[mrtg] Re: Graph scale? ...

[Darren Henderson]

On Mon, 23 Apr 2001, Perry, Duane  wrote:

> MaxBytes[interface999]: 210000  to the ones you want to be change.  I like
> consistent scaling to make it easier to spot problems but it seems strange
> that you want to ignore spikes.  I thought one of the reasons for MRTG was
> to see the spikes and deal with the problem (hacked box, bad NIC).  

There are situations where you might want an indication that there was a
peak but don't neccessarily need to see what it is as the peaks send the
normal volumes into the base line of the graph rendering them useless.

I have several boxes that have fairly stable bandwidth usage that I simply
can't see as they chew up a huge ammount of bandwidth when they are
backed-up each night. The result is I have a graph that has a single spike
during the back up times and practically nothing else. That information
isn't useful to me, I would rather clip the graph of the spike at a certain
point so normal fluctations are visible.

An alternative to clipping would be to generate logarithmic scale, wouldn't 
be my first choice but it would surfice.

I've never found a satisfactory way of dealing with it with the tools
included. I've tried recently to switch to rdtool so I can take a whack at
the graphs myself but mrtg really doesn't seem to want to deal with log file
conversions correctly and you apparently can't start from scratch using
rrdtool (was most frustraiting...I must be missing something there, need to
take another look at it when I get the chance).


-Darren

________________________________________________________________________
Darren Henderson                                  darren at bmv.state.me.us
                                            darren.henderson at state.me.us


--
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