Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 30 Jul 2001 23:23:08 -0700
From:      "Ted Mittelstaedt" <tedm@toybox.placo.com>
To:        "kesu" <kesu@kesuki.dyndns.org>
Cc:        <freebsd-questions@FreeBSD.ORG>
Subject:   RE: Need help limiting bandwith ARP uses over cable modem.
Message-ID:  <000801c11989$444d11c0$1401a8c0@tedm.placo.com>
In-Reply-To: <20010730183920.S14350-100000@kesuki.dyndns.org>

next in thread | previous in thread | raw e-mail | index | archive | help
>-----Original Message-----
>From: owner-freebsd-questions@FreeBSD.ORG
>[mailto:owner-freebsd-questions@FreeBSD.ORG]On Behalf Of kesu
>Sent: Monday, July 30, 2001 5:34 PM
>To: Ted Mittelstaedt
>> >
>> >My incoming bandwith is 560 killobits/second my outgoing bandwith is 128
>> >killobits/second now surely you can see how receieving a saturating queue
>> >of packets on the inbound is going to generate _more_ outbound packets
>> >than my bandwith limitation.
>>
>> No, actually I cannot.  If your RECIEVING a larger volume of data than your
>> sending then while you may be generating a lot of packets, those are SMALL
>> packets.
>
>I have anylized the situation, and in fact the ARP packets are causing the
>problem.  I can use the output of tcpdump to accuratle measure the
>bandwith used by ARP packets and it easily exceeds my tiny 16-KilloBytes
>second upstream cap.

Wait just a second here.  In your previous posting you said your upstream
bandwidth is 128k.  Now your saying it's 16k.

If you have a 560k downstream and a 16k upstream then I can well believe your
having a problem.

>
>The packet loss becomes user noticable when the modem resets. (as in

Obviously.  But, how often is it resetting?

>
>But the ARP packets are pointless, as it isn't a lan, and version 1.1
>causes the ARP to be selective, only sharing between local computers when
>a connection attempt is actually made between the two. DOCSIS 1.1
>compliant modems are going to make cable modems a much more user friendly
>experience, since the reduction in ARP traffic will benefit everyone.
>

I'll believe it when I see it.  The IP-over-Cable people have been
really going through some rediculous gyrations to try to get piss-poor and
ignorant networking design to work on oversubscribed segments.

You said earlier that your provider runs a full /24 subnet, ie: 254 machines
per network.  You also said your downstream is 500K.  That means that your
provider is pushing 127 Megabits into just your segment, right?  Is my
multiplication wrong?  Do you see a problem here?  I do.

>ise _extremely_ annoying.  Since the modem has a tendancy to reset
>around once a week, this is very annoying to me.
>

Your upset over a 2 minute network outage that happens ONCE A WEEK?

>crash on me every day (windows.)  Furthermore, why did you assume I use
>@home?

Because frankly I cannot imagine a small ISP to be this stupid.  @Home is
a different story, I fully expect them to be this stupid.

You already said the solution to your problem earlier although you probably
missed it.  That is, your subnet mask is too big.  This is something
that your ISP should have done by subnetting instead of throwing everyone on a
single /24.  If they had cut up that /24 into 4 /26's then you wouldn't be
seeing the kind of ARP traffic your seeing.

This does lead into a trick that you can _try_ there is no guarentee that it
will work, though.  That is you can still adjust your own mask.  Let me
explain:

Suppose that your ISP's gateway is 192.168.1.1 with a subnet mask of
255.255.255.0

Your IP numbers are 192.168.1.45,  192.168.1.46, and 192.168.1.47 and you of
course have a subnet mask of 255.255.255.0

Now, when your system gets an arp request from, say, 192.168.1.156, it knows
that this number is on it's network.  So, it responds.

However, what happens if you change your OWN subnet mask to 255.255.255.192

Well, your system will still think that the gateway at 192.168.1.1 is on it's
subnet, and thus it can still reach it.  It will still think that 192.168.1.46
is on it subnet and can reach that.

But, it will think that packets from all IP numbers from
192.168.1.65-192.168.1.255 are NOT on it's own subnet, and thus it should not
respond to them with a broadcast.

Now, if your doing network-style broadcasts then this may not work because
your ISP's gateway may not pick up your own ARP's and thus time you out.  But,
if it does it's a crude but effective hack that will get your machine to
ignore the majority of traffic on the subnet.

Frankly your ISP should have been slapped for putting an entire /24 on a
shared media like this, but what's done is done.

Good luck,


Ted Mittelstaedt                                       tedm@toybox.placo.com
Author of:                           The FreeBSD Corporate Networker's Guide
Book website:                          http://www.freebsd-corp-net-guide.com




To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-questions" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?000801c11989$444d11c0$1401a8c0>