[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Fwd: Re: Virtual Addresses on new server Fixed

For your information if you are using the new kernel and virtual ip's (ip aliasing)

Thanks to Brian Moore, I was able to get rid of invalid netmask/broadcast/etc.... as possible causes.

Turns out the new setup for networking, using /etc/network/interfaces, does not add the host to the routing table. Therefor, each interface must have a separate "up" command to do this. The problem is now fixed due to A) correct network and broadcast numbers (thanks again) and B) by adding the line:

        up      /sbin/route add -host 216.62.82.??? eth1:?

in the /etc/network/interfaces file. I've attached a copy of mine if anyone is planning to upgrade and uses virtual ip's


X-From_: bounce-debian-isp=techinfo=dailydatainc.com@lists.debian.org Wed Feb 21 16:27:53 2001
Resent-Date: Wed, 21 Feb 2001 16:27:52 -0600
X-Envelope-Sender: bem@rom.org
Date: Sat, 19 Aug 2000 09:07:18 -0700
From: brian moore <bem@rom.org>
To: debian-isp@lists.debian.org
Subject: Re: Virtual Addresses on new server
Mail-Followup-To: brian moore <bem@rom.org>, debian-isp@lists.debian.org
User-Agent: Mutt/1.2.5i
X-Uptime: 9:00am  up 56 days, 18:55,  1 user,  load average: 1.00, 1.00, 1.00
X-Moon: The Moon is Waning Gibbous (81% of Full)
X-Fingerprint: 9934 9836 64C9 2B72 C4E7  A1A1 377F CE24 8832 2B51
X-GPG: 1024D/88322B51 at http://www.cmc.net/~bem/bem.asc
X-Vile: yes
X-11: R6.3
X-10: yes
Resent-From: debian-isp@lists.debian.org
X-Mailing-List: <debian-isp@lists.debian.org> archive/latest/4221
X-Loop: debian-isp@lists.debian.org
Resent-Sender: debian-isp-request@lists.debian.org

On Sat, Aug 19, 2000 at 01:45:35AM -0500, R. W. Rodolico wrote:
> I have five IP's with my DSL (very, very small ISP). On my current server,
> I have all five working. However, when setting up a new server, two of the
> five are not visible.
> My range is - 125, with 126 being the gateway. This is
> defined as the last three bits of the range, ie 0111 1xxx. I have set up a
> netmask of .248 (1111 1000), and a broadcast of various values ranging from
> .7 ( 0000 0111), .127 (0111 1111), .135 (1000 0111).
> The addresses that work are 121-123 (0111 1001 - 0111 1011), ie I only get
> in trouble when the third bit is set (goes to 1).
> Any ideas? Obviously, I do not understand broadcast values as I have been
> playing around with them a lot (I also used .255 once). From what I've seen
> and read, I really don't think that is the problem, though.
> All these addresses work just fine on the old box ( 2.0.38 kernel ). I am
> having problems with potato ( 2.2.17 kernel) I have set ipchains to accept
> everything.
> I am attaching /etc/network/interfaces from the new box, and
> /etc/init.d/setvirtual from the old one. Note that the latter is something
> I kludged together.
> Any help would be appreciated.

iface eth1 inet static

Your 'network' is wrong.  It should be

Now, the way these numbers work:

     IP number:    216. 62. 82.121
     Netmask:    (binary 'and')
     Network:      216. 62. 82.120

That means that only the 'leftmost' bits are considered.  Draw it out in
binary (at least the last octet) and you'll see the magic.

The broadcast (at least on modern systems -- there was a period of
confusion in the olden days where broadcast was sometimes the "all 0's"
host instead of the "all 1's") is the network with the "host" portion of
the address set to 1's.

To see the broadcast, invert the netmask to: in this case (ie,
the 0's become 1's) and do a binary 'or' with either the network
(computed above) or a the host... ie, set the proper number of 'low'
bits to 1.

Yes, thinking in binary helps. :)

To UNSUBSCRIBE, email to debian-isp-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org

Attachment: interfaces
Description: Binary data

R. W. "Rod" Rodolico
Daily Data, Inc.
POB 140465 - Dallas  TX  75214-0465
214.827.2170  -- 214.821.5051 (fax)

Entia non sunt multiplicanda, praeter necessitatem
					William of Occam

Reply to: