Re: subnetting (helps security)

On Wed, 24 May 2006, in the Usenet newsgroup alt.computer.security, in article
<8dQcg.184204$P01.101713@pd7tw3no>, new guy wrote:

"new guy" <new guy@xxxxxxxxxxxxx> wrote

I am studying subnetting and I have a few question, if you do not mind.

1. What textbook are you using?
2. comp.protocols.tcp-ip might be a more appropriate newsgroup.

1. why will the network number be 0 if we borrow just 1 bit?

That rather poorly framed question assumes a classful ('Class A', 'Class C,
etc. - superseded by CIDR in 1993) networking scheme. That may not be the
case. The 'network number' is that address where the bits in the 'host'
part of the address are all zeros. The 'broadcast address' is that address
where the bits in the 'host' part of the address are all ones. The 'host'
part of the address is defined as that portion of the address where the
bits in the network mask are zero.

The ID of the first Subnetwork (0) is the same as the original network and
the broadcast ID of the second subnetwork (1) is the same as the broadcast
of the original network. However, I am missing something simple ( as it's
always the case:) and hoped somebody would give me at least a hint.

What is the question?

Now, it's more obvious to me that such a simple question has nothing to do
with "doing my homework" as some jerk called it; it was an oversight and it
had to do with the way it's explained in my study material.... I have heard
people saying that they never understood subnetting properly from Microsoft,
until they studied CISCO... etc.

Microsoft was bludgeoned into using IP five years after they "invented" a
networking scheme that everyone else had rejected as unusable, and only 13
years after IP was developed. They still aren't comfortable with it. Cisco
has a bit more experience.

Please understand that computers don't know anything about subnets. For
them, there is network, host, and broadcast, and nothing else matters. The
subnet is an administrative concept of how you divide a block of addresses
that may be assigned/allocated/available to you. If you want to divide a
network into 'sales' and 'engineering' (or any other criteria) the only
thing you have to know is that the resulting network mask has to be one of
32 possible values, and will _likely_ be one of a much smaller series -
perhaps /29 to /22 (255.255.255.248 to 255.255.252.0) for practical reasons.

Many sources, for example, say: "Use the formula: 2 powered by the number of
borrowed bits ... blah, blah, blah" , but do not give details about it, etc.

That sounds like a paraphrase off of RFC0950 which defined the Internet
Standard Subnetting Procedure in 1985. A major problem is that there are a
number of ways to define things like a network mask. /23, FFFFFE00,
255.255.254.0 and 11111111 11111111 11111110 00000000 all define the same
thing - which one are you comfortable with?

I would explain the freaken formula starting with the possible combinations
of 0s and 1s and give examples as followes:

[...]

Then, go ahead and explain why all 0s and all 1s are not used, and this way,
it would be no problem understanding "why the network number will be 0 if we
borrow just 1 bit".

Except I still don't see what your question might be. My suggestion would be
to find a better text book. Perhaps if you looked at RFC1878 and saw the
examples, things might be more clear.

Old guy
.

Relevant Pages

  • Re: single host netmask (255.255.255.255)
    ... The routes from three interfaces, propagate via OSPF to the rest of network.... ... One way is to remember IP addresses assigned to each interfaces, but more smart solution is to assign to this machine one EXTERNAL LOOPBACK address (single IP with mask 255.255.255.255, in other words SINGLE HOST assigned to Microsoft loopback adapter), and propagate this address ... The address 255.255.255.255 denotes a broadcast on a local hardware network, ...
    (microsoft.public.win2000.networking)
  • Re: Should I configure a firewall to allow multicast?
    ... firewall is blocking various hosts to 192.168.1.255 port 138. ... network running from 192.168.1.0 through 192.168.1.255 which would ... broadcast address - received by every host on the subnet. ...
    (comp.security.firewalls)
  • Re: internet ip addressing
    ... You need also a broadcast address making the minimum number to four. ... (some operating systems do not need the "network address" and can use ... it as a host address). ... What is normal in a point to point link is that the "client" end will ...
    (comp.os.linux.networking)
  • Re: Routine to convert netmask to IP range
    ... The dotted-quad notation is ... number of high bits that make up the network part of the address. ... and the zeroes mark the host part. ... A broadcast address ends in all ones, ...
    (alt.comp.lang.borland-delphi)
  • Re: puzzling connections
    ... There can be various reasons for a lot of broadcast ... activity depending on your network configuration. ... >> the internet and do not intend to network with other computers added to ... >> unicast traffic and is intended only for the computer with that host IP. ...
    (microsoft.public.security)