[mrtg] Re: problem with statistic values

Larry Sheldon lsheldon at creighton.edu
Tue Sep 14 00:18:23 MEST 1999

> > > The values
> > >  - avarage in / out
> > >  - max in / out
> > > in all 4 graphics were not correctly updated and "loose" old values.
> > Unless the traffic is a flat line at the higher value, I would expect the
> > kind of thing you report.
> > Don't forget, the top graph is five-minute averages, the next one down
> > is thirty-minute averages of the five-minute averages, and so on.
> > Turn on the WithPeak option to see what the biggest five-minute average was
> > in the thirty-minute average.
> maybe I don't understand it correctly, but I had no problems with the old mrtg
> release I used before (2.5.x), and the graphics and statistic values were all
> correct during the past month.

I'm not a bit sure that I understand things in general here or what you are
seeing in particular, but . . . .

> As I understand it, the "max" value displays the general maximum used
> bandwidth, NOT directly for the past 5 or 30 minutes - for the time period of
> the past 24-hours, past week, past month and past year.
> All 4 graphics should display the used ammount of bandwidth during the
> specified time (day, week, ....) and NOT loose the values, especially not loose
> the max and avarage!

One of the things I think I am sure of is that the minimum time-quantum in
the graphs is five-minutes (assuming you have not changed that--and having
actually thought that and typed it and seen it in writing.......)

I'll finish that thought in a minute--got to deal with a high priority 
interupt here--given the large numbers I think you were talking about--is
it possible that in the old one you had modified the code for intervals less
than five minutes and not carried that change forwaerd to the new?

Anyway--assuming a vanilla MRTG, the smallest period recorded is five-minutes,
so the biggest number you can see is the five-minute average for the busiest

In the next graph down, six adjacent 5-minute averages are averaged--the result
will be something equal to or less than the biggest 5-minute--much less if
the biggest is not pretty typical of the other five.

> I used mrtg since >12 month and it was a very usefull tool. I could check all
> days / weeks and see how much bandwidth was used (thanks to backup.pl from
> contrib) everytime, see how much bytes in and out, and specify the time this
> happened. NO value was lost or overwritten. ;-)
> And I could solve some basic problems by noticing a very high bandwidth using
> during the night shift or weekend. (such would be unimpossible now!)
> All this is no longer possible. The graphic bars and statistic values were
> changed often, also max in/out and average in/out values changed (to lower
> values).
> Past friday I created a high bandwidth usage, which was displayed by mrtg in
> the graphic bars and statistic values - today I can't find any notice about
> this high peak into the week / month / year graphics. It happened really and
> was stored into the files, but where is it now ?? Why was it overwritten??
> Such didn't happen with the old release.
> I am really confused, but don't want to downgrade without understanding what
> is going wrong. Or update my understanding how mrtg works. ;-)
> And I will try the WithPeak option to see if this will change anything.

As I said, I'm not a bit sure I understand the situation--are the graphs
publically visible?  Where?

I sure wish Tobi or somebody else who actually knows what they are talking
about would jump in here.
.                                                                       .
- L. F. (Larry) Sheldon, Jr.                                            -
. Unix Systems and Network Administration                               .
- Creighton University Computer Center-Old Gym                          -
. 2500 California Plaza                                                 .
- Omaha, Nebraska, U.S.A.  68178       Two identifying characterstics   -
. lsheldon at creighton.edu                  of System Administrators:     .
- 402 280-2254 (work)                Infallibility, and the ability to  -
. 402 681-4726 (cellular)               learn from their mistakes.      .
- 402 332-4622 (residence)                                              -
. http://www.creighton.edu/~lsheldon    Adapted from Stephen Pinker     .

* To unsubscribe from the mrtg mailing list, send a message with the
  subject: unsubscribe to mrtg-request at list.ee.ethz.ch
* The mailing list archive is at http://www.ee.ethz.ch/~slist/mrtg

More information about the mrtg mailing list