How much data is needed for Metrics Advisor to start 'Smart detection'?
I did onboard data like this
- Granularity : Minutely
- Ingest data since : 2021-08-05T00:0:00Z
And I got this message
For Minutely metrics, smart anomaly detection will start when the service get data of 2021-09-02T00:00:00Z. To get the result of the latest data 2021-08-09T06:59:00Z, you may try a start time earlier than 2021-07-12T06:59:00Z
I referred to docs below.
(https://video2.skills-academy.com/ko-kr/azure/applied-ai-services/metrics-advisor/faq#how-much-data-is-needed-for-metrics-advisor-to-start-anomaly-detection-)
The number of data required to use the 'smart detection' is different from the information shown in the docs above.
When the granularity was 1 minute, the recommended amount of data required for anomaly detection was 4 days (according to docs). However, it seems that about 28 days of data are needed for smart detection. Is there another standard for smart detection? I can't find it in the docs, so I'm asking for it. Thank you.