PHB ID Codes: Per Hop Behavior Identification Codes
A 16-bit code that names a Differentiated Services Per Hop Behavior, or a set of them, inside protocol messages.
Black, D., Brim, S., Carpenter, B., and F. Le Faucheur, "Per Hop Behavior Identification Codes", RFC 3140, DOI 10.17487/RFC3140, June 2001, <https://www.rfc-editor.org/info/rfc3140>.
@misc{rfc3140,
series = {Request for Comments},
number = 3140,
howpublished = {RFC 3140},
publisher = {RFC Editor},
doi = {10.17487/RFC3140},
url = {https://www.rfc-editor.org/info/rfc3140},
author = {D. Black and S. Brim and B. Carpenter and F. Le Faucheur},
title = {{Per Hop Behavior Identification Codes}},
pagetotal = 8,
year = 2001,
month = jun,
} RFC text
Contents
Network Working Group D. Black
Request for Comments: 3140 S. Brim
Obsoletes: 2836 B. Carpenter
Category: Standards Track F. Le Faucheur
June 2001
Per Hop Behavior Identification Codes
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 (2001). All Rights Reserved.
Abstract
This document defines a 16 bit encoding mechanism for the
identification of differentiated services Per Hop Behaviors in
protocol messages. It replaces RFC 2836.
Table of Contents
1. Introduction.................................................2
1.1. Usage Scenarios............................................2
2. Encoding.....................................................3
3. Signalling the Class Selector Codepoints.....................4
4. IANA Considerations..........................................5
5. Security Considerations......................................5
Changes from RFC 2836...........................................5
Acknowledgements................................................6
References......................................................6
Authors' Addresses..............................................6
Intellectual Property...........................................7
Full Copyright Statement........................................8
1. Introduction
Differentiated Services [RFC 2474, RFC 2475] introduces the notion of
Per Hop Behaviors (PHBs) that define how traffic belonging to a
particular behavior aggregate is treated at an individual network
node. In IP packet headers, PHBs are not indicated as such; instead
Differentiated Services Codepoint (DSCP) values are used. There are
only 64 possible DSCP values, but there is no such limit on the
number of PHBs. In a given network domain, there is a locally
defined mapping between DSCP values and PHBs. Standardized PHBs
recommend a DSCP mapping, but network operators may choose
alternative mappings.
In some cases it is necessary or desirable to identify a particular
PHB in a protocol message, such as a message negotiating bandwidth
management or path selection, especially when such messages pass
between management domains. Examples where work is in progress
include communication between bandwidth brokers, and MPLS support of
diffserv.
In certain cases, what needs to be identified is not an individual
PHB, but a set of PHBs. One example is a set of PHBs that must
follow the same physical path to prevent re-ordering. An instance of
this is the set of three PHBs belonging to a single Assured
Forwarding class, such as the PHBs AF11, AF12 and AF13 [RFC 2597].
This document defines a binary encoding to uniquely identify PHBs
and/or sets of PHBs in protocol messages. This encoding MUST be used
when such identification is required.
This document replaces RFC 2836, which omitted considerations for the
Class Selector codepoints.
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. Usage Scenarios
Diffserv services are expected to be supported over various
underlying technologies which we broadly refer to as "link layers"
for the purpose of this discussion. For the transport of IP packets,
some of these link layers make use of connections or logical
connections where the forwarding behavior supported by each link
layer device is a property of the connection. In particular, within
the link layer domain, each link layer node will schedule traffic
depending on which connection the traffic is transported in.
Examples of such "link layers" include ATM and MPLS.
For efficient support of diffserv over these link layers, one model
is for different Behavior Aggregates (BAs) (or sets of Behavior
Aggregates) to be transported over different connections so that they
are granted different (and appropriate) forwarding behaviors inside
the link layer cloud. When those connections are dynamically
established for the transport of diffserv traffic, it is very useful
to communicate at connection establishment time what forwarding
behavior(s) is (are) to be granted to each connection by the link
layer device so that the BAs transported experience consistent
forwarding behavior inside the link layer cloud. This can be
achieved by including in the connection establishment signaling
messages the encoding of the corresponding PHB, or set of PHBs, as
defined in this document. Details on proposed usage of PHB encodings
by some MPLS label distribution protocols (RSVP and LDP) for support
of Diff-Serv over MPLS, can be found in [MPLS-DS].
In another approach, the ATM Forum has a requirement to indicate
desired IP QOS treatments in ATM signaling, so that ATM switches can
be just as supportive of the desired service as are IP forwarders.
To do so the Forum is defining a new VC call setup information
element is which will carry PHB identification codes (although will
be generalized to do more if needed).
2. Encoding
PHBs and sets of PHBs are encoded in an unsigned 16 bit binary field.
The 16 bit field is arranged as follows:
Case 1: PHBs defined by standards action, as per [RFC 2474].
The encoding for a single PHB is the recommended DSCP value for that
PHB, left-justified in the 16 bit field, with bits 6 through 15 set
to zero. Note that the recommended DSCP value MUST be used, even if
the network in question has chosen a different mapping.
The encoding for a set of PHBs is the numerically smallest of the set
of encodings for the various PHBs in the set, with bit 14 set to 1.
(Thus for the AF1x PHBs, the encoding is that of the AF11 PHB, with
bit 14 set to 1.)
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
+---+---+---+---+---+---+---+---+---+---+---+---+---+---+---+---+
| DSCP | 0 0 0 0 0 0 0 0 X 0 |
+---+---+---+---+---+---+---+---+---+---+---+---+---+---+---+---+
Case 2: PHBs not defined by standards action, i.e., experimental or
local use PHBs as allowed by [RFC 2474]. In this case an arbitrary
12 bit PHB identification code, assigned by the IANA, is placed
left-justified in the 16 bit field. Bit 15 is set to 1, and bit 14
is zero for a single PHB or 1 for a set of PHBs. Bits 12 and 13 are
zero.
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
+---+---+---+---+---+---+---+---+---+---+---+---+---+---+---+---+
| PHB id code | 0 0 X 1 |
+---+---+---+---+---+---+---+---+---+---+---+---+---+---+---+---+
Bits 12 and 13 are reserved either for expansion of the PHB
identification code, or for other use, at some point in the future.
In both cases, when a single PHBID is used to identify a set of PHBs
(i.e., bit 14 is set to 1), that set of PHBs MUST constitute a PHB
Scheduling Class (i.e., use of PHBs from the set MUST NOT cause
intra-microflow traffic reordering when different PHBs from the set
are applied to traffic in the same microflow). The set of AF1x PHBs
[RFC 2597] is an example of a PHB Scheduling Class. Sets of PHBs
that do not constitute a PHB Scheduling Class can be identified by
using more than one PHBID.
3. Signalling the Class Selector Codepoints
[RFC 2474] defines the eight DS codepoint values of the form 'xxx000'
(where x may be '0' or '1') as the Class Selector Codepoints.
Codepoint 000000 is the recommended DSCP value for the Default PHB,
and hence the Case 1 PHBID constructed from that codepoint is used to
signal the Default PHB (see Section 2 above).
For convenience and consistent operation with networks that employ IP
Precedence [RFC 1812], the Case 1 format PHBIDs constructed from the
other seven Class Selector Codepoints may also be used to signal
PHBs. In each case, the PHB signaled by such a PHBID is the PHB to
which the embedded class selector codepoint (or IP Precedence value
that corresponds to it in non-diffserv domains) is mapped in the
recipient's network. Note that different networks will employ
different mappings; see Section 4 of [RFC 2474] for further
discussion.
Any specified use of PHBIDs SHOULD allow the use of the eight Case 1
PHBIDs constructed from the Class Selector Codepoints.
4. IANA Considerations
IANA is requested to create a new assignment registry for "Per-Hop
Behavior Identification Codes", initially allowing values in the
range 0 to 4095 decimal.
Assignment of values in this field require:
- the identity of the assignee
- a brief description of the new PHB, with enough detail to
distinguish it from existing standardized and non-standardized
PHBs. In the case of a set of PHBs, this description should
cover all PHBs in the set.
- a reference to a stable document describing the PHB in detail.
During the first year of existence of this registry, IANA is
requested to refer all requests to the IETF diffserv WG for review.
Subsequently, requests should be reviewed by the IETF Transport Area
Directors or by an expert that they designate.
If the number of assignments begins to approach 4096, the Transport
Area Directors should be alerted.
5. Security Considerations
This encoding in itself raises no security issues. However, users of
this encoding should consider that modifying a PHB identification
code may constitute theft or denial of service, so protocols using
this encoding must be adequately protected.
Just signalling a PHBID SHOULD NOT be sufficient to grant the sender
access to a PHB that it would otherwise not be able to use. In cases
where this is an issue, receivers SHOULD treat received PHBIDs as
requests for service, and use local policy to determine whether to
grant or deny such requests.
Changes from RFC 2836
[RFC 2836] did not consider the Class Selector code points, which are
covered by section 3 of the present document. A clarification has
been added at the end of section 2 for the case of PHB Scheduling
Classes. The second paragraph of section 5 has been added.
Acknowledgements
Useful comments were made by members of the IETF Diffserv working
group.
References
[RFC 2119] Bradner, S., "Key words for use in RFCs to Indicate
Requirement Levels", BCP 14, RFC 2119, March 1997.
[RFC 2474] Nichols, K., Blake, S., Baker, F. and D. Black,
"Definition of the Differentiated Services Field (DS
Field) in the IPv4 and IPv6 Headers", RFC 2474, December
1998.
[RFC 2475] Blake, S., Black, D., Carlson, M., Davies, E., Wang, Z.
and W. Weiss, "An Architecture for Differentiated
Services", RFC 2475, December 1998.
[RFC 2597] Heinanen, J., Baker, F., Weiss, W. and J. Wroclawski,
"Assured Forwarding PHB Group", RFC 2597, June 1999.
[RFC 2836] Brim, S., Carpenter, B. and F. Le Faucheur, "Per Hop
Behavior Identification Codes", RFC 2836, May 2000.
[MPLS-DS] Le Faucheur, F., et al., "MPLS Support of Differentiated
Services", Work in Progress.
Authors' Addresses
David L. Black
EMC Corporation
42 South St.
Hopkinton, MA 01748
EMail: black_david@emc.com
Scott W. Brim
146 Honness Lane
Ithaca, NY 14850
USA
EMail: sbrim@cisco.com
Brian E. Carpenter
IBM
c/o iCAIR
Suite 150
1890 Maple Avenue
Evanston, IL 60201
USA
EMail: brian@icair.org
Francois Le Faucheur
Cisco Systems
Petra B - Les Lucioles
291, rue Albert Caquot
06560 Valbonne
France
EMail: flefauch@cisco.com
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 implementors 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.
Full Copyright Statement
Copyright (C) The Internet Society (2001). 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.
ISO Plain Language
On this page
§1Why PHBs Need Identification Codes
Differentiated Services, as described in RFC 2474 and RFC 2475, introduces the notion of Per Hop Behaviors (PHBs), which define how traffic belonging to a particular behavior aggregate is treated at an individual network node. In IP packet headers, PHBs are not indicated as such; instead, Differentiated Services Codepoint (DSCP) values are used. There are only 64 possible DSCP values, but there is no such limit on the number of PHBs. In a given network domain there is a locally defined mapping between DSCP values and PHBs. Standardized PHBs recommend a DSCP mapping, but network operators may choose alternative mappings.
In some cases it is necessary or desirable to identify a particular PHB in a protocol message, such as a message negotiating bandwidth management or path selection, especially when such messages pass between management domains. Work in progress includes communication between bandwidth brokers and MPLS support of Diff-Serv.
In certain cases what needs to be identified is not an individual PHB but a set of PHBs. One example is a set of PHBs that must follow the same physical path to prevent re-ordering: the three PHBs belonging to a single Assured Forwarding class, AF11, AF12 and AF13, from RFC 2597.
This document defines a binary encoding to uniquely identify PHBs, or sets of PHBs, in protocol messages, and that encoding MUST be used when such identification is required. It replaces RFC 2836, which omitted the considerations for the Class Selector codepoints.
Packet headers can only carry 64 codepoints, but there are more forwarding behaviours than that, and each network can map codepoints to behaviours its own way. This document gives protocol messages a code of their own, so a request can say exactly which behaviour it wants.
§1.1Where the Codes Are Used
Differentiated Services is expected to be supported over various underlying technologies, which the RFC broadly refers to as link layers. For the transport of IP packets, some of these link layers make use of connections or logical connections where the forwarding behaviour supported by each link layer device is a property of the connection. Within the link layer domain, each link layer node schedules traffic depending on which connection the traffic is transported in. Examples of such link layers include ATM and MPLS.
For efficient support of Diff-Serv over these link layers, one model is for different Behavior Aggregates (BAs), or sets of Behavior Aggregates, to be transported over different connections, so that they are granted different and appropriate forwarding behaviours inside the link layer cloud. When those connections are dynamically established for the transport of Diff-Serv traffic, it is very useful to communicate at connection establishment time what forwarding behaviour or behaviours are to be granted to each connection by the link layer device, so that the BAs transported experience consistent forwarding behaviour inside the link layer cloud. This can be achieved by including the encoding of the corresponding PHB, or set of PHBs, in the connection establishment signalling messages. Proposed use of these encodings by the RSVP and LDP label distribution protocols for Diff-Serv over MPLS is described in a work in progress.
The RFC also describes another approach: the ATM Forum has a requirement to indicate desired IP quality of service treatments in ATM signalling, so that ATM switches can be just as supportive of the desired service as IP forwarders are. To do so, a new VC call setup information element is being defined that will carry PHB identification codes.
The code is meant for signalling, not for packet headers. When a link layer such as ATM or MPLS sets up a connection to carry traffic, the signalling message can carry this code so the switch at the far end knows what treatment to apply.
§2The 16-Bit Encoding
PHBs and sets of PHBs are encoded in an unsigned 16-bit binary field, arranged in one of two cases.
Case 1 covers PHBs defined by standards action, as per RFC 2474. The encoding for a single PHB is the recommended DSCP value for that PHB, left-justified in the 16-bit field, with bits 6 through 15 set to zero. The recommended DSCP value MUST be used, even if the network in question has chosen a different mapping. The encoding for a set of PHBs is the numerically smallest of the encodings for the various PHBs in the set, with bit 14 set to 1. Thus for the AF1x PHBs, the encoding is that of the AF11 PHB with bit 14 set to 1.
Case 2 covers PHBs not defined by standards action, that is, experimental or local use PHBs as allowed by RFC 2474. An arbitrary 12-bit PHB identification code, assigned by IANA, is placed left-justified in the 16-bit field. Bit 15 is set to 1, and bit 14 is zero for a single PHB or 1 for a set of PHBs. Bits 12 and 13 are zero; they are reserved either for expansion of the PHB identification code, or for other use, at some point in the future.
The 16 bits hold either a standard behaviour or a locally assigned one. Bit 15 says which case it is, and bit 14 says whether the code names one behaviour or a group of them.
| Field | Name | Size | Value | Description |
|---|---|---|---|---|
DSCP | DSCP | 6 bits | The recommended DSCP value for the PHB, left-justified. | |
ZEROS | Zero | 1 byte | Bits 6 through 13 are set to zero. | |
SET-FLAG | Set flag | 1 bits | Bit 14. One identifies a set of PHBs, zero a single PHB. | |
CASE-FLAG | Case flag | 1 bits | Bit 15. Zero for a PHB defined by standards action. |
| Field | Name | Size | Value | Description |
|---|---|---|---|---|
PHB-ID | PHB id code | 12 bits | A 12-bit PHB identification code assigned by IANA, left-justified. | |
RESERVED | Reserved | 2 bits | Bits 12 and 13 are zero, reserved for future expansion. | |
SET-FLAG | Set flag | 1 bits | Bit 14. One identifies a set of PHBs, zero a single PHB. | |
CASE-FLAG | Case flag | 1 bits | Bit 15. One for a PHB not defined by standards action. |
| Field | Case 1 | Case 2 |
|---|---|---|
| Who defines the PHB | Standards action, per RFC 2474 | Experimental or local use |
| Bits 0 to 5 | The recommended DSCP value | The first part of the 12-bit id code |
| Bits 6 to 13 | Set to zero | The rest of the id code, with bits 12 and 13 zero |
| Bit 14 | One for a set, zero for one PHB | One for a set, zero for one PHB |
| Bit 15 | Zero | One |
§2Sets Must Be Scheduling Classes
In both cases, when a single PHBID is used to identify a set of PHBs, that is, when bit 14 is set to 1, the set of PHBs MUST constitute a PHB Scheduling Class. That means the use of PHBs from the set MUST NOT cause intra-microflow traffic reordering when different PHBs from the set are applied to traffic in the same microflow. The set of AF1x PHBs is an example of a PHB Scheduling Class.
Sets of PHBs that do not constitute a PHB Scheduling Class can be identified by using more than one PHBID.
One code can only name a group of behaviours that are safe to use together, meaning packets of the same flow must not be reordered by mixing them. A group that does not meet that rule has to be named with several codes instead.
§3Signalling the Class Selector Codepoints
RFC 2474 defines the eight DS codepoint values of the form 'xxx000', where x may be 0 or 1, as the Class Selector Codepoints. Codepoint 000000 is the recommended DSCP value for the Default PHB, and hence the Case 1 PHBID constructed from that codepoint is used to signal the Default PHB.
For convenience and consistent operation with networks that employ IP Precedence, as in RFC 1812, the Case 1 format PHBIDs constructed from the other seven Class Selector Codepoints may also be used to signal PHBs. In each case, the PHB signalled by such a PHBID is the PHB to which the embedded class selector codepoint, or the IP Precedence value that corresponds to it in non-DiffServ domains, is mapped in the recipient's network. Different networks employ different mappings, and RFC 2474 discusses this further.
Any specified use of PHBIDs SHOULD allow the use of the eight Case 1 PHBIDs constructed from the Class Selector Codepoints.
Eight of the Diff-Serv codepoints were reserved for class selectors, and the zero value among them stands for the default behaviour. The RFC allows those eight codes to be used to name behaviours too, so that old networks that only understand IP precedence still work.
§4The IANA Registry
IANA is requested to create a new assignment registry for 'Per-Hop Behavior Identification Codes', initially allowing values in the range 0 to 4095 decimal.
Assignment of values in this field requires the identity of the assignee, a brief description of the new PHB with enough detail to distinguish it from existing standardized and non-standardized PHBs, which in the case of a set of PHBs should cover all the PHBs in the set, and a reference to a stable document describing the PHB in detail.
During the first year of existence of this registry, IANA is requested to refer all requests to the IETF DiffServ working group for review. Subsequently, requests should be reviewed by the IETF Transport Area Directors or by an expert they designate. If the number of assignments begins to approach 4096, the Transport Area Directors should be alerted.
Codes for locally defined behaviours are handed out from a public registry, starting at 4095 and working down. Each request has to describe the new behaviour clearly enough that it cannot be confused with an existing one.
§5Security Considerations
The RFC says the encoding in itself raises no security issues. Users of the encoding should, however, consider that modifying a PHB identification code may constitute theft or denial of service, so protocols that use this encoding must be adequately protected.
Just signalling a PHBID SHOULD NOT be sufficient to grant the sender access to a PHB that it would otherwise not be able to use. Where this is an issue, receivers SHOULD treat received PHBIDs as requests for service, and use local policy to decide whether to grant or deny such requests.
The code itself is harmless, but changing one in flight could let someone take traffic they should not have. So a receiver should treat a requested behaviour as a request and apply its own policy before granting it.
§5Changes From RFC 2836
RFC 2836 did not consider the Class Selector code points, which are covered by Section 3 of the present document. A clarification has been added at the end of Section 2 for the case of PHB Scheduling Classes, and the second paragraph of Section 5 has been added.
The earlier version left out the class selector codepoints, and this version adds them along with a clearer rule about which groups of behaviours may share one code.