interactive troubleshooter below to identify your ENA network performance issue by symptom, review the raw error log, understand the root cause, and apply the recommended fix.
🚨 Step 1: What specific error symptom are you experiencing?
Please click the most accurate description:
Quick Reference Table
| # | Scenario | Key Error Signal | Root Cause | The Fix |
|---|---|---|---|---|
| 1 | Applications that send and receive packets cannot communicate properly due to dropped packets. | conntrack_allowance_exceeded: 0 | Connection tracking exceeded the maximum for the instance and new connections could not be established. | Scale either the number of instances registered with the load balancer or the size of the instances registered with the load balancer. |
| 2 | Packet loss impacting traffic to the DNS service, the Instance Metadata Service, and the Amazon Time Sync Service. | linklocal_allowance_exceeded: 0 | The PPS of the traffic to local proxy services exceeded the maximum for the network interface. | N/A |
| 3 | ENA Express falls back from SRD to standard ENA transmission, increasing the possibility of dropped packets. | ena_srd_resource_utilization: 0 | The instance has approached 100% of the maximum allowed memory utilization for concurrent SRD connections. | Scale the instance out to improve network performance. |
| 4 | Network packets are queued and then dropped due to hitting instance-level packet-per-second ceilings. | pps_allowance_exceeded: 0 | The bidirectional PPS exceeded the maximum for the instance. | N/A |
| 5 | ENA Express fails to operate effectively or process packets optimally. | ethtool -G device rx 8192 | The receiver (Rx) ring size is not optimized for ENA Express packet handling. | ethtool -G device rx 8192 |
| 6 | Network packets are queued or dropped because the bandwidth limit is hit. | bw_in_allowance_exceeded: 0 | The inbound or outbound aggregate bandwidth exceeded the maximum for the instance. | N/A |