subnet mask 255.255.255.128 & DHCP
Matej Mihelic
I have split our class C network into two subnets with network mask
of 255.255.255.128. There are two DHCP server - each on it's own
subnet, separated by a router with disabled bootp forwarding.
I had no problems with this setup while using W4W 3.11 clients with
MS-TCP32 installed.
Now a few Win95 machines arrived into both subnets. The DHCP
configuration of Win95 stations in the 2nd subnetwork (>*.*.*.129)
works like it should, whereas Win95 stations in the 1st subnet
(<*.*.*.127) only work with fixed IP configuration - DHCP
configuration does not work. From event log on a DHCP server I was
able to determine, that DHCP server hands out new addresses correctly,
however it is reporting NACK a few seconds latter.
Can anybody help?
Best regards,
Matej,
matej....@eunet.si
Marino Holguin
Matej:
You cant' use 255.255.255.128 as a subnet mask for a class C, Try
255.255.255.192 instead, that will split your class C into 2 usable
nets, 62 hosts each.
Marino
Caesar
Marino Holguin (m.ho...@codetel.net.do) wrote:
: Matej:
:
: You cant' use 255.255.255.128 as a subnet mask for a class C, Try
: 255.255.255.192 instead, that will split your class C into 2 usable
: nets, 62 hosts each.
:
: Marino
Sure you can:
earth C:\users\Caesar>ipconfig | more
Windows NT IP Configuration
Ethernet adapter El59x1:
IP Address. . . . . . . . . : 206.109.4.24
Subnet Mask . . . . . . . . : 255.255.255.128
Default Gateway . . . . . . : 206.109.4.126
There's nothing wrong with using a 255.255.255.128 netmask.
--
William S. Duncanson
wil...@neosoft.com
NeoSoft Operations
(888) NEOSOFT or (713) 968-5800
Caesar
Caesar (Cae...@earth.ops.neosoft.com) wrote:
: Marino Holguin (m.ho...@codetel.net.do) wrote:
: : Matej:
: :
: : You cant' use 255.255.255.128 as a subnet mask for a class C, Try
: : 255.255.255.192 instead, that will split your class C into 2 usable
: : nets, 62 hosts each.
: :
: : Marino
:
: Sure you can:
To follow up my own post, a 255.255.255.192 subnet mask will yield 4
networks, not two. See RFC 1878 for details. Although 1878 is not one of
the RFC's that is listed as being part of Microsoft's TCP/IP
implementation, at least as of NT3.51, practical applications have shown
that it is compliant. You might also look into Classless addressing, as
this should help to explain the situation more fully.
John R Buchan
In article <5a1jnd$m...@uuneo.neosoft.com>, Cae...@earth.ops.neosoft.com
says...
>
>Caesar (Cae...@earth.ops.neosoft.com) wrote:
>: Marino Holguin (m.ho...@codetel.net.do) wrote:
>: : Matej:
>: :
>: : You cant' use 255.255.255.128 as a subnet mask for a class C, Try
>: : 255.255.255.192 instead, that will split your class C into 2 usable
>: : nets, 62 hosts each.
>: :
>: : Marino
>:
>: Sure you can:
>:
>
>To follow up my own post, a 255.255.255.192 subnet mask will yield 4
>networks, not two. See RFC 1878 for details. Although 1878 is not one of
>the RFC's that is listed as being part of Microsoft's TCP/IP
>implementation, at least as of NT3.51, practical applications have shown
>that it is compliant. You might also look into Classless addressing, as
>this should help to explain the situation more fully.
Yes, you can use a 128 mask on a C block to yield 2 subnets, a 192 mask
to yield 4 subnets, etc. However, please note that the following quote
from RFC 950 certainly seems to recommend against the practice:
Special Addresses:
From the Assigned Numbers memo [9]:
"In certain contexts, it is useful to have fixed addresses
with functional significance rather than as identifiers of
specific hosts. When such usage is called for, the address
zero is to be interpreted as meaning "this", as in "this
network". The address of all ones are to be interpreted as
meaning "all", as in "all hosts". For example, the address
128.9.255.255 could be interpreted as meaning all hosts on
the network 128.9. Or, the address 0.0.0.37 could be
interpreted as meaning host 37 on this network."
It is useful to preserve and extend the interpretation of these
special addresses in subnetted networks. This means the values
of all zeros and all ones in the subnet field should not be
assigned to actual (physical) subnets.
In the example above, the 6-bit wide subnet field may have
any value except 0 and 63.
I haven't run into any equipment that enforces these special addresses,
(and to the best of my knowledge, MS netwroking still doesn't), but I
have seen a number of others post examples of such equipment or NOSs.
As long as you are in an isolated network and use no equipment that
recognizes these 'special addresses', you should be fine using the masks
mentioned above. However.....
--
John R Buchan Independent Consultant Orlando, Florida USA
MCSE -++- MVP ...................... j.bu...@worldnet.att.net
H. Juergen Werner
>You cant' use 255.255.255.128 as a subnet mask for a class C, Try
>255.255.255.192 instead, that will split your class C into 2 usable
>nets, 62 hosts each.
>
>Marino
>
Hallo Marino,
can you explain the situation a bit more?
I do not understand why it schould not be possible to divid a class C
network in two networks with 128 hosts each, too.
Thanks.
regards,
H. Juergen Werner
Bochum; Germany
E-Mail (ROT13): <U.W.J...@Jreare-Arg.ob.RHarg.qr>
FAX: +49 (0)234 90410-13
H. Juergen Werner
Anthony Stever
Why can't 255.255.255.128 be used to split a class C into two 128 node
subnets (actually, two 126 node subnets since the first and last address in
each subnet are not usable)?
> Matej:
John R Buchan
In article <5a9g3i$3...@uuneo.neosoft.com>, Cae...@earth.ops.neosoft.com says...
>
>Larry Kahn (ka...@lgk.com) wrote:
>: try
>: http://www.pipex.net/people/ianm/classc.html
>:
>: but basically:
>:
>:
>: *You are not able to use the subnets with all zeros or ones in the network
>: portion as these relate to the subnet address and the broadcast address.
>:
>: it may work on some hardware/software but is not guaranteed to work
>: and many routers etc. don't support it..
>:
>: since with a subnet mask of 128 you have 1 network bit and 7 hosts bits
>: so your two networks are either all 0's ie. 0 or all 's ie 1
>: and both are not allowed...
>
>This reasoning is invalid, unless they happen to be using the 0 network, or
>the 255.255.255 network, neither of which are considered class C (which is
>legacy terminology, since the Internet is moving towards classless
>routing).
>
>Actually, any modern router should support it, since it is part of the CIDR
>specification. For example, we are using a 255.255.255.128 netmask here.
>Our network address is 206.109.4.0, and my host is 24. The network portion
>of the address, 11001110 01101101 00000100 0 is not all 0's or all 1's.
>The host portion, 0011000, is also not all 0's or all 1's. The only
>networks that would be affected by the all 0's or all 1's problem is the
>broadcast addresses of the network (in this case, 206.109.4.0 and
>206.109.4.127) and the networks 0 and 255.255.255.
IP: 206.109.4.24 is a class C IP (ie, the high order bits of the first
octet are 110). So the network ID is specified in the first 3 octets and
the host ID is the forth.
Network ID Host ID
---------- -------
206.109.4 24
Since you are using a mask of 255.255.255.128, you are dividing the host
ID into a subnet ID and a host ID (the most significant bit designates
the subnet ID, the 7 least significant designate the host ID). The result
is:
Network ID Subnet ID Host ID
---------- --------- -------
206.109.4 0 24
The special definition of network and host IDs of all 1s and all 0s does
extend to subnet IDs as well (see RFC 950). Since the mask in this
example restricts the subnet ID to a single bit, it must be either
all 1s or all 0s (in the sample value of 24, its all 0s).
Robert Gillies
Cae...@earth.ops.neosoft.com (Caesar) wrote:
>Larry Kahn (ka...@lgk.com) wrote:
>: try
>: http://www.pipex.net/people/ianm/classc.html
>:
>: but basically:
>:
>:
>: *You are not able to use the subnets with all zeros or ones in the network
>: portion as these relate to the subnet address and the broadcast address.
>:
>: it may work on some hardware/software but is not guaranteed to work
>: and many routers etc. don't support it..
>:
>: since with a subnet mask of 128 you have 1 network bit and 7 hosts bits
>: so your two networks are either all 0's ie. 0 or all 's ie 1
>: and both are not allowed...
>
>This reasoning is invalid, unless they happen to be using the 0 network, or
>the 255.255.255 network, neither of which are considered class C (which is
>legacy terminology, since the Internet is moving towards classless
>routing).
>
>Actually, any modern router should support it, since it is part of the CIDR
>specification. For example, we are using a 255.255.255.128 netmask here.
>Our network address is 206.109.4.0, and my host is 24. The network portion
>of the address, 11001110 01101101 00000100 0 is not all 0's or all 1's.
>The host portion, 0011000, is also not all 0's or all 1's. The only
>networks that would be affected by the all 0's or all 1's problem is the
>broadcast addresses of the network (in this case, 206.109.4.0 and
>206.109.4.127) and the networks 0 and 255.255.255.
If you follow this reasoning, I can guarantee that you will miss questions on
the MCP test "Internetworking Microsoft TCP/IP using Microsoft Windows NT".
It may work in some circumstances with some routers, but I don't think it is
valid if you go strictly by the RFCs.
--
Robert Gillies
MCSE, (NT 3.5x and 4.0)
BackOffice MVP
rgil...@hiwaay.net