graphite-dev team mailing list archive

[gyre007 <question218688@xxxxxxxxxxxxxxxxxxxxx>]

New question #218688 on Graphite:
https://answers.launchpad.net/graphite/+question/218688

Hi there,
I have started playing around with Graphite recently and I noticed one strange behaviour I can't really understand.
At the moment I have only one storage schema. It looks like this:

[stats]
priority = 100
pattern = ^stats\..*
retentions = 10s:12h,60s:7d,10m:5y

I played around with displaying rates of requests per second on idle server - with literally almost NO requests per second, I tried to graph the rates like this:

derivative(stats.app01.live.staging.nginx.handles)

What I noticed is the following behaviour: when I graph these values for past 24 hours down to 12 hours I get a reasonable results (which is 70req/min -> 1.16req/s) - this is awesome, HOWEVER when I try to graph the same thing for period < 12 hours (this is equal to the first retention time) I get completely nonsense values -> something like 12req/min -> 0.2req/s. The rate should be exactly the same as reported when the graphing period > 12 hours (I can say that based on the access logs).

I don't understand the discrepancy between these graphs at all but I reckon that it has something to do with the data retention settings ? I'm using version 0.9.10 and sending the samples every 10 seconds.
Not sure if this is a bug or something I'm not clear on.

-- 
You received this question notification because you are a member of
graphite-dev, which is an answer contact for Graphite.