# [mrtg] Re: Ds value for CRC, Resets and Reliability

Janet Post j.post at sparton.ca
Wed Dec 17 20:06:55 MET 2003

```I monitor everything but the Reliablility of the router, (what is the
OID?  I think I would like to watch this!)

I use the Gauge option like Alex van den Bogaerdt suggested.  I have my
MAX set to some ridiculously low number: like 50 or 100.  But then I set
the ABSMAX to some ridiculously high number like 500,000.  This means
that you will see activity on the graph right away, and the graph will
scale nicely if the errors start piling up.  The graph will continue to
rise until you reset your routers.

I've thought about using THRESHOLD to alert me when one of these
counters goes above 5 or 10, but I haven't played with it yet.  If
anyone out there has done this,  please let me know.

Janet Post

>>> Alex van den Bogaerdt <alex at ergens.op.het.net> 12/16/03 03:26 AM >>>
On Mon, Dec 15, 2003 at 09:18:16PM -0800, Chitman Kaur wrote:
> Hi All
> I want to moniter the CRC , Resets and Reliability of router..... I
have got the OIDs and its getting the data but the problem is that I am
not getting wether to give them as GAUGE, COUNTER or DERIVE and what to
give as MIN and MAX .......

GAUGE:   read the number from the device and use this as the rate.
COUNTER: read the number, subtract the previous number, divide by
time, use this as the rate.

The difference between COUNTER and DERIVE is that DERIVE can show
a negative rate whereas COUNTER will only show positive rates.

Example: a 5-digit counter will go from 99999 to 99998.   DERIVE
will see a counter that counted back by 1, COUNTER will see a
counter that increased 99999 (and wrapped).  In both cases the
resulting change is then divided by the time lapsed.

(Note: COUNTER will work with 32 bit and 64 bit counters only.
The example won't work, this was just to explain the mechanism).

Reliability will be a number, meaning something like "in the past
five minutes, the reliability was 255 out of 255".  Use GAUGE for this.
This will make MRTG show "255" (when all is right).

The number of Resets will be low.  One reset every five minutes is
highly unlikely.  OTOH using a GAUGE here isn't good as well.  _If_
visible at all, you will have to spot the line going up by one.
Probably the best you can do is to use COUNTER together with the
PerHour option.

The number of CRC errors _should_ be low.  Low enough so that the
same behaviour is seen as with Resets.  This means you should use
the same approach.  If the number is higher, you may want to change
PerHour into PerMinute, or maybe drop the option altogether.

HTH
Alex

--
http://www.googlism.com/index.htm?ism=alex+van+den+bogaerdt&type=1
This message was produced without any <iframe tags

--
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