Docs » Integrations Guide » Monitor Kubernetes » Example: Finding the Root Cause of a Problem

Example: Finding the Root Cause of a Problem 🔗

Finding root causes of issues in a distributed microservices environment is often a complex and arduous task. For example, a problem you are experiencing might be caused by a deployment on a service that is a couple of hops away from you, or it could be a noisy neighbor eating away at the CPU.

The following example illustrates one way in which SignalFx and the Kubernetes Navigator can help you quickly pinpoint underlying issues that otherwise would take hours to find.

The following image shows a simple e-commerce application. Imagine that someone at the company deploys some new code. Within seconds of a problem starting, you receive an active alert that the application is slowing down.


It is a critical alert indicating that the transaction latency is too high. Click on the alert icon on the dashboard to open the alert details page.


The most important thing is to identify the underlying culprit and restore the health of your system. Use the Kubernetes Navigator to look at the correlated infrastructure and guide you to the root cause more quickly.


Pods that match the filter are highlighted, allowing you to drill down further.


Click the container. The Info tab in the sidebar gives you detailed information, including what resource limits it is using.


In this case, the container has no memory limit. That is probably why it is using all of the memory on this node.


Now that we know which container is causing the problem, we can learn which workload it belongs to and we can get in touch with the right people to get it fixed. So, we solved the mystery of what was going wrong with our service — a noisy neighbor that got scheduled to a downstream service caused our API service to send an alert.