Two Class C Networks?

Yep huliyar, you can do that.. though some people say that 1 bit subnetting is illegal (well specifically Cisco says its illegal.. buts its quite a useful mask if you ask me)

However in relation to the post, it won't give enough IPs for the scenario

Ok, this is going to be a long one.

Actually, I believe that you can use 1 bit for the subnet. Apparently, this is something that is being changed. Older software may not work with this.

The problem was that all 0's and all 1's on the network side was illegal. This had to do with how the routing table was set up. Now this has apparently changed, somewhat. I know that in our old system we did have our system working with all zeros in our subnet mask and it worked fine.

The designation ((2^N)-2) is used when you are taking out the 0's and 1's addresses and the (2^N) is used when you are not.

The problem with the ((2^N)-2) method, is that you lose all those addresses (this would lose 252 addresses for a 128 mask). The (2^N) method would apparently work for Knights situation as you would have 126 addresses for each subnet (you still need that router, however).

I was looking at a couple of articles that show the rfc 1878, which mentions this. One of the articles www.mcmcse.com/articles/subnetting.shtml expains it. Unfortunately, if you look at page 7 of the rfc that this article points to:

www.cis.ohio-state.edu/cgi-bin/rfc/rfc1878.html

you will see that it mentions that all 0s and 1s of the host being excluded is obolete. This is not the case and I believe this is in error. I am not sure why the 0s and 1s were illegal for the network portion (obviously, they have changed this), but I do know why the hosts can not be all 0's and 1s. All zeros mean the network. If you look at 2 routing tables

Here is a route print with a mask of 255.255.255.0:

[code:1]C:\WINNT\system32>route print
===========================================================================
Interface List
0x1 ........................... MS TCP Loopback interface
0x2 ...44 45 53 54 42 00 ...... NOC Extranet Access Adapter
0x1000004 ...00 02 b3 5c ec 0a ...... Intel(R) PRO/100+ Management Adapter
===========================================================================
===========================================================================
Active Routes:
Network Destination Netmask Gateway Interface Metric
0.0.0.0 0.0.0.0 192.168.122.1 192.168.122.7 1
127.0.0.0 255.0.0.0 127.0.0.1 127.0.0.1 1
192.168.122.0 255.255.255.128 192.168.122.7 192.168.122.7 1
192.168.122.7 255.255.255.255 127.0.0.1 127.0.0.1 1
192.168.122.255 255.255.255.255 192.168.122.7 192.168.122.7 1
224.0.0.0 224.0.0.0 192.168.122.7 192.168.122.7 1
255.255.255.255 255.255.255.255 192.168.122.7 2 1
Default Gateway: 192.168.122.1
===========================================================================
[/code:1]

If you look at the 3rd line of the routes, you will see 192.168.122.0 in the first column (which is titled "Network Destination"), if you had an address of 0, your network and host would match and would cause problems. You will find W2K will not even allow you to do this.

Here is a route print with a mask of 255.255.255.128 (1 bit set):

[code:1]C:\WINNT\system32>route print
===========================================================================
Interface List
0x1 ........................... MS TCP Loopback interface
0x2 ...44 45 53 54 42 00 ...... NOC Extranet Access Adapter
0x1000004 ...00 02 b3 5c ec 0a ...... Intel(R) PRO/100+ Management Adapter
===========================================================================
===========================================================================
Active Routes:
Network Destination Netmask Gateway Interface Metric
127.0.0.0 255.0.0.0 127.0.0.1 127.0.0.1 1
192.168.122.128 255.255.255.128 192.168.122.135 192.168.122.135 1
192.168.122.135 255.255.255.255 127.0.0.1 127.0.0.1 1
192.168.122.255 255.255.255.255 192.168.122.135 192.168.122.135 1
224.0.0.0 224.0.0.0 192.168.122.135 192.168.122.135 1
255.255.255.255 255.255.255.255 192.168.122.135 2 1
===========================================================================
[/code:1]

Here look at line 2 of the routes and you will see 192.168.122.128. If you were to set your host to 0, you would have the same address as this line.

You cannot have all 1s, as this is the broadcast address for this network. This is the address that ARP uses to broadcast to all hosts on the network.

Now, according to the above article, the MCSE would use the ((2^N)-2) method (although W2K uses the (2^N) method), but Ciscos new track apparently uses the (2^N) method. Chris or Sahirh could tell you more about this, as they just took the test.

So all the people have not gotten together on this yet.

Gotta love math to do networkin!

Good to see how dedicated every1 is to helping others..

firewall.cx Forums is my home page now!

tfs: when I did my ccna this year I used the ((2^N)-2) method worked well, but yes intresting to see if the new method is (2^N)

I don't know if the new certification is using the new method or not. Just what I read.

Now the ((2^N)-2) is used for the dermination of the number of hosts. That was where I believe the one paper was wrong, as that can't change (and we are only talking about losing 2 addresses as opposed to losing a bunch if we drop 0's and 1's from the network side).

Guys I believe that the (2^n)-2 is valid for hosts and networks.

I remember reading it when I did my ccna recertification and also when performing the onhands labs in the exam, they specifically noted NOT to use subnet zero, that is, the first subnet in the network.

On the other hand, I haven't heard anything about a 'new' rule where the '-2' is absent;

This is one of those things where Cisco behaves really strange..
On the tests I'm 105% sure that its (2^n)-2... they've always had it that way (maybe because the math changes a bit if you deal with the 128 mask)... its very unlikely that they changed it for the new exams.

However Cisco routers by default are configured with the
'ip subnet zero' command... which allows you to use subnet 0 !! Strange...

Anyway, nice to see tfs posting his mini e-books again