Why has an anomaly been triggered?

Refer to exhibit:

Why has an anomaly been triggered?
A . Controller reached 2500 IOPS.
B. Observed IOPS exceed normal values.
C. Normal Controller behavior has increased.
D. Observed values do not match predicted values.

View Answer

Answer: D

Explanation:

Nutanix anomaly detection is a feature that monitors seasonal trends for time-series data (e.g. CPU usage, memory usage, latency, etc.) and establishes a “band” of expected values based on historical data12. Only values that hit outside the “band” will trigger an event / alert1.

Nutanix leverages a method for determining the bands called ‘Generalized Extreme Studentized Deviate Test’. A simple way to think about this is similar to a confidence interval where the values are between the lower and upper limits established by the algorithm.

Another web source3 shows an example of how anomaly detection works in Nutanix Prism Central. In the video, you can see that when the observed value of a metric deviates significantly from the predicted value based on historical data, an anomaly event is triggered and displayed on a chart.

Therefore, by comparing the observed values with the predicted values based on historical data, Nutanix anomaly detection can identify abnormal behavior and alert you accordingly.