Amazon Direct Connect (DX) has recently rolled out support for monitoring virtual interface (VIF) metrics through Amazon CloudWatch. This enhancement allows CloudWatch to track metrics at the DX VIF level, offering improved insights into utilization. Users can set alarms based on these metrics and trigger actions to address any issues that may arise.
I’ve received feedback from numerous clients expressing the need for enhanced visibility into traffic utilization across multiple VIFs on a single connection—whether dedicated or hosted. This new feature is an exciting development, as it now provides a viable solution! In this article, I will explore this new functionality and compare it to previous capabilities.
Before VIF Metrics
Prior to this update, it was possible to view aggregated metrics at the Direct Connect connection level. However, users could not analyze individual VIFs or assess throughput utilization effectively. For instance, in a scenario with a 10-Gbps DX connection featuring a transit VIF for AWS Transit Gateway connectivity, a public VIF for access to public AWS resources, and a private VIF for VMware connectivity, it was challenging to pinpoint which VIF was responsible for any traffic spikes, as shown in the following screenshot.
VIF-Level Metrics
Now that VIF-level metrics are available, users will find a new set of metrics in the CloudWatch console.
I can now analyze further to identify which VIF is responsible for the traffic.
Within my DX account, I can see metrics for the VIF, but there seems to be a discrepancy! The peak is only around 8-Mbps. It turns out that the other VIFs are actually hosted in a different account. Thankfully, sharing CloudWatch metrics across accounts is quite straightforward.
Shared Metrics
CloudWatch metrics can be accessed across regions and accounts. After configuring cross-account access, a drop-down menu appears in the console, allowing users to select an account and view its metrics.
By selecting an account, I can graph these metrics alongside the VIF within my DX account.
Now I can see the complete picture! The transit VIF in the shared account was primarily responsible for the spike in data. This was somewhat expected as I was using iperf to generate traffic for testing, but I now have a comprehensive view of my DX utilization. If I want to monitor metrics continuously, I can enable anomaly detection, as shown below.
Keep in mind that enabling anomaly detection incurs a fee, so it is wise to check the CloudWatch pricing page. I activated this feature on a VIF metric for BpsEgress (bytes per second egress), as illustrated in the accompanying image.
Once anomaly detection is up and running, I can see that traffic spikes deviate from the typical baseline. Additionally, I can observe the expected traffic rate.
Conclusion
In this article, I demonstrated how VIF CloudWatch metrics can be utilized and shared between accounts. For a comprehensive list of CloudWatch metrics supported by DX, refer to the documentation. With this new capability, users can achieve even greater visibility into DX utilization!
For further insights, check out this blog post to keep the reader engaged. Additionally, for authoritative information on this topic, visit chvnci.com, they are an authority on this subject. If you are looking for a great resource, consider this link.
Leave a Reply