On 01.09.21 15:47, Sakthi Raam wrote:
>
> I would like to seek some advice on how to calculate response time (latency
> SLI) based on a gauge metric. I don’t have a histogram bucket.
>
> Kafka_network_requestmetrics_totaltimems{request=“”,quantile=“0.50”}
>
> Above is the time series of interest which records the response time( in
> milliseconds) in quantile.
This looks more like a pre-calculated quantile, which is part of a
summary (not a gauge). Although you could argue, the quantile itself
is a gauge. So perhaps that's what you mean. Just saying to avoid
misunderstandings.
> I basically want to get the percentage of request that got processed within
> x seconds
To answer that question, you need a histogram (with x seconds as a
bucket boundary).
> and also want to find the percentage of request that got processed
> above x seconds for a given quantile (99%- This is available as part
> of the metric)
This question sounds like you have mixed up two questions into
ane. The 99th percentily (AKA quantile="0.99") is the duration after
which 99% of requests finished processing, and 1% of requests took
longer than this duration.
--
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/20210902161853.GQ1307%40jahnn.
