[mrtg] Re: Unexpected spikes in 64 bit graph

[Rich Adamson]

> I have recently upgraded a graph from 32 bit to 64 bit counters, and am now
> getting unexpected (and I think incorrect) spikes in the graphs. All ideas
> welcome.
> I have a network with multiple VLANs. A layer3 wirespeed switch (an old
> 3Com Coebuilder 3500) is attached with a single Gigabit interface.

I can't speak to this exact model, but historically 3Com has had more then
its share of snmp problems over the years. I'd suggest validating the counter
values by sending a known volume of traffic through the interfaces at o-dark-
thirty and comparing that to what's being reported via snmp.

Also, a fair number of switch vendors do not report accurate vlan traffic
statistics. You might attempt to validate both.

Mrtg also had a problem around v2.9.19 (several minor version numbers had it)
that caused a spike whenever one or more polls were missed (outbound poll
was sent, but no response received). Since snmp polling relies on udp packets,
its up to the application (mrtg polling) to determine whether a packet was
missed and how to handle it. The problem was corrected somewhere around the
v2.9.22 (I don't know the exact version).



--
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