PCN Signaling Needs: Requirements for Signaling of Pre-Congestion Information in a Diffserv Domain
Requirements for pre-congestion notification signaling in Diffserv networks.
Karagiannis, G., Taylor, T., Chan, K., Menth, M., and P. Eardley, "Requirements for Signaling of Pre-Congestion Information in a Diffserv Domain", RFC 6663, DOI 10.17487/RFC6663, July 2012, <https://www.rfc-editor.org/info/rfc6663>.
@misc{rfc6663,
series = {Request for Comments},
number = 6663,
howpublished = {RFC 6663},
publisher = {RFC Editor},
doi = {10.17487/RFC6663},
url = {https://www.rfc-editor.org/info/rfc6663},
author = {G. Karagiannis and T. Taylor and K. Chan and M. Menth and P. Eardley},
title = {{Requirements for Signaling of Pre-Congestion Information in a Diffserv Domain}},
pagetotal = 7,
year = 2012,
month = jul,
} RFC text
Contents
Internet Engineering Task Force (IETF) G. Karagiannis
Request for Comments: 6663 University of Twente
Category: Informational T. Taylor
ISSN: 2070-1721 Huawei
K. Chan
Consultant
M. Menth
University of Tuebingen
P. Eardley
BT
July 2012
Requirements for Signaling of Pre-Congestion Information
in a Diffserv Domain
Abstract
Pre-Congestion Notification (PCN) is a means for protecting quality
of service for inelastic traffic admitted to a Diffserv domain. The
overall PCN architecture is described in RFC 5559. This memo
describes the requirements for the signaling applied within the PCN-
domain: (1) PCN-feedback-information is carried from the PCN-egress-
node to the Decision Point; (2) the Decision Point may ask the PCN-
ingress-node to measure, and report back, the rate of sent PCN-
traffic between that PCN-ingress-node and PCN-egress-node. The
Decision Point may be either collocated with the PCN-ingress-node or
a centralized node (in the first case, (2) is not required). The
signaling requirements pertain in particular to two edge behaviors,
Controlled Load (CL) and Single Marking (SM).
Status of This Memo
This document is not an Internet Standards Track specification; it is
published for informational purposes.
This document is a product of the Internet Engineering Task Force
(IETF). It represents the consensus of the IETF community. It has
received public review and has been approved for publication by the
Internet Engineering Steering Group (IESG). Not all documents
approved by the IESG are a candidate for any level of Internet
Standard; see Section 2 of RFC 5741.
Information about the current status of this document, any errata,
and how to provide feedback on it may be obtained at
http://www.rfc-editor.org/info/rfc6663.
Copyright Notice
Copyright (c) 2012 IETF Trust and the persons identified as the
document authors. All rights reserved.
This document is subject to BCP 78 and the IETF Trust's Legal
Provisions Relating to IETF Documents
(http://trustee.ietf.org/license-info) in effect on the date of
publication of this document. Please review these documents
carefully, as they describe your rights and restrictions with respect
to this document. Code Components extracted from this document must
include Simplified BSD License text as described in Section 4.e of
the Trust Legal Provisions and are provided without warranty as
described in the Simplified BSD License.
Table of Contents
1. Introduction ....................................................2
1.1. Requirements Language ......................................3
2. Signaling Requirements for Messages from the PCN-Egress-Nodes
to Decision Point(s) ............................................3
3. Signaling Requirements for Messages between Decision Point(s)
and PCN-Ingress-Nodes ...........................................5
4. Security Considerations .........................................5
5. Acknowledgments .................................................6
6. References ......................................................6
6.1. Normative References .......................................6
6.2. Informative References .....................................6
1. Introduction
The main objective of Pre-Congestion Notification (PCN) is to support
the quality of service (QoS) of inelastic flows within a Diffserv
domain in a simple, scalable, and robust fashion. Two mechanisms are
used: admission control and flow termination. Admission control is
used to decide whether to admit or block a new flow request, while
flow termination is used in abnormal circumstances to decide whether
to terminate some of the existing flows. To support these two
features, the overall rate of PCN-traffic is metered on every link in
the domain, and PCN-packets are appropriately marked when certain
configured rates are exceeded. These configured rates are below the
rate of the link, thus providing notification to boundary nodes about
overloads before any congestion occurs (hence "pre-congestion"
notification). The PCN-egress-nodes measure the rates of differently
marked PCN traffic in periodic intervals and report these rates to
the Decision Points for admission control and flow termination; the
Decision Points use these rates to make decisions. The Decision
Points may be collocated with the PCN-ingress-nodes, or their
function may be implemented in a centralized node. For more details
see [RFC5559], [RFC6661], and [RFC6662].
This memo specifies the requirements on signaling protocols:
o to carry reports from a PCN-egress-node to the Decision Point,
o to carry requests, from the Decision Point to a PCN-ingress-node,
that trigger the PCN-ingress-node to measure the PCN-sent-rate,
o to carry reports, from a PCN-ingress-node to the Decision Point.
The latter two messages are only needed if the Decision Point and
PCN-ingress-node are not collocated.
1.1. Requirements Language
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 RFC 2119 [RFC2119].
2. Signaling Requirements for Messages from the PCN-Egress-Nodes to
Decision Point(s)
The PCN-egress-node measures per ingress-egress-aggregate the rates
of differently marked PCN-traffic in regular intervals. The
measurement intervals are recommended to take a fixed value between
100 ms and 500 ms; see [RFC6661] and [RFC6662]. At the end of each
measurement interval, the PCN-egress-node calculates the congestion-
level-estimate (CLE) based on these quantities.
The PCN-egress-node MAY be configured to record a set of identifiers
of PCN-flows for which it received excess-traffic-marked packets
during the last measurement interval. The latter may be useful to
perform flow termination in networks with multipath routing.
At the end of each measurement interval, or less frequently if
"optional report suppression" is activated (see [RFC6661] and
[RFC6662]), the PCN-egress-node sends a report to the Decision Point.
For the SM edge behavior, the report MUST contain:
o the identifier of the PCN-ingress-node and the identifier of the
PCN-egress-node (typically their IP addresses); together they
specify the ingress-egress-aggregate to which the report refers,
o the rate of not-marked PCN-traffic (NM-rate) in octets/second, and
o the rate of PCN-marked traffic (PM-rate) in octets/second.
For the CL edge behavior, the report MUST contain:
o the identifier of the PCN-ingress-node and the identifier of the
PCN-egress-node (typically their IP addresses); together they
specify the ingress-egress-aggregate to which the report refers,
o the rate of not-marked PCN-traffic (NM-rate) in octets/second,
o the rate of threshold-marked PCN traffic (ThM-rate) in
octets/second, and
o the rate of excess-traffic-marked traffic (ETM-rate) in
octets/second.
The number format and the rate units used by the signaling protocol
will limit the maximum rate that PCN can use. If signaling space is
tight, it might be reasonable to impose a limit, but any such limit
may impose unnecessary constraints in the future.
The signaling report can either be sent directly to the Decision
Point or it can "piggy-back", i.e., be included within some other
message that passes through the PCN-egress-node and then reaches the
Decision Point.
As described in [RFC6661], PCN reports from the PCN-egress-node to
the Decision Point may contain flow identifiers for individual flows
within an ingress-egress-aggregate that have recently experienced
excess-marking. Hence, the PCN report messages used by the PCN CL
edge behavior MUST be capable of carrying sequences of octet strings
constituting such identifiers.
Signaling messages SHOULD have a higher priority and a lower drop
precedence than PCN-packets (see [RFC5559]) in order to deliver them
quickly and to prevent them from being dropped in case of overload.
The load generated by the signaling protocol SHOULD be minimized. We
give three methods that may help to achieve that goal:
1. piggy-backing the reports by the PCN-egress-nodes to the Decision
Point(s) onto other signaling messages that are already in place,
2. reducing the amount of reports to be sent by optional report
suppression, or
3. combining reports for different ingress-egress-aggregates in a
single message (if they are for the same Decision Point).
As PCN reports are sent regularly, additional reliability mechanisms
are not needed. This also holds in the presence of optional report
suppression, as reports are sent periodically if actions by the
Decision Point(s) are needed; see [RFC6661] and [RFC6662].
3. Signaling Requirements for Messages between Decision Point(s) and
PCN-Ingress-Nodes
Through request-response signaling between the Decision Point and
PCN-ingress-node, the Decision Point requests and in response the
PCN-ingress-node measures and reports the PCN-sent-rate for a
specific ingress-egress-aggregate. Signaling is needed only if the
Decision Point and PCN-ingress-node are not collocated.
The request MUST contain:
o the identifier of the PCN-ingress-node and the identifier of the
PCN-egress-node; together they determine the ingress-egress-
aggregate for which the PCN-sent-rate is requested, and
o the identifier of the Decision Point that requests the PCN-sent-
rate.
The report MUST contain:
o the PCN-sent-rate in octets/second, and
o the identifier of the PCN-ingress-node and the identifier of the
PCN-egress-node.
The request MUST be addressed to the PCN-ingress-node, and the report
MUST be addressed to the Decision Point that requested it.
Because they are sent only when flow termination is needed (which is
an urgent action), the request and the report SHOULD be sent with
high priority, with a lower drop precedence than PCN-packets, and in
a reliable manner.
Note that a complete system description for a PCN-domain with
centralized Decision Point includes the signaling from Decision Point
to the PCN-ingress-nodes to control flow admission and termination.
However, this is a known problem (with solutions provided in
[RFC3084] and [RFC5431], for example), and it lies outside the scope
of the present document.
4. Security Considerations
[RFC5559] provides a general description of the security
considerations for PCN. This memo relies on the security-related
requirements of the PCN signaling, provided in [RFC5559]. In
particular, the signaling between the PCN-boundary-nodes must be
protected from attacks. For example, the recipient needs to validate
that the message is indeed from the node that claims to have sent it.
Possible measures include digest authentication and protection
against replay and man-in-the-middle attacks.
Specifically for the generic aggregate RSVP protocol, additional
protection methods against security attacks are described in
[RFC4860].
5. Acknowledgments
We would like to acknowledge the members of the PCN working group for
the discussions that produced the contents of this memo.
6. References
6.1. Normative References
[RFC2119] Bradner, S., "Key words for use in RFCs to Indicate
Requirement Levels", BCP 14, RFC 2119, March 1997.
[RFC5559] Eardley, P., Ed., "Pre-Congestion Notification (PCN)
Architecture", RFC 5559, June 2009.
[RFC6661] Charny, A., Huang, F., Karagiannis, G., Twente, U., Menth,
M., and T. Taylor, Ed., "Pre-Congestion Notification (PCN)
Boundary-Node Behaviour for the Controlled Load (CL) Mode
of Operation", RFC 6661, July 2012.
[RFC6662] Charny, A., Zhang, J., Karagiannis, G., Twente, U., Menth,
M., and T. Taylor, Ed., "Pre-Congestion Notification (PCN)
Boundary-Node Behaviour for the Single Marking (SM) Mode
of Operation", RFC 6662, July 2012.
6.2. Informative References
[RFC3084] Chan, K., Seligson, J., Durham, D., Gai, S., McCloghrie,
K., Herzog, S., Reichmeyer, F., Yavatkar, R., and A.
Smith, "COPS Usage for Policy Provisioning (COPS-PR)", RFC
3084, March 2001.
[RFC4860] Le Faucheur, F., Davie, B., Bose, P., Christou, C., and M.
Davenport, "Generic Aggregate Resource ReSerVation
Protocol (RSVP) Reservations", RFC 4860, May 2007.
[RFC5431] Sun, D., "Diameter ITU-T Rw Policy Enforcement Interface
Application", RFC 5431, March 2009.
Authors' Addresses
Georgios Karagiannis
University of Twente
P.O. Box 217
7500 AE Enschede,
The Netherlands
EMail: g.karagiannis@utwente.nl
Tom Taylor
Huawei Technologies
Ottawa
Canada
EMail: tom.taylor.stds@gmail.com
Kwok Ho Chan
Consultant
EMail: khchan.work@gmail.com
Michael Menth
University of Tuebingen
Sand 13
72076 Tuebingen
Germany
Phone: +49-7071-2970505
EMail: menth@uni-tuebingen.de
Philip Eardley
BT
B54/77, Sirius House Adastral Park Martlesham Heath
Ipswich, Suffolk IP5 3RE
United Kingdom
EMail: philip.eardley@bt.com
AI ELI5
On this page
§1Quality of Service in Diffserv
Pre-Congestion Notification (PCN) is a mechanism to protect quality of service for inelastic flows—like Voice over IP (VoIP), video conferencing, and streaming—within a Diffserv domain. PCN uses two techniques: admission control, which decides whether to admit a new flow request, and flow termination, which drops existing flows when the network is overloaded. The overall PCN architecture is defined in RFC 5559. This document specifies the signaling requirements needed for PCN to work: how information flows between the network edge and the decision points that make admission and termination decisions.
The network watches for danger signs and tells a decision maker, who then says yes or no to new customers, or asks some current customers to leave.
§2PCN Components and Operation
PCN operates on three main components. PCN-boundary-nodes mark RTP or RTCP packets based on how much traffic they see. At PCN-egress-nodes (the exit points), the marked traffic is measured in regular intervals (typically 100 to 500 milliseconds). At the Decision Point, these measurements are used to make admission and termination decisions. The Decision Point may be collocated with a PCN-ingress-node (the entry point), in which case it has direct access to information, or it may be a centralized node. PCN supports two edge behaviors: Controlled Load (CL), which uses threshold marking and excess-traffic marking, and Single Marking (SM), which uses a single marking level. Each approach has different signaling requirements.
Watch stations on the way out measure the load. A decision maker gets these counts. If they are far away, the entry point also has to send what it sent out, so the decision maker can compare.
§3Feedback from the Network Edge
The PCN-egress-node must report its measurements to the Decision Point regularly, typically at the end of each measurement interval, unless optional report suppression is enabled (in which case reports are sent less frequently if the network state is stable). For Single Marking (SM) mode, each report must contain the not-marked traffic rate (in octets per second) and the PCN-marked traffic rate. For Controlled Load (CL) mode, the report must include the not-marked rate, the threshold-marked rate, and the excess-traffic-marked rate. Both reports must identify the ingress and egress nodes to specify which path the measurement applies to. The report can be sent directly to the Decision Point or carried within another signaling message that passes through the egress node. To minimize signaling load, multiple reports for different paths can be combined in a single message. Signaling messages SHOULD be prioritized above regular PCN-traffic to ensure timely delivery.
The edge sends counts of unmarked and marked cars every few seconds. If the decision maker is far away, it needs the ingress point's address too.
§4Measurements from the Ingress Node
When the Decision Point is not collocated with the PCN-ingress-node, the Decision Point may request the ingress node to measure and report the PCN-sent-rate—the total rate of PCN traffic sent toward a specific path. This signaling is only needed if the Decision Point and ingress node are separate. The request must specify the ingress and egress nodes that define the path, and the identity of the Decision Point requesting the measurement. The ingress node responds with the PCN-sent-rate in octets per second and the path identifiers. Since these requests are only sent when flow termination is urgently needed, they SHOULD be sent with high priority, low drop precedence, and in a reliable manner to ensure the message gets through.
The decision maker can ask the entry point, 'How much are you sending down the road to exit point X?' and waits for an answer.
§5Trust and Protection
PCN signaling carries network management information that could disrupt service if spoofed or modified. The signaling between PCN-boundary-nodes must be protected from attacks. Messages must be authenticated so the recipient can verify that the sender is who it claims to be. The signaling should be protected against replay attacks (where an attacker replays an old message) and man-in-the-middle attacks (where an attacker intercepts and modifies messages). Possible mechanisms include digest authentication and encryption. For generic aggregate RSVP signaling (used in some deployments), RFC 4860 provides additional security methods. The specific security mechanisms are outside the scope of this requirements document; they are handled by the signaling protocol that carries PCN feedback.
Make sure the person sending the counts is real, not someone pretending, and the counts have not been changed on the way.