[mrtg] Re: clearing of counters on Cisco - again
Jim Witherell
jwitherell at us.ibm.com
Mon Jan 27 18:59:06 MET 2003
That makes more sense. Good tip, too, as I was thinking of moving to
RRDtool at some point (when I get more sophisticated). What is the reason
for your NMS doing the counter reset? Is it a scheduled thing, or is it
triggered by other conditions? If it is scheduled, maybe the thing to do is
just expect to see the spike at a certain point on the graph, or move the
reset event to, maybe at midnight so it is easy to spot as a non-event.
Jim Witherell
Network Specialist --Live Intentionally--
IBM @ AK Steel Corp
513.425.3483
"Scott P.
Daffron" To: mrtg at list.ee.ethz.ch
<scott at cs.odu.edu cc:
> Subject: [mrtg] clearing of counters on Cisco - again
Sent by:
mrtg-bounce at list.
ee.ethz.ch
01/27/03 12:25 PM
After spending a couple of days on this now, I have determined that MRTG,
when used alone, does not show this behavior. The problem is only with
MRTG + RRDTOOL + 14all.cgi. I've even downloaded and set up the newest
versions of each only to get the same result. Does anyone how I can have
RRDTOOL and 14all.cgi follow the same logic as MRTG when creating graphs
(eliminating this spike)?
More info on my setup:
- Switches are Cisco Cat 4006 and 5XXX
- Our NMS is resetting all port counters by doing an snmpset on
private.1.9.5.1.1.13.0. Most people do not believe me, but this DOES
reset the ifOctets value to 0, by design.
---------- Forwarded message ----------
Date: Fri, 24 Jan 2003 11:25:47 -0500 (EST)
From: Scott P. Daffron <scott at cs.odu.edu>
To: mrtg at list.ee.ethz.ch
Subject: [mrtg] clearing of counters on Cisco
Oh most wise MRTG gurus,
We have noticed that our version of MRTG (a couple years old) shows a huge
spike on our Cisco interface graphs whenever a "clear counters" is done on
the device. This is becoming a big problem for us, as we have just
implemented a new NMS which clears all counters once per day.
I know this is an issue of the OID counter wrapping, leading MRTG to think
that a massive bandwidth spike has indeed occurred, but is there a way to
configure MRTG not to draw this conclusion (and alleviate all of the
massive spikes in our graphs)? Does a newer version of MRTG handle this
differently? I have looked high and low, but can't seem to find an answer
on this. Thanks in advance...
Scott Daffron
Network Engineer
Sentara Health Care
--
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
--
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
--
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
More information about the mrtg
mailing list