graphite-dev team mailing list archive

Thread
Date

[Question #114625]: bandwidth graphing issues

To: graphite-dev@xxxxxxxxxxxxxxxxxxx
From: d mckay <question114625@xxxxxxxxxxxxxxxxxxxxx>
Date: Mon, 14 Jun 2010 22:11:20 -0000
Reply-to: question114625@xxxxxxxxxxxxxxxxxxxxx
Sender: bounces@xxxxxxxxxxxxx

New question #114625 on Graphite:
https://answers.launchpad.net/graphite/+question/114625

I'd been sending raw counters for the ports. It seems like to get rates you'd need to wrap each port in some serious levels of 

scale(nonNegativeDerivative(path.to.port), some_crazy_factor)

to get the kind of data you'd want/need from this.

Curious about if that's how it's done, or if one collects rate data instead of counter data, and strategies for organizing network data in the dashboards. I'm happily tracking lots of application metrics, but the network tracking side was feeling a little squarepeg/roundhole.

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