[mrtg] Re: MRTG Graphs don't appear to match

[Eric Brander]

Davis, Steve wrote:
> I've been using MRTG for quite a period of time.  However, I'm somewhat
> confussed.  When I look at the weekly graph it shows a max of 25.1 MB/s on
> Monday but when I look at the monthly and yearly graphs they only show max's
> of 20.5 MB/s and 12/4 MB/s respectively. It doesn't appear as though they
> are being updated properly.  Does anyone have a suggestion as to why this is
> so, or am I simply misreading the data.

By default the daily graphs are 5-minute averages, the weekly are 
30-minute averages, the monthly are 2-hour averages, and the yearly are 
1-day averages.

So in your case, Monday for one 30-minute period you got an average 
maximum of 25.1 MB/s.  What do you think the average would be over a 
2-hour period?  How about over the entire day?  Keep in mind that the 
graphs are showing Averages of the readings. That is how the log file 
says small, because as data gets older, it gets consolidated in to an 
average over a longer span.

So, if you had a graph that was a flat line, and exact 25 MB/s 
constantly, for a 1 year period, then the readings for Max would be the 
same for all 4 graphs. Since this isn't likely the case, your Max's for 
each period will differ.

Just in case that's not clear.

Pretend these numbers are the 5-minute readings over 2 hours (broken 
down by 30-minute intervals for clarity):
(2 3 2 1 4 3) (3 6 7 5 2 2) (3 3 1 1 1 2) (2 3 1 2 1 1)

These are then the 30-minute averages
2.5		4.2		1.8		1.6

This would be the 2-hour average:
2.5

If all other readings were 0 (and the withzeroes option is NOT set) the 
following would be what you'd see:

-On the daily graph, the 5-minute average MAX would be 7.

-On the Weekly graph, the 30-minute average MAX would be 4.2

-On the Monthly the 2-hour average MAX would be 2.5

When dealing with rates, you have to do averaging. Even the initial 
polling is simply an average since the last poll. In order to have 
historical data that doesn't fill up your hard drive, you have to do 
averaging of the averaging which is why the MAX readings for each of the 
graph intervals is likely lower.

I hope this helps. Judging by some of the other answers to your question 
either I've misunderstood or they have. So I thought I'd throw in my $0.02.

Regards,

Eric Brander

--
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
WebAdmin    http://www.ee.ethz.ch/~slist/lsg2.cgi