>Thank you Howard!
>
>I have to ask then would the command "ip subnet-zero" be useless then?
>or are there some environments where this command is usefull and 
>others where this command isn't. It would now seem as though this 
>command would only be nedded in a classfull environment?


You've pretty much answered it.  IP subnet zero works quite well with 
everything except IGRP, RIPv1, and BGP3.

Very conservative designers will avoid it on the theory that it might 
create a problem of backwards compatibility.  See either my RFC2072 
or my Designing Addressing Architectures book for details, but here 
are some ideas.

When I renumber an enterprise from classful to classless, I have a 
few tricks.  In a classful environment, I can assume subnet zero is 
not in use.  So the first thing I do is split subnet zero into /30 
VLSM and use them for router-to-router links, first on new router 
links, then on routers with parallel links, and then on routers with 
single existing links.

Renumbering the WAN router-to-router links then frees up the previous 
classful subnets that were used, inefficiently, for router-to-router 
links. I then renumber those into small VLSM subnets (e.g., /28 or 
/29) and use the new subnets for small office LANs.


-------
No intellectual properties were disturbed by the creation of this 
message. All information was handled in a manner consistent with 
guidelines from People for the Ethical Treatment of Packets.  Any 
packets that were source quenched first were given a wetsuit and a 
snorkel.
-------

>As this message is sent I am reading the RFC's mentioned below and I 
>may end up answering my own question. Thanks again.
>
>>>>Brian
>
>
>>From: "Howard C. Berkowitz" <[EMAIL PROTECTED]>
>>Reply-To: "Howard C. Berkowitz" <[EMAIL PROTECTED]>
>>To: [EMAIL PROTECTED]
>>Subject: Re: ip subnetting
>>Date: Wed, 20 Dec 2000 09:08:10 -0500
>>
>>>Group,
>>>    In determining valid subnetworks it used to be true all were valid accept
>>>the all 1's subnet and the all 0's subnet. In an environment where all
>>>devices are compatible, I understand, that by using the configuration
>>>command "ip subnet-zero" it enables the use of the all 0's subnet.
>>
>>The first thing to remember is that modern routers don't have a
>>concept of network or subnet.  Classful notation and subnetting is
>>more a crutch for people than anything else.  Routers understand
>>prefixes, which are no more or less than contiguous numbers of bits
>>starting on the left.
>>
>>The following all are perfectly legitimate /20's:
>>
>>       10.0.0.0/20
>>       172.16.240.0/20
>>       192.255.240.0/20
>>
>>The confusion around all ones and all zeroes subnet comes from a lack
>>of prefix length information, and situations in which the router
>>either had to infer the prefix from the high order "class" bits, or
>>from local configuration information. Now, modern routing protocols
>>transmit explicit lengths.
>>
>>>The
>>>question I ask then, wouldn't that leave the all 1's subnet still an invalid
>>>subnet even when using the previous mentioned command? or is there something
>>>else that allows the all 1's subnetwork? I have been asked this question,
>>>and I don't feel I am correct. The new formula noted in the CCNA book notes
>>>the new equation for determining the amount of valid subnetworks is 2^# of
>>>subnet bits, and I figured it would be 2^#of subnet bits - 1 (for the all
>>>1's subnet). I don't pose this question without first a bit of research and
>>>everything I have read indicates the "ip subnet-zero" command only enables
>>>the use of the all 0's subnetwork. I however cannot find anything indicating
>>>the use of the all 1's subnet is permitted. Any help is appreciated.
>>>
>>>   >>>Brian
>>
>>My study question to you would be:
>>
>>      why shouldn't all 0's and all 1's work?
>>
>>You may want to consult RFC 1812, and possibly 1879.
>>
>>
>>--------------------
>>
>>No CCIE's were harmed or otherwise violated in the preparation of 
>>this message.
>>
>>_________________________________
>>FAQ, list archives, and subscription info: 
>>http://www.groupstudy.com/list/cisco.html
>>Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
>
>_________________________________________________________________
>Get your FREE download of MSN Explorer at http://explorer.msn.com

_________________________________
FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html
Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]

Reply via email to