Question:
How do we report the issues encountered? Example, If a device has exceeded the threshold for CPU utilization (I.e., > 90%). How is this calculated? When we see > 90% between polls? > 90% over the course of one minute, etc?
Answer:
The platform looks at all performance data collected, stacks the records in ascending order, then looks at the 95th value. So we are seeing the CPU at 'X' percentage 95% of the time
so every 5 minute poll includes a new data set to add to that analytic check.
