CODEWITHSUNDEEP
newrelic
user1745356
user1745356

Reputation: 4673

How is "requests per minute" calculated in New Relic?

Specifically I'm trying to understand why the number of requests per minute is not a whole number for a given minute in New Relic. I can pick a specific minute during the runtime of my app in New Relic dashboard. For the sake of example, let's say it's the minute between 10:34AM and 10:35AM. The RPM value is not necessarily a whole number. Why?

Upvotes: 0

Views: 2209

Answers (1)

David Marchelya
David Marchelya

Reputation: 3432

Throughout (RPM) is a rate, and may not result in a whole number due to sampling. Call Count on the other hand can be observed as a whole number, which might be the metric that you were expecting.

You can observe these behaviors yourself for the Requests Per Minute metric and the Call Count metric using New Relic's Metric Explorer (Data Explorer). For a given Application, find a WebTransaction Metric and compare the Call Count (whole number) to Requests Per Minute (decimal).

Possible Explanation (Additional Info)

As per: https://discuss.newrelic.com/t/throughput-rpm/498/2

rpm = requests per minute

This is the number of requests per minute that the app receives from users. A request is the entire HTTP request beginning to end.

This indicates that one possible reason that RPM would not be a whole number may be because a request (or requests) started, but did not complete, in the given interval.

Upvotes: 3

Related Questions