I don't use netview to do it, but I read the stats out of about 2300
catalyst ports (5000s, 5500s, 1200s, 2900s) and they all look reasonable (I
do extensive analysis on the data). The problem you have to look out for is
whether you are reading them often enough. They will wrap fairly quickly if
it is a high speed port. I currently read mine out at 15 minute intervals,
and carefully watch the traffic levels to watch for when I have to increase
the frequency of the readout. Gigabit ports need to be read more frequently
than 100 mbit ports, and 100 mbit ports need to be read more frequently than
10 Mbit ports. Also, bear in mind that you are reading counters that wrap,
and you need to do the math carefully to allow for the wrap.
Also, if you have engineers that insist on clearing the stats via telnet,
that will foul up your numbers, and it also clears the snmp values. You
also have to recognize invalid data, as sometimes the readout glitches (not
too often) and you get garbage.
Hope this helps...it is not as simple as it initially seems.
At 05:04 PM 1/25/99 -0700, you wrote:
>Further to my note re: collecting data on cisco 2916/24 switches last week,
>does anyone know if ifInOctets/ifOutOctets on a switch (any switch!)
>represent the actual traffic on these ports?
>The numbers I am getting from snmpCollect sure don't look real!
>And graphing in real time (from the mib browser) doesn't look any better.
>My gut feeling is that the numbers are misleading, in that they don't
>represent the traffic, but some aggregate count???
>Any advise would be greatly appreciated!
>
>--
>/in haste
>/Bob
>
*********************************************************************
" Of course the opinions expressed here are my own. "
Connie Logg CAL@SLAC.Stanford.Edu ph: 650-926-2879
Network Management and Performance Analyst
SLAC (MS 97), P.O. Box 4349, Stanford, CA 94309
"Happiness is found along the way, not at the end of the road."
|