Archived from groups: microsoft.public.broadbandnet.hardware (
More info?)
Ken you forget the a network & node can not be all zeros or all ones.
That is why the equation of (2^n) - 2 = number of hosts or subnets
(where n is in the number of bits remaining or stolen) should be used.
Thus making a 255.255.255.0 subnet only have 254 possible nodes.
for example we take the following decimal information and convert it to
binary so that a computer understands it.
===============================================================
All the numbers in Decimal format.
IP = 192.168. 2 .9
subnet = 255.255.255.0
===============================================================
All the numbers in Binary format.
IP = 11000000.10101000.00000010.00001001
subnet = 11111111.11111111.11111111.00000000
===============================================================
This means that "11000000.10101000.00000010" is the network & "00001001"
is the node on that network.
Once again a network & also a Node can not be all zeros or all ones.
Another example this time a Class A Address with the default subnet.
===============================================================
All the numbers in Decimal format.
IP = 10.0.10.9
subnet = 255.0.0.0
===============================================================
All the numbers in Binary format.
IP = 00001010.00000000.00001010.00001001
subnet = 11111111.00000000.00000000.00000000
===============================================================
This means that "00001010" is the network & "00000000.00001010.00001001"
is the node on that network.
Once again a network & also a Node can not be all zeros or all ones.
Ken wrote:
> Cool. What I'm trying to say is the DHCP server-
> assignable pool of possible IPs for Class A IPs
> (10.*.*.*) is the last THREE octets. For Class C it is
> only the last (192.168.2.*). The subnet, in other words,
> can be described within three octets for Class A, but
> only one octet for Class C.
>
> This is a little confusing, since in theory you can use
> (192.168.1.*) to (192.168.255.*) in Class C. But once
> you assign the first three octets, the DHCP server-
> assignable IPs (the range of IPs) are limtied to the last
> octet (256 possible IPs). Class C IPs are also known
> as "16-bit" addresses. Class A are known as "24-bit"
> addresses. You can guess what Class B private IPs
> (172.100.*.*) are called.
>
> Just try to add a little understnad; background to IANA
> standards
>
>
>
>>-----Original Message-----
>>Ken wrote:
>>|| Dear Maxx,
>>||
>>|| No, its not a "bug" or "known issue"...unless you're
>>|| consider the whole oragnization of the Internet a
>>|| mistake.
>>
>>What I meant was for the problem I have, been asigned an
>
> IP that is not
>
>>included in the defined pool. About my comment on IP
>
> been given in a random
>
>>manner within a defined pool, I was just saying that not
>
> all DHCP servers
>
>>are behaving this way. And I gave NT4 as an exemple. I
>
> still don't
>
>>understand what is the motive for behaving this way.
>>
>>|| It is the IANA standard for Class A IP
>>|| assignments. Not all IP classes are the same. For
>>|| example, the (192.168.2.*) private IP series that is
>
> the
>
>>|| default in the MN-700 is Class C. Here,the first 3
>>|| octets are the Net ID, with only the last octect being
>>|| the Host ID. In Class A IPs, only the first octet is
>
> the
>
>>|| Net ID, while the remaining 3 octets are all Host IDs.
>>|| So, in your case, only the first (10.*.*.*) sets the
>
> Net
>
>>|| ID. All else is upto the DHCP server.
>>
>>I'm far from been a specialist and must admit my
>
> ignorance here. I thougth
>
>>that class of an IP was defined by the subnet.
>>
>>|| You should be happy you're getting (10.10.10.*).
>>
>>
Sorry I don't understand that. You mean the router is
>
> making me a favor?
>
>>
>>
>>|| Lastly, I think basing fixed connections on physical
>>|| address alone instead of IP could be a security issue.
>>
>>Yes. But having the possibility to do it for say, one
>
> MAC to always be given
>
>>10.10.10.100 seems to me pretty safe and convinient.
>
> BTW, my previous
>
>>Linksys router was able to do it.
>>
>>
>>|| MAC addressing can be easily "spoofed". Since people
>>|| would not easily or routinely change their MAC
>
> address,
>
>>|| once the enter a WiFi zone, it would be recorded and
>>|| could be copied. Just my thought.
>>||
>>|| Hope this helps!
>>||
>>||
>>||| -----Original Message-----
>>||| I guess it is normal when it is by design. But it's
>
> not the case
>
>>||| for all other DHCP server. On NT4, the DHCP assign
>
> it in a logical
>
>>||| order. And I doubt it is normal for the DHCP server
>
> to asign IP
>
>>||| outside of its pool.
>>||| Is it a know bug?
>>||| Thanks for your reply
>>|||
>>||| Maxx
>>|||
>>||| joker wrote:
>>||||| THat is normal for the IP to be randomly selected
>
> from the pool
>
>>||||| of IP addresses.
>>|||||
>>||||| Maxx wrote:
>>|||||
>>|||||| I've set my MN-700 DHCP to use a pool from
>
> 10.10.10.100 up to
>
>>|||||| 10.10.10.253. But for some reason, the first time
>
> I got an IP
>
>>|||||| asigned it was 10.10.10.144. After a power cycle
>
> I got an IP of
>
>>|||||| 10.10.10.45 which is outside the pool I have
>
> defined.
>
>>|||||| Is anyone knows why it is doing so? And why does
>
> it seem to just
>
>>|||||| go in a random fashion instead of starting with
>
> the first
>
>>|||||| available one?
>>||||||
>>|||||| One sugestion I have for MS developers: Adding a
>
> fixed lease
>
>>|||||| based on MAC address would be very usefull. This
>
> way we don't
>
>>|||||| have to use a fixe IP for port fowarding. The
>
> reason I don't
>
>>|||||| wanna use a fixed IP is that I need it to be
>
> dynamic so I can
>
>>|||||| use it at the office and at some Internet Coffee.
>>||||||
>>|||||| Maxx
>>|||
>>|||
>>||| .
>>
>>
>>.
>>