[mrtg] Re: MRTG with RRDTool, report making?

[Rainer Bawidamann]

In article <EJEOJJDLAPOCFPMNBKNPAEBJCAAA.mluebke at resourcenetworks.com>,
	mluebke at resourcenetworks.com ("Matt Luebke") writes:

> I would like to be able to pull data from the .rrd files to make a
> textual report to present to customers.

> Monthly traffic average, monthly max, monthly minimum.
> 
> w% of traffic was in the 0-50% range
> x% of traffic was in the 50-70% range
> y% of traffic was in the 70-90% range
> z% of traffic was in the 90-100% range

> rrdtool fetch 10.1.1.1.1.rrd **** --start 957153600 --end 959832000

> I know the formula I will use to compute the values, but the thing I
> am missing is the **** from the command line. What value can I use
> here to get the actual samples from MRTG?

rrd stores data in several "resolutions": The original samples are
stored for a short time (in the mrtg/14all/rrdtool case 50 hours). Older
values are stored "consolidated", i.e. rrdtool calculates a
maximum/minimum/average of the samples of a configured interval (e.g.
30 minutes) and stores this computed value. This gives the following
picture:

every sample (mrtg default: every 5 minutes) - for two days
one (computed) value for every 30 minutes    - for two weeks
one value for every 2 hours                  - for two months
one value for every 1 day                    - for two years

So you can get the samples only for the last two days. With mrtg/14all
older values are available as maximum or as average of a (longer)
interval. So your "in the ... range" values will only be approximations.
But you can get the correct "monthly max" and "monthly average" values.
You cannot get a "monthly min" value from the mrtg/14all rrd files.

If you want to get the actual samples for a longer period you could
create the rrd files "by hand" with appropriate archives (look into
rrdtool/contrib/log2rrd for an example). Additionally you could add MIN
archives to get the monthly minimum.


To your question: ***** can be "MAX" or "AVERAGE". If you fetch the
samples of the last two days this doesn't matter. For the other cases
average might be more correct.


I hope I could help you a little bit.


Rainer

--
Unsubscribe mailto:mrtg-request at list.ee.ethz.ch?subject=unsubscribe
Archive     http://www.ee.ethz.ch/~slist/mrtg
FAQ         http://faq.mrtg.org    Homepage     http://www.mrtg.org