DHCP Subnet Selection: The IPv4 Subnet Selection Option for DHCP
A DHCP option that lets a client name the subnet it wants its address to come from.
Waters, G., "The IPv4 Subnet Selection Option for DHCP", RFC 3011, DOI 10.17487/RFC3011, November 2000, <https://www.rfc-editor.org/info/rfc3011>.
@misc{rfc3011,
series = {Request for Comments},
number = 3011,
howpublished = {RFC 3011},
publisher = {RFC Editor},
doi = {10.17487/RFC3011},
url = {https://www.rfc-editor.org/info/rfc3011},
author = {G. Waters},
title = {{The IPv4 Subnet Selection Option for DHCP}},
pagetotal = 7,
year = 2000,
month = nov,
} RFC text
Contents
Network Working Group G. Waters
Request for Comments: 3011 Nortel Networks
Category: Standards Track November 2000
The IPv4 Subnet Selection Option for DHCP
Status of this Memo
This document specifies an Internet standards track protocol for the
Internet community, and requests discussion and suggestions for
improvements. Please refer to the current edition of the "Internet
Official Protocol Standards" (STD 1) for the standardization state
and status of this protocol. Distribution of this memo is unlimited.
Copyright Notice
Copyright (C) The Internet Society (2000). All Rights Reserved.
Abstract
This memo defines a new Dynamic Host Configuration Protocol (DHCP)
option for selecting the subnet on which to allocate an address.
This option would override a DHCP server's normal methods of
selecting the subnet on which to allocate an address for a client.
Table of Contents
1. Introduction..................................................1
1.1. Motivational Example........................................2
2. Subnet Selection Option Definition............................3
3. Intellectual Property.........................................4
4. IANA Considerations...........................................4
5. Acknowledgements..............................................5
6. Security Considerations.......................................5
7. References....................................................5
8. Editor's Addresses............................................6
9. Full Copyright Statement......................................7
1. Introduction
The Dynamic Host Configuration Protocol (DHCP) [RFC2131] provides a
framework for passing configuration information to hosts on a TCP/IP
network. RFC 2132 [RFC2132] specifies DHCP option configuration
information that may be carried in DHCP packets to/from the DHCP
server and the DHCP client. This document specifies a new DHCP
option.
To select the subnet on which to allocate an address, the DHCP server
determines the subnet from which the request originated, and then
selects an address on the originating subnet or on a subnet that is
on the same network segment as the originating subnet. The subnet
from which the request originates can be determined by:
o Using the subnet address of the giaddr field in the DHCP packet
header, or if the giaddr field is zero;
o Using the subnet address of the local interface on which the DHCP
server received the packet.
This memo defines a new DHCP option, the subnet selection option,
which allows the DHCP client to specify the subnet on which to
allocate an address. This option takes precedence over the methods
that the DHCP server uses to determine the subnet on which to select
an address.
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
"SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
document are to be interpreted as described in [RFC2119].
1.1. Motivational Example
An example of where this option could be useful is in a device (e.g.:
a RAS device) that is allocating addresses on behalf of its clients.
In this case the device would be allocating addresses through DHCP
and then managing those addresses among its clients.
In this scenario, the device is connected to a private "internal"
network on which the DHCP server would be located. The device is
also connected to one or more service providing "external" networks
(i.e.: the networks that the device's clients are connected to).
Furthermore, the internal network is not IP connected to the external
networks, although inside the device there is connectivity between
the internal and external networks (e.g.: though the backplane).
Recall that the device is allocating addresses for its clients on the
external networks and that there is no IP connectivity between the
internal network and the external networks. The DHCP requests cannot
originate from the external networks since packets cannot be routed
between the external network and the internal network. Thus, the
DHCP requests must originate from the internal network. The problem
with originating the DHCP requests from the internal network is that
the DHCP server will allocate addresses on the internal network's
subnet, when what is required are addresses on the external subnets.
The subnet selection option provides a solution to this problem.
The device would send its DHCP request on the internal subnet, but
would include the subnet selection option containing the address of
the external subnet on which it requires the address. The subnet
selection option instructs the DHCP server to allocate the address on
the requested subnet as opposed to the normal operation of allocating
the address on the subnet from which the DHCP request originated.
2. Subnet Selection Option Definition
The subnet selection option is a DHCP option. The option contains a
single IPv4 address that is the address of a subnet. The value for
the subnet address is determined by taking any IPv4 address on the
subnet and ANDing that address with the subnet mask (i.e.: the
network and subnet bits are left alone and the remaining (address)
bits are set to zero). When the DHCP server is configured to respond
to this option, is allocating an address, and this option is present
then the DHCP server MUST allocate the address on either:
o the subnet specified in the subnet selection option, or;
o a subnet on the same network segment as the subnet specified in the
subnet selection option.
The format of the option is:
Code Len IPv4 Address
+-----+-----+-----+-----+-----+-----+
| 118 | 4 | A1 | A2 | A3 | A4 |
+-----+-----+-----+-----+-----+-----+
Servers configured to support this option MUST return an identical
copy of the option to any client that sends it, regardless of whether
or not the client requests the option in a parameter request list.
Clients using this option MUST discard DHCPOFFER or DHCPACK packets
that do not contain this option.
This option does not require changes to operations or features of the
DHCP server other than to select the subnet on which to allocate an
address. For example, the handling of DHCPDISCOVER for an unknown
subnet should continue to operate unchanged.
When this option is present and the server is configured to support
this option, the server MUST NOT offer an address that is not on the
requested subnet or network segment. Servers that do not understand
this option will allocate an address using their normal algorithms
and will not return this option in the DHCPOFFER or DHCPACK. In this
case the client will discard the DHCPOFFER or DHCPACK. Servers that
understand this option but are administratively configured to ignore
the option MUST ignore the option, use their normal algorithms to
allocate an address, and MUST NOT return this option in the DHCPOFFER
or DHCPACK. In this case the client will discard the DHCPOFFER or
DHCPACK.
During an address renew, the DHCP server may send a DHCPACK directly
to the allocated address, however packets from the DHCP server may
not be routable to the address. Thus, in all packets that the DHCP
client sends that contain the subnet selection option, the giaddr
field in the BOOTP header MUST be set to an IPv4 address on which the
DHCP client will accept DHCP packets (e.g.: the address on the subnet
connected to the internal network).
The IPv4 address to which a DHCP server sends a reply to MUST be the
same as it would chose when this option is not present.
3. Intellectual Property
The IETF takes no position regarding the validity or scope of any
intellectual property or other rights that might be claimed to
pertain to the implementation or use of the technology described in
this document or the extent to which any license under such rights
might or might not be available; neither does it represent that it
has made any effort to identify any such rights. Information on the
IETF's procedures with respect to rights in standards-track and
standards-related documentation can be found in BCP-11.
Copies of claims of rights made available for publication and any
assurances of licenses to be made available, or the result of an
attempt made to obtain a general license or permission for the use of
such proprietary rights by implementers or users of this
specification can be obtained from the IETF Secretariat.
The IETF invites any interested party to bring to its attention any
copyrights, patents or patent applications, or other proprietary
rights which may cover technology that may be required to practice
this standard. Please address the information to the IETF Executive
Director.
4. IANA Considerations
IANA has assigned a value of 118 for the DHCP option code described
in this document.
5. Acknowledgements
This document is the result of work undertaken the by DHCP working
group. Thanks to Ted Lemon, Tim Aston and Ralph Droms for their
helpful comments in this work.
W. Mark Townsley and Pratik Gupta originally published a subnet
selection option Internet Draft in July 1997. The work in this
document was not based on the original work but it does achieve the
same goals.
6. Security Considerations
DHCP currently provides no authentication or security mechanisms.
Potential exposures to attack are discussed is section 7 of the
protocol specification [RFC2131].
The subnet selection option allows for the DHCP client to specify the
subnet on which to allocate an address. This would allow a client to
perform a more complete address-pool exhaustion attack since the
client would no longer be restricted to attacking address-pools on
just its local subnet.
Servers that implement the subnet selection option MUST by default
disable use of the feature; it must specifically be enabled through
configuration. Moreover, a server SHOULD provide the ability to
selectively enable use of the feature under restricted conditions,
e.g., by enabling use of the option only from explicitly configured
client-ids, enabling its use only by clients on a particular subnet,
or restricting the subnets (as indicated in the subnet selection
option) from which addresses may be requested.
7. References
[RFC2119] Bradner, S., "Key words for use in RFCs to Indicate
Requirement Levels", BCP 14, RFC 2119, March 1997.
[RFC2131] Droms, R. "Dynamic Host Configuration Protocol", RFC 2131,
March 1997.
[RFC2132] Alexander, S. and R. Droms, "DHCP Options and BOOTP Vendor
Extensions", RFC 2132, March 1997.
8. Editor's Address
Glenn Waters
Nortel Networks
310-875 Carling Avenue,
Ottawa, Ontario K1S 5P1
Canada
Phone: +1 613-765-0249
EMail: gww@nortelnetworks.com
9. Full Copyright Statement
Copyright (C) The Internet Society (2000). All Rights Reserved.
This document and translations of it may be copied and furnished to
others, and derivative works that comment on or otherwise explain it
or assist in its implementation may be prepared, copied, published
and distributed, in whole or in part, without restriction of any
kind, provided that the above copyright notice and this paragraph are
included on all such copies and derivative works. However, this
document itself may not be modified in any way, such as by removing
the copyright notice or references to the Internet Society or other
Internet organizations, except as needed for the purpose of
developing Internet standards in which case the procedures for
copyrights defined in the Internet Standards process must be
followed, or as required to translate it into languages other than
English.
The limited permissions granted above are perpetual and will not be
revoked by the Internet Society or its successors or assigns.
This document and the information contained herein is provided on an
"AS IS" basis and THE INTERNET SOCIETY AND THE INTERNET ENGINEERING
TASK FORCE DISCLAIMS ALL WARRANTIES, EXPRESS OR IMPLIED, INCLUDING
BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE INFORMATION
HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED WARRANTIES OF
MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.
Acknowledgement
Funding for the RFC Editor function is currently provided by the
Internet Society.
AI ELI5
On this page
§1How a DHCP Server Picks a Subnet
The Dynamic Host Configuration Protocol (RFC 2131) hands out configuration to hosts on a TCP/IP network, and RFC 2132 defines the options that DHCP packets can carry. RFC 3011 adds one more option.
Normally the server works out which subnet the request came from, then allocates an address on that subnet or on a subnet on the same network segment. It finds the originating subnet in one of two ways: from the subnet address of the giaddr field in the DHCP packet header, or, if giaddr is zero, from the subnet address of the local interface that received the packet.
The new subnet selection option lets the client state the subnet itself, and it takes precedence over both of those methods. The RFC uses the key words MUST and SHOULD as defined in RFC 2119.
Usually the helper decides which section of the hall you sit in, based on the door you came through. This option lets you say which section you want instead.
§1.1The Motivating Example
The RFC's example is a device, such as a RAS device, that allocates addresses on behalf of its own clients. It gets the addresses through DHCP and then manages them among those clients.
The device sits on a private internal network, where the DHCP server lives. It is also connected to one or more external networks, which is where its clients are. The internal network is not IP connected to the external ones, although inside the device there is connectivity between them, for example through the backplane.
That is the problem. The DHCP requests cannot come from the external networks, because packets cannot be routed between them and the internal network. So they must come from the internal network, and the server would then allocate addresses on the internal subnet, which is not what is wanted. The device sends its request on the internal subnet but includes the subnet selection option with the address of the external subnet it needs. The server then allocates on the requested subnet instead.
Imagine a receptionist who can only phone the ticket office from the back room, but is buying tickets for people sitting in the front rooms. The receptionist says which room the ticket is for, so the seat is in the right place.
§2The Option Format
The option carries a single IPv4 address that is the address of a subnet. To get that value, take any IPv4 address on the subnet and AND it with the subnet mask: the network and subnet bits are left alone and the remaining address bits are set to zero.
On the wire it is three parts: the code 118, a length of 4, and the four bytes of the address. IANA assigned the value 118 for this option code.
When the server is configured to respond to the option, is allocating an address, and the option is present, it MUST allocate the address on the subnet named in the option, or on a subnet on the same network segment as that subnet.
The option is a tiny label with a number saying what kind of label it is, how long it is, and the name of the subnet you want.
| Field | Name | Size | Value | Description |
|---|---|---|---|---|
CODE | Code | 1 byte | 118 | Option code assigned by IANA. |
LEN | Len | 1 byte | 4 | Length of the data that follows, in bytes. |
ADDRESS | IPv4 Address | 4 bytes | Subnet address: any address on the subnet ANDed with the subnet mask. |
§3Server and Client Behavior
Servers set up to support the option MUST return an identical copy of it to any client that sends it, whether or not the client listed it in a parameter request list. Clients that use the option MUST discard any DHCPOFFER or DHCPACK that does not contain it.
That rule handles servers that do not play along. A server that does not understand the option allocates with its normal algorithms and does not return the option, so the client throws the reply away. A server that understands the option but is administratively set to ignore it MUST ignore it, allocate normally, and MUST NOT return it, with the same result for the client. A server that does honor it MUST NOT offer an address outside the requested subnet or network segment.
The option changes nothing else about the server. For example, handling of a DHCPDISCOVER for an unknown subnet continues to work as before.
Replies need a route back. During a renewal the server may send a DHCPACK directly to the allocated address, but packets from the server might not be routable there. So in every packet the client sends that contains the option, the giaddr field in the BOOTP header MUST be set to an IPv4 address on which the client will accept DHCP packets, for example the address on the subnet connected to the internal network. The address the server replies to MUST be the same one it would choose without the option.
If the helper does not understand your request for a section, it just gives you a seat anywhere, and you hand the ticket back because it is not what you asked for. You also have to leave a note saying where to send answers, since the answers cannot reach your seat.
- DHCPDISCOVER + option 118Sent from the internal subnet, naming the subnet where the address is wanted. giaddr is set to an address where the client accepts DHCP packets.
- DHCPOFFER + identical option 118The offered address is on the requested subnet or its network segment.
§4Security Considerations
The RFC points out that DHCP provides no authentication or security mechanisms, and refers to section 7 of RFC 2131 for the exposures.
The option adds its own risk. Because a client can pick the subnet, it can run a more complete address-pool exhaustion attack: it is no longer limited to the pools on its local subnet.
Therefore servers that implement the option MUST disable it by default, so it has to be switched on through configuration. A server SHOULD also let administrators enable it only under restricted conditions, such as only for explicitly configured client-ids, only for clients on a particular subnet, or only for requests naming certain subnets.
If anyone could ask for any section, a mischievous person could use up all the seats everywhere. So the hall keeps the option turned off unless the manager turns it on for particular people.