Question 860
At the Westwick financial cooperative, an API is slow only when a downstream function calls an external billing gateway. How should the team reduce the chance of the issue recurring?
Explanation
Correct Answer
The continuing risk is that aggregate CPU and HTTP counts cannot reveal the slow step in a distributed request path. To sustain the correction, maintain distributed tracing coverage through new application integrations. This directly addresses the reported issue at the Westwick financial cooperative.
Incorrect Answers
That practice helps control a different risk: CPU-only monitoring can misdiagnose instability caused by memory pressure. The recurring control here should be to maintain distributed tracing coverage through new application integrations. For the Westwick financial cooperative, the priority is to analyze Application Insights distributed traces and dependency spans to localize latency.
The alternative suggests that the team monitor Container Insights coverage after cluster and namespace changes; it would not directly prevent the observed issue from returning. For the Westwick financial cooperative, the priority is to analyze Application Insights distributed traces and dependency spans to localize latency.
This maintenance work is reasonable for application insights telemetry. The continuing exposure described here is that aggregate CPU and HTTP counts cannot reveal the slow step in a distributed request path. For the Westwick financial cooperative, the priority is to analyze Application Insights distributed traces and dependency spans to localize latency.