Question 830
At the Fairhaven customer support provider, an AKS cluster reaches memory pressure but operators see only node CPU and cannot identify failing pods. Which technical or process problem best explains the failure?
Explanation
Correct Answer
The issue is not simply a failed task; node-wide averages can hide individual container resource exhaustion. That is why the practical change is to enable suitable Azure Monitor Container Insights and collect relevant pod and node metrics. This directly addresses the reported issue at the Fairhaven customer support provider.
Incorrect Answers
That statement concerns azure pipelines event alerting; here the reported facts point to a different issue: node-wide averages can hide individual container resource exhaustion. For the Fairhaven customer support provider, the priority is to enable suitable Azure Monitor Container Insights and collect relevant pod and node metrics.
Sampled data may not represent literal unsampled request counts without adjustment. This can occur in other projects, but it is not the primary explanation at the Fairhaven customer support provider.
This explanation belongs to distributed application insights tracing. The immediate weakness in the question is that node-wide averages can hide individual container resource exhaustion. For the Fairhaven customer support provider, the priority is to enable suitable Azure Monitor Container Insights and collect relevant pod and node metrics.