On 12.07.21 07:49, stav alfi wrote: > > 1. can histogram_quantile return a value which is different than all the > bucket values? for example, if the buckets are: 1,2,3. can it return 1.7?
Yes, it can. See the documentation at https://prometheus.io/docs/prometheus/latest/querying/functions/#histogram_quantile : "The histogram_quantile() function interpolates quantile values by assuming a linear distribution within a bucket." So if the bucket `le="2"` contains 10 samples, and the seventh happened to coincide with the median (ar whatever you are calculating), histogram_quantile will return 1.7. > 2. in the folllowing screenshot (green=first query - call duration, > yellow=second query - quantile), I'm seeing that the 99% quantile took more > than the actual API call at the same time (look at 11:46 - yellow > (quantile) > green (call duration)). *how can it be?* Your 1st "green" query calculates _average_ duration (mean value of all observed call durations). Your 2nd "yellow" query calculates the 99th percentile duration (99% of the observed calls had a duration shorter than that). It's not surprising that the average is lower than the 99th percentile. -- Björn Rabenstein [PGP-ID] 0x851C3DA17D748D03 [email] [email protected] -- You received this message because you are subscribed to the Google Groups "Prometheus Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/prometheus-users/20210714172821.GE3656%40jahnn.
