[rrd-users] Re: limiting hight of a graph...

Palmans Pepijn pepijn at kangaroot.net
Thu Oct 26 01:44:30 MEST 2000


What I was actually trying to do is to limit the peeking of the graphs to
destroy the rest of the data.

the problem in this case still stays a little, since I don't know exactly
what will be peeks... Depending on howmany clients are connected to the
line forinstance.  If i have one client with 100K of bandwith, a peek of
1000K will destroy my graph for example.  but if I have 10 of those
users 1000K is ok and only 100.000K will do some harm (yeah, the numbers
aint that correct :)

I suppose I can change the first 1000 in the line below (CDEF as quoted)
into some function to take the average bandwith + 10% for instance and all
others should be peeked off at avg bandwith + 15% to show they are peeks.

bit late for such stuff, if you know the direct answer and you don't have
to spend to much time, feel free to post.  Otherwise I'll start digging
into the docu's again tomorrow :)

kind regards,

pepijn palmans

On Thu, 26 Oct 2000, Alex van den Bogaerdt wrote:

> 
> Palmans Pepijn wrote:
> 
> > with the CDEF:blah=inoctets,0,1000,LIMIT I can limit the graph to only
> > display the values in between 0 and 1000.  The sad thing is that if the
> > line is rather thing (meaning just one value is high) you don't see that
> > it isn't there.  In other words, you don't notice the peek.
> > 
> > is there a way to make LIMIT still draw the value but only up to the
> > limited value so that you see a peek, but just not that high as it would
> > be otherwise ?
> 
> Something like
> 
> CDEF:blah=inoctets,1000,GT,1000,inoctets,IF
> 
> Then graph blah and do your calculations with inoctets (for correct
> averages, maxima, minima etcetera.


--
Unsubscribe mailto:rrd-users-request at list.ee.ethz.ch?subject=unsubscribe
Help        mailto:rrd-users-request at list.ee.ethz.ch?subject=help
Archive     http://www.ee.ethz.ch/~slist/rrd-users
WebAdmin    http://www.ee.ethz.ch/~slist/lsg2.cgi



More information about the rrd-users mailing list