🚀 go-pugleaf

Mark Landin (landin@cherokee.nsuok.edu) wrote:
> Can you subnet a Class C address, say by using a mask like 255.255.255.128?

  Not 128 specifically, but certainly other values, such as 192, 224 etc.
  The problem with 128 per se is that that creates a one-bit subnet id.
  Since all-zero and all-one values have special meaning, subnet id and host
  id fields must comprise at least __two__ bits each.

-- Ken Mintz

PS:  Some implementations do permit one-bit subnet ids, but this is non-
conformant.

Ken Mintz (mintz@cup.hp.com) wrote:
> Mark Landin (landin@cherokee.nsuok.edu) wrote:
> > Can you subnet a Class C address, say by using a mask like 255.255.255.128?

>   Not 128 specifically, but certainly other values, such as 192, 224 etc.
>   The problem with 128 per se is that that creates a one-bit subnet id.
>   Since all-zero and all-one values have special meaning, subnet id and host
>   id fields must comprise at least __two__ bits each.

> -- Ken Mintz

> PS:  Some implementations do permit one-bit subnet ids, but this is non-
> conformant.
We used 255.255.255.128 successfully, but we are already subnetting
because our adress is actually class B.
--
Steve Fosdick                           Post:   Room 210, B67, BT Labs,
E-mail: stevef@aom.bt.co.uk                     Martlesham Heath,
Tel:    +44 473 642987                          IPSWICH   IP5 7RE
Fax:    +44 473 644607                          England

Mark Landin (landin@cherokee.nsuok.edu) wrote:
> Can you subnet a Class C address, say by using a mask like 255.255.255.128?
> We have two Novell networks that we are with an Ethernet bridge between them.
> Right now, it looks like we will need three Class C address for this:
> one for each ring, and one for the Ethernet bridge. Since the rings are only
> going to have like 20 or 30 clients on each, it seems extremely wasteful to
> use 3*255 potential addresses to serve 60 clients.

No problems, I've subnetted my Class C into 8 subnets. Who wants more than 32
machines at home anyway ????

--
__________________email ken@kgcc.demon.co.uk ____compuserve 100101,3046_______
Ken Green Computer Consultancy
	Bodega, Jasmine Cres. Princes Risborough, Bucks. HP27 0AB, U.K.

landin@cherokee.nsuok.edu (Mark Landin) writes:

>Can you subnet a Class C address, say by using a mask like 255.255.255.128?
>We have two Novell networks that we are with an Ethernet bridge between them.
>Right now, it looks like we will need three Class C address for this:
>one for each ring, and one for the Ethernet bridge. Since the rings are only
>going to have like 20 or 30 clients on each, it seems extremely wasteful to
>use 3*255 potential addresses to serve 60 clients.
Absolutely.  We're masking at the .192 level (0xC0) and for our backbone
at the .240 (0xF0) level.

However -- we've found out -- the very hard way -- over the last two days
that CISCO routers don't like varying subnet masking.  What we had was:



  ----------CISCO-----NOVELL-----ws
  IPX=13487     IXP=13496   IPX=13411
  IP=x.x.240.2              IP=x.x.246.2
  255.255.255.0  255.255.255.240     255.255.255.192

The WS on the left can see the CISCO.  The WS on the right can see the
CISCO.  From inside Mr. CISCO you can see both.  The WS on the left
can login to the netware server.  The WS on the left can't ping the WS
on the right (nor, more important, can the WS on the right talk to the
MVS host on the left)...

Ouch...

So, be careful out there, test everything!

Seriously, though, it does work.  It's a software limitation of the
CISCO, which will be fixed RSN (6-8 months)...

-----Burton

In article <CM5JM7.D19@cup.hp.com> of comp.sys.novell,
  mintz@cup.hp.com (Ken Mintz) writes:
> Mark Landin (landin@cherokee.nsuok.edu) wrote:
> > Can you subnet a Class C address, say by using a mask like 255.255.255.128?
> 
>   Not 128 specifically, but certainly other values, such as 192, 224 etc.
>   The problem with 128 per se is that that creates a one-bit subnet id.
>   Since all-zero and all-one values have special meaning, subnet id and host
>   id fields must comprise at least __two__ bits each.

Nope, sorry.  Not for a *network* number.  However, all 0 and all 1
for a *host* number are the clueless and the broadcast addresses,
respectively.

Hence, a one-bit subnet mask is perfectly valid, albeit a bit silly.

However, the subnet mask 255.255.255.254 would be invalid, since each
subnet would consist of only the clueless and the broadcast address.

> PS:  Some implementations do permit one-bit subnet ids, but this is non-
> conformant.

No, they are not.  Any application that does *not* permit a one-bit
subnet id is nonconformant.

	/hpa

-- 
INTERNET: hpa@nwu.edu               FINGER/TALK: hpa@ahab.eecs.nwu.edu
IBM MAIL: I0050052 at IBMMAIL       HAM RADIO:   N9ITP or SM4TKN
FIDONET:  1:115/511 or 1:115/512    STORMNET:    181:294/101
... .- -- ...   -.. ---   .. -   --- -.   - .... .   .- .. .-.   .-.-.-

> >Can you subnet a Class C address, say by using a mask like 255.255.255.128?
> >We have two Novell networks that we are with an Ethernet bridge between them.
> >Right now, it looks like we will need three Class C address for this:
> >one for each ring, and one for the Ethernet bridge. Since the rings are only
> >going to have like 20 or 30 clients on each, it seems extremely wasteful to
> >use 3*255 potential addresses to serve 60 clients.
> Absolutely.  We're masking at the .192 level (0xC0) and for our backbone
> at the .240 (0xF0) level.
>
> However -- we've found out -- the very hard way -- over the last two days
> that CISCO routers don't like varying subnet masking.  What we had was:
>
>
>
>   ----------CISCO-----NOVELL-----ws
>   IPX=13487     IXP=13496   IPX=13411
>   IP=x.x.240.2              IP=x.x.246.2
>   255.255.255.0  255.255.255.240     255.255.255.192
>
> The WS on the left can see the CISCO.  The WS on the right can see the
> CISCO.  From inside Mr. CISCO you can see both.  The WS on the left
> can login to the netware server.  The WS on the left can't ping the WS
> on the right (nor, more important, can the WS on the right talk to the
> MVS host on the left)...
>
> Ouch...
>
> So, be careful out there, test everything!
>
> Seriously, though, it does work.  It's a software limitation of the
> CISCO, which will be fixed RSN (6-8 months)...

Baloney.

Certainly a class C net can be subnetted. You can have variable length subnet
masks - as long as you use a sensible routing protocol like OSPF. RIP does not
carry the necessary information (subnet mask length).

Steinar Haug, system/networks administrator
SINTEF RUNIT, University of Trondheim, NORWAY
Email: Steinar.Haug@runit.sintef.no, Steinar.Haug@delab.sintef.no

In article <CM8Kny.8Ho@eecs.nwu.edu>, H. Peter Anvin N9ITP <hpa@nwu.edu> wrote:
 > > PS:  Some implementations do permit one-bit subnet ids, but this is non-
 > > conformant.
 >
 > No, they are not.  Any application that does *not* permit a one-bit
 > subnet id is nonconformant.

RFC 1009 states "No subnet should be assigned the value 0 or -1 (all one
bits)".  See second paragraph of page 6. Has this been superseded by a later
RFC?

// marc
--
// marc@dumbcat.sf.ca.us or marc@ascend.com

In article <CM5JM7.D19@cup.hp.com> mintz@cup.hp.com (Ken Mintz) writes:
>From: mintz@cup.hp.com (Ken Mintz)
>Subject: Re: Subnetting Class C IP networks
>Date: Fri, 4 Mar 1994 18:06:07 GMT

>Mark Landin (landin@cherokee.nsuok.edu) wrote:
>> Can you subnet a Class C address, say by using a mask like 255.255.255.128?

>  Not 128 specifically, but certainly other values, such as 192, 224 etc.
>  The problem with 128 per se is that that creates a one-bit subnet id.
>  Since all-zero and all-one values have special meaning, subnet id and host
>  id fields must comprise at least __two__ bits each.

>-- Ken Mintz

>PS:  Some implementations do permit one-bit subnet ids, but this is non-
>conformant.

In article <CM5JM7.D19@cup.hp.com> mintz@cup.hp.com (Ken Mintz) writes:
>From: mintz@cup.hp.com (Ken Mintz)
>Subject: Re: Subnetting Class C IP networks
>Date: Fri, 4 Mar 1994 18:06:07 GMT

>Mark Landin (landin@cherokee.nsuok.edu) wrote:
>> Can you subnet a Class C address, say by using a mask like 255.255.255.128?

>  Not 128 specifically, but certainly other values, such as 192, 224 etc.
>  The problem with 128 per se is that that creates a one-bit subnet id.
>  Since all-zero and all-one values have special meaning, subnet id and host
>  id fields must comprise at least __two__ bits each.

>-- Ken Mintz

>PS:  Some implementations do permit one-bit subnet ids, but this is non-
>conformant.
We have been using the mask 255.255.255.224 with complete sucess.  This gives
us 6 subnets with 30 hosts each.

Wendell Read

In article <2l8fcr$5kc@medusa.aom.bt.co.uk> stevef@aom.bt.co.uk (Steve Fosdick) writes:
>Ken Mintz (mintz@cup.hp.com) wrote:
>> Mark Landin (landin@cherokee.nsuok.edu) wrote:
>> > Can you subnet a Class C address, say by using a mask like 255.255.255.128?

>>   Not 128 specifically, but certainly other values, such as 192, 224 etc.
>>   The problem with 128 per se is that that creates a one-bit subnet id.
>>   Since all-zero and all-one values have special meaning, subnet id and host
>>   id fields must comprise at least __two__ bits each.

This is correct.

>We used 255.255.255.128 successfully, but we are already subnetting
>because our adress is actually class B.

....and since you have Class B address space instead of Class C, you're
using a nine-bit subnet mask, which is fine. The first two bytes are
your *network* mask in a Class B, while the remaining two are the real
*subnet* mask. In a Class C network, three bytes are the network mask,
and all subnetting must be done in the final byte. (Yes, a Class A network
has one byte of network mask, and can use most of the remaining three for
subnetting.)

Subnetting is done using bits within your address space--the bits over the
assigned network aren't included. This is somewhat confusing, unfortunantly;
masks are interpreted differently depending upon what type of network they're
being used in. They break down like this:

		   (decimal)			     (binary)
Mask:		255.255.255.240		11111111.11111111.11111111.11110000

Class A:	NNN SSS SSS S/H		NNNNNNNN SSSSSSSS SSSSSSSS SSSSHHHH

Class B:	NNN NNN SSS S/H		NNNNNNNN NNNNNNNN SSSSSSSS SSSSHHHH

Class C:	NNN NNN NNN S/H		NNNNNNNN NNNNNNNN NNNNNNNN SSSSHHHH

Where "N" indicates network mask, "S" indicates subnet mask, and "H" is host
address bits. So, the mask 255.255.255.240 would provide a 20-bit subnet mask
in a Class A network, a 12-bit in a Class B, and only a 4-bit on a Class C.





--
Gary Heston      SCI Systems, Inc.       gary@sci.com       site admin
   The Chairman of the Board and the CFO speak for SCI. I'm neither.
"Quit while you're ahead. All the best gamblers do."  Baltasar Gracian

In article <CM8Kny.8Ho@eecs.nwu.edu> hpa@nwu.edu (H. Peter Anvin) writes:
>In article <CM5JM7.D19@cup.hp.com> of comp.sys.novell,
>  mintz@cup.hp.com (Ken Mintz) writes:
 
>>   Not 128 specifically, but certainly other values, such as 192, 224 etc.
>>   The problem with 128 per se is that that creates a one-bit subnet id.
>>   Since all-zero and all-one values have special meaning, subnet id and host
>>   id fields must comprise at least __two__ bits each.

>Nope, sorry.  Not for a *network* number.  However, all 0 and all 1
>for a *host* number are the clueless and the broadcast addresses,
>respectively.

There are differences between "network" and "subnet" masks; they aren't
the same thing. Ken is correct in his two-bit minimum statement.

>Hence, a one-bit subnet mask is perfectly valid, albeit a bit silly.

No, it's specifically stated as not acceptable. Check RFC 950, page 6.

>However, the subnet mask 255.255.255.254 would be invalid, since each
>subnet would consist of only the clueless and the broadcast address.

>> PS:  Some implementations do permit one-bit subnet ids, but this is non-
>> conformant.

>No, they are not.  Any application that does *not* permit a one-bit
>subnet id is nonconformant.

They *are* non-conformant, and will stay that way until someone writes
a new RFC and gets it approved--something I don't consider likely.




-- 
Gary Heston      SCI Systems, Inc.       gary@sci.com       site admin
   The Chairman of the Board and the CFO speak for SCI. I'm neither.
"Quit while you're ahead. All the best gamblers do."  Baltasar Gracian

In article <2ldkpk$sq5@dumbcat.sf.ca.us>, marc@dumbcat.sf.ca.us (Marco S Hyman) writes:
> In article <CM8Kny.8Ho@eecs.nwu.edu>, H. Peter Anvin N9ITP <hpa@nwu.edu> wrote:
>  > > PS:  Some implementations do permit one-bit subnet ids, but this is non-
>  > > conformant.
>  >
>  > No, they are not.  Any application that does *not* permit a one-bit
>  > subnet id is nonconformant.
>
> RFC 1009 states "No subnet should be assigned the value 0 or -1 (all one
> bits)".  See second paragraph of page 6. Has this been superseded by a later
> RFC?

No.  The current "Requirements for Internet Hosts -- Communication Layers"
(RFC 1122), which supplements, clarifies, and/or supersedes the primary
protocol documents, states in section 3.2.1.3 on addressing on page 31:

            IP addresses are not permitted to have the value 0 or -1 for
            any of the <Host-number>, <Network-number>, or <Subnet-
            number> fields (except in the special cases listed above).
            This implies that each of these fields will be at least two
            bits long.

The purpose of RFC 1122 (actually, of any of the "standard" RFCs) is to
help vendors and implementors avoid problems of lack of interoperability,
bad performance and/or robustness.  So there's nothing wrong with an
implementation allowing the use of subnets 0 and/or all-ones *as an option*,
especially if there is a big demand for it, as there is in this case in
these days of increasingly scarce IP addresses.  However, to allow it by
*default* would be a bad idea, since

o   commonly-used "classful" routing protocols (e.g., RIP) can't distin-
    guish a network route from a subnet route for subnet zero, and

o   some hosts may confuse IP addresses on the all-ones subnet with
    broadcast addresses.  True, the code would have to be pretty sleazoid
    to do this, but any confusion about broadcast addresses is potentially
    disasterous.

Another poster mentioned varying-length subnet masks.  This is a separate
issue, but they also require classless routing protocols like OSPF.  Often
one can workaround these kinds of problems by using static routes or assigning
multiple subnets on the same "wire."  Cisco routers (and probably other brands)
have had the latter capability for quite some time -- see the description of
"secondary address" in the Cisco docs.

I hope this helps.  I wanted to direct followups to comp.protocols.tcp-ip
but my news software wouldn't let me, too many x-posted groups already...
--
James Harvey   harvey@iupui.edu   IUPUI OIT Technical Support/VMS/Unix/Networks
Disclaimer:  These are my own opinions.  I do not speak for Indiana University.
"Contrariwise," continued Tweedledee, "if it was so, it might be, and if it
were so, it would be; but as it isn't, it ain't.  That's logic!" -Lewis Carroll