[ previous ] [ next ] [ threads ]
 From:  "Mitch \(WebCob\)" <mitch at webcob dot com>
 To:  "Manuel Kasper" <mk at neon1 dot net>, "Christopher M. Iarocci" <iarocci at eastendsc dot com>
 Cc:  m0n0wall at lists dot m0n0 dot ch
 Subject:  RE: [m0n0wall] Something odd in status page
 Date:  Sat, 29 Nov 2003 11:11:39 -0800
Same problem exists for most SNMP counters. 32 bt counters are standard, and
with a 100Mb feed, you have to check your counters every 5 minutes and log
them for difference checking so you can detect the roll-overs. 1Gb SNMP
typicalls supports 64 bit counters, but that's not the norm for most of us.
I suspect this is why most MRTG polls run every 5 minutes.

I have a PDF I found somewhere that discusses dealing with counter roll and
so on - can send it to anyone who is interested, but basically the theory is
check every 5 minutes, subtract the previous value unless the previous value
is smaller which indicates a counter roll has occurred, if so, add 4G to
your current number and try the subtraction again.

Make sense?

Sure it's a limitation of FreeBSD, but that seems to exist elsewhere and is
how others seem to deal with it.

My two cents.


-----Original Message-----
From: Manuel Kasper [mailto:mk at neon1 dot net]
Sent: Saturday, November 29, 2003 1:21 AM
To: Christopher M. Iarocci
Cc: m0n0wall at lists dot m0n0 dot ch
Subject: Re: [m0n0wall] Something odd in status page

On 29.11.2003, at 06:41, Christopher M. Iarocci wrote:

> I noticed that when your total traffic over an interface reaches 4GB,
> it resets to zero when looking at the status page for the interfaces.
> Is that by design, or did I happen to notice something wrong?

No - seems like FreeBSD only uses 32-bit integers to store the number
of bytes received/sent through an interface. Or is it only netstat
(which is what m0n0wall uses to get the stats)?

- Manuel

To unsubscribe, e-mail: m0n0wall dash unsubscribe at lists dot m0n0 dot ch
For additional commands, e-mail: m0n0wall dash help at lists dot m0n0 dot ch