[mrtg] Re: Accuracy of data

[PAUL WILLIAMSON]

MRTG reports the data it is given.  I have had this discussion with 
so many people where I work that I have actually written a little 
blurb about how mrtg collects the data and why it is not wrong.
If anything, we've found more instances where the device 
is reporting incorrectly that we've gotten a few bugs resolved 
with Cisco et al.

3MB - that would be 3 megabytes per second?  Are you using 
32 bit counters or 64?  I realize that probably would not 
have a rollover factor involved.  

If you doubt it, write a script that grabs the counters every 30 
seconds and put it into a text file.  Then compare that to the 
data MRTG has grabbed.  I've done that on a number of 
occasions as well to detect interfaces that are getting high 
bursts of data.  Once we find those, we ratchet down our 
interval to 1 minute.  1 minute is generally a small enough 
slice of time to detect the bursts we experience.

Remember, the data you see is an average throughput 
during a 5 minute window.  A whole lot of data can 
pass through an interface in that 5 minutes, and the 
spikes experienced within that timeframe can be 
quite large.

Paul

>>> "Paul Harlow" <pharlow at skld.com> 06/03/05 11:57 AM >>>
I have had MRTG up for quite a while not and am a bit uneasy as to how
accurate MRTG really is. It seems to me that we should actually be
using
more bandwidth than what MRTG is reporting. The only hard example that
I
have is the Firebox 2500 that I have behind a router that I have
attached to the Internet. This router hosts a multi linked T1 circuit
with two T1 lines for roughly 3MB worth of bandwidth. The bandwidth
monitor on the Firebox shows an average between 300k and 500k an many
times there are a few minutes worth of 1mb+ traffic spikes. Despite
that, MRTG shows less than 100k on average as far as traffic is
concerned. Why is this?
 
Paul Harlow
System and Network Administrator
SKLD Information Services LLC
720 S. Colorado Blvd. Ste1000N
Denver, CO   80246
303.820.0861 - desk
303.913.2804 - cell
 
--
Unsubscribe mailto:mrtg-request at list.ee.ethz.ch?subject=unsubscribe 
Archive     http://lists.ee.ethz.ch/mrtg 
FAQ         http://faq.mrtg.org    Homepage     http://www.mrtg.org 
WebAdmin    http://lists.ee.ethz.ch/lsg2.cgi 



--
Unsubscribe mailto:mrtg-request at list.ee.ethz.ch?subject=unsubscribe
Archive     http://lists.ee.ethz.ch/mrtg
FAQ         http://faq.mrtg.org    Homepage     http://www.mrtg.org
WebAdmin    http://lists.ee.ethz.ch/lsg2.cgi