One document matched: draft-westberg-pcn-load-control-03.txt
Differences from draft-westberg-pcn-load-control-02.txt
PCN L. Westberg
Internet-Draft A. Bhargava
Intended status: Standards Track A. Bader
Expires: August 28, 2008 Ericsson
G. Karagiannis
University of Twente
February 25, 2008
LC-PCN: The Load Control PCN Solution
draft-westberg-pcn-load-control-03
Status of this Memo
By submitting this Internet-Draft, each author represents that any
applicable patent or other IPR claims of which he or she is aware
have been or will be disclosed, and any of which he or she becomes
aware will be disclosed, in accordance with Section 6 of BCP 79.
Internet-Drafts are working documents of the Internet Engineering
Task Force (IETF), its areas, and its working groups. Note that
other groups may also distribute working documents as Internet-
Drafts.
Internet-Drafts are draft documents valid for a maximum of six months
and may be updated, replaced, or obsoleted by other documents at any
time. It is inappropriate to use Internet-Drafts as reference
material or to cite them other than as "work in progress."
The list of current Internet-Drafts can be accessed at
http://www.ietf.org/ietf/1id-abstracts.txt.
The list of Internet-Draft Shadow Directories can be accessed at
http://www.ietf.org/shadow.html.
This Internet-Draft will expire on August 28, 2008.
Copyright Notice
Copyright (C) The IETF Trust (2008).
Abstract
There is an increased interest of simple and scalable resource
provisioning solution for Diffserv network. The Load Control PCN
(LC-PCN) addresses the following issues:
Westberg, et al. Expires August 28, 2008 [Page 1]
Internet-Draft LC-PCN February 2008
o Admission Control for real time data flows in stateless Diffserv
Domains
o Flow Termination: Termination of flows in case of exceptional
events, such as severe congestion after re-routing.
Admission control in a Diffserv stateless domain can be performed
using two methods:
o Admission Control based on data marking, whereby in congestion
situations the data packets are marked to notify the PCN-egress-
node that a congestion occurred on a particular PCN-ingress-node
to PCN-egress-node path.
o Probing, whereby a probe packet is sent along the forwarding path
in a network to determine whether a flow can be admitted based
upon the current congestion state of the network
The scheme provides the capability of controlling the traffic load in
the network without requiring signaling or any per-flow processing in
the PCN-interior-nodes. The complexity of Load Control is kept to a
minimum to make implementation simple.
Westberg, et al. Expires August 28, 2008 [Page 2]
Internet-Draft LC-PCN February 2008
Table of Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 5
2. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . 6
3. LC-PCN Overview . . . . . . . . . . . . . . . . . . . . . . . 6
3.1. Admission control . . . . . . . . . . . . . . . . . . . . 6
3.1.1. Admission control based on data marking . . . . . . . 7
3.1.1.1. PCN-interior-node features . . . . . . . . . . . . 7
3.1.1.2. PCN-egress-node features . . . . . . . . . . . . . 8
3.1.1.3. PCN-ingress-node features . . . . . . . . . . . . 9
3.1.2. Admission control based on probing . . . . . . . . . . 9
3.1.2.1. PCN-interior-node features . . . . . . . . . . . . 9
3.1.2.2. PCN-egress-node features . . . . . . . . . . . . . 10
3.1.2.3. PCN-ingress-node features . . . . . . . . . . . . 10
3.1.3. ECMP solution . . . . . . . . . . . . . . . . . . . . 10
3.2. Flow Termination . . . . . . . . . . . . . . . . . . . . . 11
3.2.1. PCN-interior-node . . . . . . . . . . . . . . . . . . 11
3.2.2. PCN-egress-node . . . . . . . . . . . . . . . . . . . 11
3.2.3. PCN-ingress-node . . . . . . . . . . . . . . . . . . . 12
3.2.4. ECMP solution . . . . . . . . . . . . . . . . . . . . 13
3.3. Operational states in LC-PCN . . . . . . . . . . . . . . . 13
3.3.1. Operational states in PCN-interior-nodes . . . . . . . 13
3.3.2. Operational states in PCN-egress-nodes . . . . . . . . 15
3.3.3. Operational states in PCN-ingress-nodes . . . . . . . 17
3.4. Encoding of PCN traffic . . . . . . . . . . . . . . . . . 18
4. LC-PCN detailed description . . . . . . . . . . . . . . . . . 19
4.1. Admission control based on data marking for
unidirectional flows . . . . . . . . . . . . . . . . . . . 19
4.1.1. Operation in PCN-interior-nodes . . . . . . . . . . . 19
4.1.2. Operation in PCN-egress-nodes . . . . . . . . . . . . 21
4.1.3. Operation in PCN-ingress-nodes . . . . . . . . . . . . 23
4.2. Admission control based on probing for unidirectional
flows . . . . . . . . . . . . . . . . . . . . . . . . . . 23
4.2.1. Operation in PCN-interior-nodes . . . . . . . . . . . 24
4.2.2. Operation in PCN-egress-nodes . . . . . . . . . . . . 24
4.2.3. Operation in PCN-ingress-nodes . . . . . . . . . . . . 25
4.3. Flow Termination for unidirectional flows . . . . . . . . 25
4.3.1. Operation in the PCN-interior-node . . . . . . . . . . 25
4.3.1.1. Optimization mode features for Flow termination . 26
4.3.1.2. ECMP solutions . . . . . . . . . . . . . . . . . . 28
4.3.2. Operation in PCN-egress-nodes . . . . . . . . . . . . 28
4.3.2.1. ECMP solutions . . . . . . . . . . . . . . . . . . 31
4.3.3. Operation in PCN-ingress-nodes . . . . . . . . . . . . 31
4.4. Admission control based on data marking and probing
for bi-directional flows . . . . . . . . . . . . . . . . . 32
4.5. Flow Termination handling for bi-directional flows . . . . 33
5. Security Considerations . . . . . . . . . . . . . . . . . . . 38
6. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 38
Westberg, et al. Expires August 28, 2008 [Page 3]
Internet-Draft LC-PCN February 2008
7. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . 38
8. Informative References . . . . . . . . . . . . . . . . . . . . 38
Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 40
Intellectual Property and Copyright Statements . . . . . . . . . . 41
Westberg, et al. Expires August 28, 2008 [Page 4]
Internet-Draft LC-PCN February 2008
1. Introduction
The amount of traffic carried on the Internet is now greater than the
traffic on the world's telephony network. Still, Internet-based
communication services generate less income than plain old telephony
services. Enabling value-added services over the Internet is
therefore crucial for service providers. One significant class of
such value-added services requires real-time packet transportation.
It can be expected that these real-time services will be popular as
they replicate or are natural extensions of existing communication
services like telephony. Exact and reliable resource management
(e.g., admission control) is essential for achieving high utilization
in networks with real-time transportation capabilities. The problem
is difficult mainly due to scalability issues.
With the introduction of differentiated services (DS) [RFC2475], it
is now possible to provide large scale, real-time services. The
basic idea of DiffServ is that, rather than classifying packets at
each router, packets are only classified at the edge devices. The
result - the required packet treatment - is stored and carried in the
packet headers, and core routers can carry out appropriate
scheduling.
The current definition of DiffServ, however, does not contain any
simple, scalable solution to the problem of resource provisioning and
control. A number of approaches to solving the problem already exist
[RFC3175], [Berson97], [Stoica99], [Bernet99]. The scheme presented
in this document does not require any state aggregation in the core
and aims at extreme simplicity and low cost of implementation along
with good scaling properties. Load control operates edge-to-edge in
a DS domain, or between two RSVP or NSIS capable routers, where only
the edge devices keep flow state and do per-flow processing. The
main purpose of Load Control is to provide a simple and scalable
solution to the resource provisioning problem.
The original Load Control concept, submitted in April 2000,
[Westberg00], has been developed further to a signaling concept named
Resource Management in Diffserv. RMD was incorporated by NSIS
working group, where the protocol details were worked out for using
NSIS as external protocol [RMD]. Recently new drafts have been
submitted aiming to standardize new Diffserv PHB that provides
controlled load services in Diffserv domains [CL-PHB], [CL-ARCH],
[Babi07], [Char07]. These concepts are very similar to the original
two-bit marking scheme of Load Control.
We believe that the LC-PCN features supported by, at least, PCN-
interior-nodes can be combined with features supported by the above
listed concepts.
Westberg, et al. Expires August 28, 2008 [Page 5]
Internet-Draft LC-PCN February 2008
This document aims to develop a common framework that could be used
both with RSVP and NSIS external protocols.
The remainder of this draft is structured as follows. After the
terminology in Section 2, we give an overview of the LC-PCN in
Section 3. In Section 4 we give a detailed description of the LC-
PCN. Section 5 discusses security issues.
2. Terminology
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. The terms
specified in [Eard08] are used.
3. LC-PCN Overview
Load Control PCN (LC-PCN) is achieved by two actions: Admission
Control and/or Flow Termination. The LC-PCN can be applied within
either a single PCN domain, see Figure 1, or multiple neighboring PCN
domains, when a trust relationship exists between these multiple PCN
domains.
PCN-Ingress-Node PCN-Egress-Node
(PCN-Interior-Nodes; I-Nodes)
| | |
| | |
V V V
+-------+ Data +------+ +------+ +------+ +------+
|-------|--------|------|------|------|-------|------|---->|------|
| | Flow | | | | | | | |
|Ingress| |I-Node| |I-Node| |I-Node| |Egress|
| | | | | | | | | |
+-------+ +------+ +------+ +------+ +------+
=================================================>
<=================================================
Signaling
Figure 1: Actors in the LC-PCN
3.1. Admission control
Admission control can be accomplished in LC-PCN in two ways:
Westberg, et al. Expires August 28, 2008 [Page 6]
Internet-Draft LC-PCN February 2008
o Admission control based on data marking: whereby in congestion
situations, the admission control is accomplished using excess
rate marking and metering to detect and to decide either a new
flow request should be accepted or denied.
o Admission control based on probing: where probing is required to
accomplish the admission control procedure.
Note that the two admission control features can be used either
independently or combined. The Admission Control features can be
applied to flows that are aggregated between PCN-ingress-nodes and
PCN-egress-nodes. In this way edge-to-edge (i.e., ingress-egress)
pair PCN aggregates can be maintained by PCN-ingress-nodes and PCN-
egress-nodes. Two PCN-domain-wide constraints are used. One of them
is denoted as "N", used to indicate the proportionality between the
measured out of profile packets (or bytes) and the remarked packets
(or bytes). If "N" is used in the algorithm, then it must have the
same value in all Diffserv nodes that use this mechanism. The
parameter N is higher or equal to 1 (N >= 1).
Another PCN-domain-wide constraint, see [Char07], has to be used on
the ratio U between the configured-admissible-rate on a link and the
level of PCN load on the link that should trigger the Flow
Termination. This level represents the configured-termination-rate,
which is not explicitly configured on the PCN_interior node. The
value is typically set to U = 1,2, see [Char07].
Furthermore, it is important to note that in this draft we denote the
not congested PCN packets (or bytes) as PCN unmarked packets (or
bytes).
3.1.1. Admission control based on data marking
The admission control based on data marking is using features located
at the PCN_ingress_edge, PCN-interior-node and PCN-egress-node.
3.1.1.1. PCN-interior-node features
The PCN-interior-node performs measurements on the PHB aggregated PCN
traffic. When the PCN-interior-node detects that the measured PHB
aggregated PCN traffic is higher than a preconfigured threshold, say
configured-admissible-rate, then it is considered that the PCN-
interior-node changes operational state from Normal state to
Admission Control state, see Section 3.3.1. Furthermore, the
measured PHB aggregated PCN traffic rate that is above the
configured-admissible-rate is considered to be excess rate, which is
marked using PCN_marking.
Westberg, et al. Expires August 28, 2008 [Page 7]
Internet-Draft LC-PCN February 2008
This can be accomplished using different metering and marking
features. It is important to note that the excess rate measurements
SHOULD be done before a queuing mechanism used by a PCN-interior-
node, drop packets before/during buffer overflow. The constant N
should be used such that the marked excess rate can represent also
high levels of excess rate. This means that before marking the
excess rate, the measured excess rate should be divided by N (when N
>= 1). This can be e.g., implemented by marking every N-th packet
(or byte) instead of marking each packet (or byte).
The PCN_marking SHOULD be done after the queuing mechanism drops the
packets before/during buffer overflow. Several implementation
alternatives of this algorithm are possible. One implementation
alternative can be based on the algorithms discussed in [Char07]. In
particular, a token bucket can be used with the rate configured with
the rate equal to configured-admissible-rate. Other implementation
alternatives can e.g., be based on rate measurements and marking. In
particular, the PCN-interior-nodes packets are using the PCN_marking,
whenever the measured PHB aggregated PCN traffic rate exceeds a pre-
configured rate threshold denoted as configurable-admissible-rate.
It is important to note that, for optimization purposes, the
PCN_marking encoded packets SHOULD NOT be preferentially dropped by
queuing mechanisms in PCN-interior-nodes. This can be accomplished
by for example using two virtual queues. One virtual queue can be
used for the PCN unmarked traffic (and PCN_Affected_Marking encoded,
when the affected marking solution is supported) and another one for
PCN_marking encoded packets. The virtual queues can for example, be
implemented by using one Drop Tail physical queue and by maintaining
queuing information and also one queuing threshold for each of the
virtual queues. The physical queue uses the same scheduling
algorithm, but the length of each of the virtual queue defines the
packet dropping probability of a virtual queue.
3.1.1.2. PCN-egress-node features
The PCN-egress-node measures the rate of the received PHB aggregated
PCN unmarked and PCN_marking encoded packets. Based on these
measurements, the PCN-egress-node can use a similar functionality as
the one specified in [Char07] and [CL-ARCH] to calculate the
Congestion Level Estimate (CLE), which is the fraction of the marked
traffic received from one PCN-ingress-node.
If the value of CLE is higher than a certain value, e.g., 1%, then
the PCN-egress-node is changing its operational state from Normal
state to Admission Control state. By using an external to PCN,
signaling protocol the admission control procedure is accomplished by
using a combination of the PCN operational state of the PCN-egress-
Westberg, et al. Expires August 28, 2008 [Page 8]
Internet-Draft LC-PCN February 2008
node and an admission control request provided by the external to
PCN, signaling protocol. When the admission control request arrives
at a PCN-egress-node that operates in Admission Control state then
the request is rejected. If it operates in Normal state it is
accepted.
3.1.1.3. PCN-ingress-node features
If the external to PCN signaling protocol is also used by the PCN-
ingress-node, then the PCN-ingress-node SHOULD be informed that an
admission control request has been admitted or rejected by the PCN-
egress-node. If the PCN-ingress-node is notified that the admission
request is rejected, then the PCN-ingress-node rejects the admission
control request. Otherwise it is accepted.
3.1.2. Admission control based on probing
The admission control function based on probing can be used to
implement a simple measurement-based admission control within a PCN
domain. The main reason of why this admission control feature should
be used is to solve the possible ECMP (Equal Cost Multi-Path) issue.
Furthermore, this feature can provide admission control support even
when the edge-to-edge pair PCN aggregate is not yet initiated at one
of the edges.
3.1.2.1. PCN-interior-node features
The PCN-interior-node features that are used to detect the PCN
operational states are the same as the ones described in Section
3.1.1.1. In this scenario an IP packet is used as a probe packet,
meaning that the DSCP (and/or ECN) field, see Section 3.4, in the
header of the IP packet is re-marked when the measured PHB aggregated
PCN traffic rate exceeds a predefined congestion threshold, i.e,
configured-admissible-rate. Note that a message used by an external,
to PCN, on path signaling protocol, e.g., RSVP, can be used as a
probe packet.
The PCN-interior-nodes SHOULD detect a probe packet by observing the
Router Alert option, which is carried by the IP packet data packet.
Note that a PCN-ingress-node sets the Router Alert option of all
packets that are used as probe packets. This also holds for
signaling protocol messages (e.g. RSVP PATH message) that are used
by LC-PCN as probe packets. Thus if a PCN-interior-node receives a
probe packet then, due to the Router Alert option it has to handle it
differently than the user packets.
The PCN-interior-node has to PCN_marking encode the probe packet if
it is operating in Admission Control state (or Flow Termination
Westberg, et al. Expires August 28, 2008 [Page 9]
Internet-Draft LC-PCN February 2008
state). Otherwise the probe packet does not change its encoding
state.
3.1.2.2. PCN-egress-node features
The PCN-egress-node measures the rate of the received PHB aggregated
PCN unmarked and PCN_marking encoded packets. When the probe packet
arrives at the PCN-egress-node that is belonging to a certain edge-
to-edge pair PCN aggregate, and it is PCN_marking encoded then the
request is rejected. Otherwise it is accepted.
Note that if an edge-to-edge pair aggregated state is not available
at the PCN-egress-node, then the PCN-egress-node cannot determine
whether a PCN-egress-node associated with the edge-to-edge pair PCN
aggregate operates in Normal state, Admission Control state or Flow
Termination state. However, even in this case, when a probe packet
arrives at the PCN-egress-node, then this request should be rejected
if the probe packet is PCN_marking encoded. Otherwise, i.e., if the
probe packet is not PCN_marking encoded, it should be accepted.
3.1.2.3. PCN-ingress-node features
The PCN-ingress-node MUST set the Router_Alert option to all packets
that have to be used as probe packets. Furthermore, if the external
to PCN signaling protocol is also used by the PCN-ingress-node, then
the PCN-ingress-node SHOULD be informed that an admission control
request has been admitted or rejected by the PCN-egress-node. If the
PCN-ingress-node is notified that the admission request is rejected,
then the PCN-ingress-node rejects the admission control request.
Otherwise it is accepted.
3.1.3. ECMP solution
By using probing, the ECMP (Equal Cost Multi Path) problem that is
associated with the admission control feature can be, to a certain
degree, solved by being able to identify which flows are passing
through the congested node. This is because a probe packet can be
PCN_marking encoded only by congested PCN-interior-nodes. Note that
the ECMP problem is related to the fact that flows that are not
passing through a congested PCN-interior-node can belong to an
aggregate that detects a congestion.
Any measures that are taken on such flows will not solve the
congestion problem, since such flows are not contributing and causing
the congestion in the PCN-interior-node.
Westberg, et al. Expires August 28, 2008 [Page 10]
Internet-Draft LC-PCN February 2008
3.2. Flow Termination
The Flow Termination function is able to terminate flows in case of
exceptional events, such as severe congestion after re-routing. The
exceptional event, or severe congestion can be detected using a
remarking approach where the PCN_marking is proportional to the
excess rate. The Flow Termination features, similar to the Admission
Control features, can be applied to flows that are aggregated between
PCN-ingress-nodes and PCN-egress-nodes. In this way edge-to-edge
aggregates can be maintained by PCN-ingress-nodes and PCN-egress-
nodes.
Furthermore, the "N" and "U" PCN-domain-wide constraints, specified
in 3.1 are also used during Flow Termination.
Moreover, it is important to note that in this draft we denote the
not congested PCN packets (or bytes) as PCN unmarked packets (or
bytes).
3.2.1. PCN-interior-node
The PCN-interior-nodes can support two types of Flow Termination
modes, a base mode and an optimization mode. The Flow Termination
base mode that is supported by the PCN-interior-nodes can be
accomplished using the admission control features described in
Section 3.1.1.
However, inaccuracies in excess rate measurements might occur due to
the delay between the metering and marking events that occur at the
PCN-interior-nodes, the decisions that are made at PCN-egress-nodes,
and the termination of flows that are performed by PCN-ingress-nodes,
see Section 6 of [CsTa05].
In order to reduce these excess rate inaccuracies an optimization
feature can be added to the Flow Termination base mode, which is
denoted as optimization mode. The main addition that this
optimization mode requires is that an additional operational state
has to be maintained by the PCN-interior-node, i.e., a Flow
Termination state, see Section 3.3.1. In particular, when the
measured PHB aggregated PCN traffic is higher than the threshold
equal to (U * configured-admissible-rate), then the PCN-interior-node
changes from the Admission Control state to the Flow Termination
state.
3.2.2. PCN-egress-node
The PCN-egress-node measures the rate of the received PHB aggregated
PCN unmarked and PCN_marking encoded packets (or bytes).
Westberg, et al. Expires August 28, 2008 [Page 11]
Internet-Draft LC-PCN February 2008
The PCN-egress-node uses a similar functionality as discussed in
[Char07] to activate/trigger the Flow Termination with a trigger
computed from the ratio of the PCN_marking encoded and PCN unmarked
packets (or bytes). It is important that the format of the equation
of the above described ratio between the PCN_marking encoded and PCN
unmarked traffic, depends on the values of the U and the
configurable-admissible-rate at the PCN-interior-nodes.
In particular if the product U * configurable-admissible-rate is
smaller than the packet dropping threshold used by queuing mechanisms
at the PCN_interior nodes, then the above ratio is calculated by
dividing the product N * PCN_marking encoded rate of packets (or
bytes) and the rate of PCN unmarked packets (or bytes). If the
product (U * configurable-admissible-rate) is equal or higher than
the packet dropping threshold, then the above ratio (i.e., ratio of
the PCN_marking encoded and PCN unmarked packets (or bytes)) is
calculated by dividing the product N * PCN_marking encoded rate of
packets (or bytes) and the sum of the rates of N * PCN_marking
encoded and PCN unmarked packets (or bytes). Note that the rate of
PCN_marking packets (or bytes) has to be multiplied with N to match
the division by N performed by the PCN-interior-node.
The trigger is detected when the above given ratio between the
PCN_marking encoded and PCN unmarked encoded packets (or bytes) is
higher than the value of (U-1), see [Char07]. When this trigger is
detected then the PCN-egress-node has to store the value of the
absolute rate of the PCN_marking encoded traffic, say: configured-
termination-rate-egress, see also Section 3.3.2.
The N * (measured excess rate) that is above this threshold, is used
to calculate the number of flows to be terminated, such that the
excess rate is severely reduced until it drops below the Flow
Termination trigger.
Note that the excess rate and the flows to be terminated are
associated with the same edge-to-edge (i.e., ingress-egress) pair PCN
aggregate. For the flows that should be terminated the PCN-egress-
node informs the associated PCN-ingress-node to terminate them.
3.2.3. PCN-ingress-node
The flows that are Flow Termination notified by the PCN_egress-node
have to be terminated by the PCN_ingress-node. Furthermore, the PCN-
ingress-node rejects all new flow admission requests that are
associated with the same edge-to-edge pair PCN aggregate. In
addition, the PCN_ingress-node informs the associated flow sender
about the occurred exceptional/severe congestion.
Westberg, et al. Expires August 28, 2008 [Page 12]
Internet-Draft LC-PCN February 2008
3.2.4. ECMP solution
In order to solve the ECMP issue that may occur during Flow
Termination operational state, the LC-PCN solution could use an
additional PCN marking encoding approach, denoted as:
PCN_Affected_Marking.
This means that the descriptions of Section 3.2.1 and 3.2.2 have to
be slightly modified.
Regarding the description provided in section 3.2.1, the
PCN_Affected_Marking is used in the PCN-interior-node in the
following way. When the measured PHB aggregated PCN traffic is
higher than the threshold equal to (U*configured-admissible-rate),
then the PCN-interior-node changes from the Admission Control state
to Flow Termination state, see Section 3.3.2. In Flow Termination
state, the PCN-interior-node encodes all PCN unmarked (i.e., not
congested PCN encoded) packets that are passing through the PCN-
interior-node by using the PCN_Affected_Marking.
Regarding the PCN-egress-node description provided in section 3.2.2,
the rate of the PCN unmarked (or bytes), has to be replaced, by the
rate of PCN_Affected_Marking encoded packets (or bytes). This rate
is used to calculate the Flow Termination trigger at the PCN-egress-
node. Furthermore, the PCN-egress-node uses the PCN_Affected_Marking
to identify which flows were affected by the exceptional/severe
congestion. In this way the PCN-egress-node, when operating in Flow
Termination state, is able to terminate only the flows that received
one or more PCN_Affected_Marking packets.
3.3. Operational states in LC-PCN
This section describes the LC-PCN operational states that are used to
identify when and how a PCN node is triggered to either remain or
change into an operational state, i.e., Normal, Admission Control and
Flow Termination.
3.3.1. Operational states in PCN-interior-nodes
Per each PHB supported with the PCN domain, the PCN-interior-node
supports the operational states diagram depicted in Figure 2.
Westberg, et al. Expires August 28, 2008 [Page 13]
Internet-Draft LC-PCN February 2008
---------------------------------------------
| event B |
| V
---------- ------------- ----------
| Normal | event A | Admission | event B | Flow |
| state |---------->| Control |-------->|Termination|
| | | state | | state |
---------- ------------- ----------
^ ^ | |
| | event C | |
| ----------------------- |
| event D |
------------------------------------------------
Figure 2: States of Operation
The terms used in Figure 2 and applied for PCN-interior-nodes are:
* Normal state: represents the normal operation conditions of the
node, i.e. no congestion
* Flow Termination state: this state is only applied either when the
optimization mode solution is applied to solve the excess rate
measurement inaccuracies, see Section 3.2.1, or when the ECMP
solution described in Section 3.2.4 is used. This state represents
the state related to a certain PHB when the PCN-interior-node is
severely congested.
* Admission Control state: state where the load is relatively high,
close to the level when pre-congestion can occur
* event A: this event occurs when the measured PHB aggregated PCN
traffic is higher than the configured-admissible-rate. The measured
PHB aggregated PCN traffic rate that is above the configured-
admissible-rate is considered to be excess rate, which is encoded
using PCN_marking.
* event B: this event is only applied either when the optimization
solution is applied to solve the excess rate measurement inaccuracies
or when the ECMP solution is used. This event occurs when the
measured PHB aggregated PCN traffic is higher than the threshold
equal to (U * configured-admissible-rate).
* event C: this event occurs when the measured PHB aggregated PCN
traffic is equal or lower than the configured-admissible-rate.
* event D: this event is only applied either when the optimization
solution is applied to solve the excess rate measurement inaccuracies
Westberg, et al. Expires August 28, 2008 [Page 14]
Internet-Draft LC-PCN February 2008
or when the ECMP solution is used. This event occurs when the
measured PHB aggregated PCN traffic is equal or lower than the
threshold equal to (U * configured-admissible-rate).
3.3.2. Operational states in PCN-egress-nodes
Per edge-to-edge (ingress-egress) PCN aggregate, the PCN-egress-node
supports the same operational states diagram as depicted in Figure 2.
The terms used in Figure 2 and applied for PCN-egress-nodes are:
* Normal state: represents the normal operation conditions of the
node, i.e. no congestion.
* Flow Termination state: it represents the state related to a
certain edge-to-edge (ingress-egress) pair PCN aggregate to identify
the situation that a severe/exceptional event occurred and ongoing
flows need to be terminated in order to solve this severe congestion.
* Admission Control state: state where the load is relatively high,
close to the level when pre-congestion can occur.
* event A: this event is activated when the ratio between the
incoming_PCN_marked_rate and the total received PHB aggregated PCN
traffic is higher than a predefined value, e.g., 1%, see [Char07].
Where:
o incoming_PCN_marking_rate: the rate of the PCN_marking encoded
packets (or bytes) and it is equal to the product of N and the
measured incoming PCN_marking rate The measured rate of
PCN_marking encoded packets (or bytes) has to be multiplied with N
to match the division by N performed by the PCN-interior-node.
o Total received PCN traffic is the sum of the received
incoming_PCN_marking_rate and the PCN unmarked_rate. The PCN
unmarked_rate is the rate of the not congested PCN packets (or
bytes). When PCN_Affected_Marking is used, see Section 3.2.4,
then this rate represents the measured PCN_Affected_Marking
encoded packets (or bytes).
* event B: is activated comparing the ratio of the PCN_marked and PCN
unmarked packets (or bytes) to a certain preconfigured threshold. An
example of such a threshold is given in [Char07], with a value of
(U-1). The format of the equation of the above described ratio
between the PCN marked and PCN unmarked traffic may depend on the
values of the U and the configurable-admissible-rate at the PCN-
interior-nodes. In order to describe the conditions that influence
the event B trigger we use a pseudo code description:
Westberg, et al. Expires August 28, 2008 [Page 15]
Internet-Draft LC-PCN February 2008
If ((U * configured-admissible-rate) < Dropping_threshold)) THEN
{
IF incoming_PCN_marking_rate/unmarked_rate > (U - 1) THEN
{
activate event B
}
}
ELSE
{
IF incoming_PCN_marking_rate/(unmarked_rate +
incoming_PCN_marking_rate) >(U - 1) THEN
{
activate event B
}
}
Dropping_threshold = packet dropping threshold used by
queuing mechanisms at the PCN_interior nodes
unmarked_rate is the rate of the not congested PCN packets
(or bytes). When PCN_Affected_Marking is used, see
Section 3.2.4, then this rate represents the measured
PCN_Affected_Marking encoded packets (or bytes).
When this trigger is detected then the egress has to store the value
of the absolute rate of the PCN_marking encoded traffic, say:
configured-termination-rate-egress
* event C: this event occurs when the ratio between the
incoming_PCN_marked_rate and the total received PCN traffic is lower
or equal than the predefined value used to trigger event A.
* event D: this event is activated by comparing the ratio of the
PCN_marked and PCN unmarked packets (or bytes) to a certain
preconfigured threshold, see event B. In order to describe the
conditions that influence the event D trigger, we use a pseudo code
description, see also event B.
Westberg, et al. Expires August 28, 2008 [Page 16]
Internet-Draft LC-PCN February 2008
If ((U * configured-admissible-rate) < Dropping_threshold)) THEN
{
IF incoming_PCN_marking_rate/unmarked_rate =< (U - 1) THEN
{
activate event D
}
}
ELSE
{
IF incoming_PCN_marking_rate/(unmarked_rate +
incoming_PCN_marking_rate) =< (U - 1) THEN
{
activate event D
}
}
3.3.3. Operational states in PCN-ingress-nodes
Per each edge-to-edge pair of PCN aggregates the PCN-ingress-nodes
support the same operational states diagram as depicted in Figure 2.
The terms used in Figure 2 and applied for PCN-ingress-nodes are:
* Normal state: represents the normal operation conditions of the
node, i.e. no congestion.
* Flow Termination state: it represents the state related to a
certain edge-to-edge (ingress-egress) pair PCN aggregate to identify
the situation that a severe/exceptional event occurred and ongoing
flows need to be terminated in order to solve this severe congestion.
In Flow Termination, the PCN-ingress-node SHOULD block all new
admission flow requests that are associated with the same edge-to-
edge pair of PCN aggregates.
* Admission Control state: state where the load is relatively high,
close to the level when pre-congestion can occur. The PCN-ingress-
node rejects a flow that is requesting admission to the PCN domain.
* event A: this event occurs when the PCN-ingress-node receives a
response from the PCN-ingress-node that a flow that is requesting
admission to the PCN domain is rejected.
* event B: this event occurs when the PCN-ingress-node receives one
response from the PCN-ingress-node that a flow has to be terminated
due to the fact that the PCN-ingress-node operates in the Flow
Termination operational state.
* event C: this event occurs after the PCN-ingress-node rejected the
Westberg, et al. Expires August 28, 2008 [Page 17]
Internet-Draft LC-PCN February 2008
flow that was requesting admission and informed the flow sender about
it.
* event D: this event occurs when the PCN-ingress-node does not
receive anymore responses from the PCN-ingress-node that flows have
to be terminated.
3.4. Encoding of PCN traffic
The encoding that can be used for LC-PCN can be based either on DSCP
or on a combination between ECN and DSCP IP fields. In the current
version of the draft it is assumed that the encoding is based on only
the DSCP IP field. In particular, the encoding can be accomplished
in the following way. The PCN traffic can be distinguished from the
non PCN traffic by using a first additional DSCP, say
not_congested_PCN_DSCP, to identify the not congested PCN traffic.
The single marking state used during PCN_marking encoding can use a
second additional DSCP, say PCN_marking_DSCP. When the
PCN_Affected_Marking is used then an additional third DSCP is needed,
say PCN_Affected_Marking_DSCP.
The first, second and third additional DSCP values are representing
DSCP values that are assigned by IANA as DSCP experimental values.
It is important to note that when the LC-PCN is applied in multiple
neighboring PCN domains where a trust relationship exists between
these multiple PCN domains and a packet is received by the edge
router of another trusted domain (new PCN domain, that might be
managed by another operator), remarking of the
not_congested_PCN_DSCP, PCN_marking DSCP and PCN_Affected_Marking
DSCP to other DSCPs, say not_congested_PCN_new_DSCP,
PCN_marking_new_DSCP and PCN_Affected_Marking_new_DSCP, respectively,
might be necessary. This is because the neighbor PCN operator may
use different Diffserv mapping schemes.
When DSCP is used for PCN encoding and no trust relationships exist
between the PCN-domains, then for packets that are forwarded outside
the PCN-domain, the PCN-egress-nodes and PCN-ingress-nodes SHOULD
restore the original DSCP values of the PCN remarked packets,
otherwise multiple actions for the same event might occur. This
value MAY be left in its remarking form if there is an SLA agreement
between domains that a downstream domain handles the remarking
problem. When no trust relationship exists between multiple
neighboring PCN domains then the PCN-ingress-nodes SHOULD PCN encode
the incoming traffic that is used as incoming PCN traffic using the
not congested PCN DSCP.
Westberg, et al. Expires August 28, 2008 [Page 18]
Internet-Draft LC-PCN February 2008
4. LC-PCN detailed description
This section describes the details of the used LC-PCN algorithms.
Section 4.1, 4.2 and 4.3 describe the "Admission control based on
data marking", "Admission control based on probing" and "Flow
Termination" scenario, respectively, for the situation that the end-
to-end sessions are using unidirectional reservations. Section 4.4
describes the two admission control procedures and Section 4.5
describes the flow termination scenario for the situation that the
end-to-end sessions are using bi-directional reservations.
4.1. Admission control based on data marking for unidirectional flows
This type of admission control uses excess rate marking and metering
to provide admission control for unidirectional flows. In pre-
congestion situations the data packets are marked to notify the PCN-
egress-node that a congestion occurred on a particular PCN-ingress-
node to PCN-egress-node path.
4.1.1. Operation in PCN-interior-nodes
The PCN-interior-node performs measurements on the PHB aggregated PCN
traffic, see Figure 3, and changes operational state from Normal to
Admission Control state when the event A trigger occurs, see Section
3.3.1.
As mentioned in Section 3.1.1.1, the measured aggregated PCN traffic
rate that is above the configured-admissible-rate is considered to be
excess rate, which is marked using PCN_marking. When the PCN-
interior-node operates in Admission Control state and the configured-
admissible-rate is exceeded then PCN unmarked packets are
proportionally to the excess rate re-marked, using the PCN_marking
encoding, see event A, in Section 3.3.1.
The above described functionalities can be accomplished using
different metering and marking features. Several implementation
alternatives of this algorithms are possible. One implementation
alternative can be based on the algorithms discussed in [Char07]. In
particular, a token bucket can be used with the rate configured with
the rate equal to configured-admissible-rate.
Another implementation alternative can for example be based on rate
measurements and marking. In particular, the PCN-interior-nodes
packets using the PCN_marking, whenever the measured PHB aggregated
PCN traffic rate exceeds a pre-configured rate threshold denoted as
configurable-admissible-rate. An example of the detailed operation
of this later procedure is described below. The predefined
configured-admissible-rate, see Section 3.1.1.1 is set according to,
Westberg, et al. Expires August 28, 2008 [Page 19]
Internet-Draft LC-PCN February 2008
and usually less than, an engineered bandwidth limitation, i.e., real
admission threshold, based on e.g. agreed Service Level Agreement or
a capacity limitation of specific links. The difference between the
configured-admissible-rate and the engineered bandwidth limitation,
i.e., real admission threshold, provides an interval where the
signaling information on resource limitation is already sent by a
node but the actual resource limitation is not reached.
During admission control the PCN-interior-node calculates, per
traffic class (PHB), the incoming rate that is above configured-
admissible-rate, denoted as signaled_overload_rate, in the following
way:
* before queuing and eventually dropping the packets, at the end of
each measurement interval of T seconds, the PCN-interior-node should
count the total number of PCN unmarked, PCN_marking (and
PCN_Affected_Marking bytes, when the ECMP solution is used, see
Section 3.2.4) received. Denote this number as total_received_bytes.
Note that there are situations when more than one PCN-interior-nodes
in the same communication path become admission control congested and
operate in Admission Control state. Therefore, any PCN-interior-node
located behind a PCN- interior-node that operates in Admission
Control state may receive PCN_marking (and PCN_Affected_Marking, when
the ECMP solution is used, see Section 3.2.4) bytes.
Then the PCN-interior-node calculates the current estimated excess
rate (i.e., overloaded rate), say signaled_overload_rate, by using
the following equation:
signaled_overload_rate =
((total_received_bytes) / T) - configured-admissible-rate)
To provide reliable estimation of the encoded information several
techniques can be used, see [AtLi01], [AdCa03], [ThCo04], [AnHa06].
The bytes that have to be remarked to satisfy the signaled overload
rate, e.g., signaled_remarked_bytes, are calculated as follows:
Westberg, et al. Expires August 28, 2008 [Page 20]
Internet-Draft LC-PCN February 2008
IF (measured PHB rate > configured-admissible-rate
THEN
{
IF (incoming_PCN_marking_rate <> 0)
THEN
{ signaled_remarked_bytes =
((signaled_overload_rate -
incoming_PCN_marking_rate) * T) / N
}
ELSE signaled_remarked_bytes =
signaled_overload_rate * T / N
}
Where the "incoming_PCN_marking_rate" is calculated as follows:
incoming_PCN_marking_rate =
N * (input_PCN_marking_bytes) / T
where input_PCN_marking_bytes represents the measured
number of bytes carried by PCN_marking encoded packets.
When incoming PCN_marking encoded packets (or bytes) are dropped, the
operation of the admission control algorithm may be affected, e.g.,
the algorithm may become in certain situations slower. An
implementation of the algorithm may assure as much as possible that
the incoming PCN_marking encoded packets (or bytes) are not dropped.
This could for example be accomplished by using different dropping
rate thresholds for PCN_marking encoded and PCN unmarked (and
PCN_Affected_Marking encoded, when ECMP solution is used) bytes, see
Section 3.1.1.1.
4.1.2. Operation in PCN-egress-nodes
The PCN-egress-node measures the rate of the received PHB aggregated
PCN unmarked and PCN_marking marked packets. The measurements on the
PCN unmarked and unmarked traffic can be implemented using several
alternatives. One alternative is to use a similar functionality as
the one specified in [Char07] and [CL-ARCH] to calculate the
Congestion Level Estimate (CLE), which is the fraction of the marked
traffic received from one PCN-ingress-node. If the value of CLE is
higher than a certain value, e.g., 1%, then the PCN-egress-node is
changing its operational state from Normal state to Admission Control
state, see Section 3.3.2.
Another alternative is to use other type of rate measurements, than
the Exponential Weighted Moving Average (EWMA) measurements used to
calculate the CLE. The detailed description of this procedure is
given below. In this case, during a measurement interval T, the PCN-
Westberg, et al. Expires August 28, 2008 [Page 21]
Internet-Draft LC-PCN February 2008
egress-node measures the input_PCN_marking_bytes by counting, during
the interval T, the bytes that are carried in PCN_marking encoded
packets.
The incoming_PCN_marking_rate can be then calculated as follows:
incoming_PCN_marking_rate =
N * input_PCN_marking_bytes / T
where input_PCN_marking_bytes represents the measured number of
bytes carried by the PCN_marking encoded packets
To provide reliable estimation of the encoded information several
techniques can be used, see [AtLi01], [AdCa03], [ThCo04], [AnHa06].
If the ratio: incoming_PCN_marking_rate/(incoming_PCN_marking_rate +
PCN unmarked rate) is higher than a predefined value, say 1% then the
communication path between PCN-ingress-node and PCN-egress-node is
considered to be pre-congested.
PCN-ingress-node PCN-interior-node PCN-interior-node PCN-egress-node
user | | | |
data | user data | | |
------>|----------------->| user data | |
| |---------------->| user data |
| | |----------------->|
user | | | |
data | user data | | |
------>|----------------->| user data | user data |
| |---------------->S(# marked bytes) |
| | S----------------->|
| | S(# unmarked bytes)|
| | S----------------->|
| | S |
request for reservation | S |
------->| probe packet S |
|----------------------------------->S |
| | S probe packet |
| | S----------------->|
| |response |
|<------------------------------------------------------|
response | | |
<------| | | |
Figure: 3 Admission control based on data marking and probing
The admission control procedure is accomplished by using a
Westberg, et al. Expires August 28, 2008 [Page 22]
Internet-Draft LC-PCN February 2008
combination of the PCN operational state of the PCN-egress-node and
an admission control request provided by an external to PCN,
signaling protocol. When the admission control request arrives at a
PCN-egress-node that operates in admission control state then the
request SHOULD be rejected. If it operates in Normal state it SHOULD
be accepted. When DSCP is used for PCN encoding and no trust
relationships exist between the PCN-domains, then for packets that
are forwarded outside the PCN-domain, the PCN-egress-node SHOULD
restore the original DSCP values of the PCN remarked packets,
otherwise multiple actions for the same event might occur, see
Section 3.4.
4.1.3. Operation in PCN-ingress-nodes
The PCN-ingress-node can receive a reservation request message
belonging to an external to PCN, signaling protocol, e.g., RSVP.
This reservation request message can be used during the admission
control process. If the PCN-ingress-node receives a response, from
the PCN-egress-node, which notifies that the reservation request
message belonging to the external signaling protocol was successfully
processed, then the reservation request SHOULD be admitted.
Otherwise it SHOULD be rejected, see Section 3.3.3. Both situations
SHOULD be notified to the sender of the flow.
When DSCP is used for PCN encoding and no trust relationships exist
between the PCN-domains, then for packets that are forwarded outside
the PCN-domain, the PCN-ingress-node SHOULD restore the original DSCP
values of the PCN remarked packets, otherwise multiple actions for
the same event might occur, see Section 3.4. Furthermore, when the
DSCP encoding is used to encode the not congested PCN state, see
Section 3.4, then the PCN- ingress-node SHOULD remark to not
congested PCN encoding state, all incoming to PCN domain, packets
associated to flows that need to use the LC-PCN features.
4.2. Admission control based on probing for unidirectional flows
This type of admission control uses probing, whereby a probe packet
is sent along the forwarding path in a network to determine whether a
unidirectional flow can be admitted based upon the current congestion
state of the network. In pre-congestion situations the probe packets
are PCN_marking encoded to notify the PCN-egress-node that a
congestion occurred on a particular PCN-ingress-node to PCN-egress-
node path. The Admission control based on probing feature is used to
solve the ECMP issue that might occur during the process of admission
control, see Section 3.1.3.
Westberg, et al. Expires August 28, 2008 [Page 23]
Internet-Draft LC-PCN February 2008
4.2.1. Operation in PCN-interior-nodes
The PCN-interior-node features that are used to detect the PCN
operational states, are the same as the ones described in section
4.1.1. In this scenario an IP packet is used as a probe packet, see
Figure 3. A probe packet that passes through a PCN-interior-node
that operates in Admission Control state (or in Flow Termination
state, when either the Flow Termination optimization mode or the ECMP
solution described in Section 3.2.4 are used) MUST remark the PCN
unmarked encoded probe packet to PCN_marking encoded probe packet.
The PCN-interior-nodes SHOULD detect a probe packet by observing the
Router Alert option, which is carried by the probe packet. Note that
a PCN-ingress-node sets the Router Alert option of all packets that
are used as probe packets. This also holds for signaling protocol
messages that are used by LC-PCN as probe packets. Thus if a PCN-
interior-node receives a probe packet then, due to the Router Alert
option it has to handle it differently then the user packets. If the
PCN-interior-node operates in Admission Control state (or in Flow
Termination state, when either the Flow Termination optimization mode
or the ECMP solution described in Section 3.2.4 are supported) then
PCN-interior-node SHOULD PCN_marking encode the probe packet.
Otherwise, the encoding state of the probe packet SHOULD NOT change.
4.2.2. Operation in PCN-egress-nodes
The PCN-egress-node measures the rate of the received aggregated PCN
unmarked and PCN_marking encoded packets. When the probe packet
arrives at the PCN-egress-node that is belonging to a certain
ingress-egress PCN aggregate, and it is PCN_marking encoded then the
request SHOULD be rejected. In this way it is ensured that the probe
packet passed through the node that it is congested and therefore, it
can be used to solve the associated ECMP issue, see Section 3.4.
This feature is very useful when ECMP based routing is used to detect
only flows that are passing through the pre- congested router. Note
that even when no edge-to-edge pair PCN aggregate state is available
at the PCN-egress-node and when a probe packet arrives at the PCN-
egress-node, then this request SHOULD be rejected if the probe packet
is PCN_marking encoded. Otherwise, i.e., if the probe packet is not
PCN_marking encoded, it SHOULD be accepted. When DSCP is used for
PCN encoding and no trust relationships exist between the PCN-
domains, then for packets that are forwarded outside the PCN-domain,
the PCN-egress-node SHOULD restore the original DSCP values of the
PCN remarked packets, otherwise multiple actions for the same event
might occur, see Section 3.4.
Westberg, et al. Expires August 28, 2008 [Page 24]
Internet-Draft LC-PCN February 2008
4.2.3. Operation in PCN-ingress-nodes
Similar, to Section 4.1.3, the PCN-ingress-node can receive a
reservation request message belonging to an external to PCN,
signaling protocol, e.g., RSVP PATH message. Subsequently, the PCN-
ingress-node sends a probe packet, see Figure 3, towards the PCN-
egress-node. When RSVP is used, the RSVP PATH message is the probe
packet. Note that the probe packet should use the same flow ID
information and encoding state (ECN and/or DSCP) as the data packets
associated with the received reservation request message. The PCN-
ingress-node SHOULD set the Router Alert option carried by the probe
packet.
Note that probe packets can be either user data packets or messages
used by an external, to PCN, on path signaling protocol, e.g., RSVP
PATH. If the PCN-ingress-node receives a response that notifies that
the probe was successfully processed, then the reservation request is
admitted. In case of RSVP, the response is RSVP RESV message.
Otherwise it is rejected, see Section 3.3.3. Both situations have to
be notified to the sender of the flow.
When DSCP is used for PCN encoding and no trust relationships exist
between the PCN-domains, then for packets that are forwarded outside
the PCN-domain, the PCN-ingress-node SHOULD restore the original DSCP
values of the PCN remarked packets, otherwise multiple actions for
the same event might occur, see Section 3.4. Furthermore, when the
DSCP encoding is used to encode the not congested PCN state, see
Section 3.4, then the PCN- ingress-node SHOULD remark to not
congested PCN encoding state, all incoming to PCN domain, packets
associated to flows that need to use the LC-PCN features.
4.3. Flow Termination for unidirectional flows
The Flow Termination handling method requires the following
functionalities.
4.3.1. Operation in the PCN-interior-node
The PCN-interior-nodes can measure the PHB aggregated PCN traffic
that exceeds a configured-admissible-rate and mark this excess PCN
traffic, see Figure 4. This can be accomplished using different
metering and marking features, see Section 4.1.1.
The admission control features described in Section 4.1.1 can be
applied also for the situation that the PCN-interior-node operates in
the base mode of the Flow Termination state.
However, in the Flow termination base mode, inaccuracies in excess
Westberg, et al. Expires August 28, 2008 [Page 25]
Internet-Draft LC-PCN February 2008
rate measurements might occur due to the delay between the metering
and marking event that occurs at the PCN-interior-nodes, the
decisions that are made at PCN-egress-nodes, and the termination of
flows that are performed by PCN-ingress-nodes, see section 6 in
[CsTa05].
Furthermore, until the overload decreases at the PCN-interior-node
that operates in Flow Termination state, an additional trip time from
the PCN-ingress-node to this PCN-interior-node must expire. This is
because immediately before receiving the flow termination
notification, the PCN-ingress-node may have sent out packets in the
flows that were selected for termination. That is, a terminated flow
may contribute to congestion for a time longer that is taken from the
PCN-ingress-node to the PCN-interior-node. Without considering the
above, PCN-interior-nodes would continue marking the packets until
the measured utilization falls below the flow termination threshold.
In this way, at the end more flows will be terminated than necessary,
i.e., an over-reaction takes place. In order to reduce these
inaccuracies an optimization mode can be added to the base mode of
Flow Termination.
PCN-ingress-node PCN-interior-node PCN-interior-node PCN-egress-node
user | | | |
data | user data | | |
------>|----------------->| user data | user data |
| |---------------->S(# marked bytes) |
| | S----------------->|
| | S(# unmarked bytes)|
| | S----------------->|Term.
| notification for termination |flow?
|<-----------------|-----------------S------------------|YES
release | S |
| -----------------|----------------------------------->|
| | | |
Figure: 4 LC-PCN Flow Termination handling
4.3.1.1. Optimization mode features for Flow termination
In order to solve the excess rate inaccuracies, an additional
optimization mode feature can be used, see also [CsTa05]. The main
addition that this optimization mode solution requires is that an
additional operational state has to be maintained by the PCN-
interior-node, i.e., a Flow Termination state, see Section 3.4.1. In
particular, when the measured PHB aggregated PCN traffic is higher
than the threshold equal to (U * configured-admissible-rate), then
the PCN-interior-node changes from the Admission Control state to the
Westberg, et al. Expires August 28, 2008 [Page 26]
Internet-Draft LC-PCN February 2008
Flow Termination state.
Furthermore, when operating in Flow Termination state, the PCN-
interior-nodes use a sliding window memory to keep track of the
signaling overload in a couple of previous measurement intervals. At
the end of a measurement intervals, T, before encoding and signaling
the excess rate as PCN_marking encoded packets, the actual overload
is decreased with the sum of already signaled overload stored in the
sliding window memory, since that overload is already being handled
in the flow termination handling control loop. The sliding window
memory consists of an integer number of cells, i.e, n = maximum
number of cells. Guidelines for configuring the sliding window
parameters are given in [CsTa05].
At the end of each measurement interval, the newest calculated
overload is pushed into the memory, and the oldest cell is dropped.
If Mi is the overload_rate stored in ith memory cell (i = [1..n]),
then at the end of every measurement interval, the overload rate that
is signaled to the PCN-egress-node, i.e., signaled_overload_rate is
calculated as follows:
Sum_Mi =0
For i =1 to n
{
Sum_Mi = Sum_Mi + Mi
}
signaled_overload_rate = measured_overload_rate - Sum_Mi,
where Sum_Mi is calculated as above.
Next, the sliding memory is updated as follows:
for i = 1..(n-1): Mi < - Mi+1
Mn < - signaled_overload_rate
The bytes that have to be remarked to satisfy the signaled overload
rate: signaled_remarked_bytes, are calculated as follows:
Westberg, et al. Expires August 28, 2008 [Page 27]
Internet-Draft LC-PCN February 2008
IF (measured PHB rate > U * configured-admissible-rate)
THEN
{
IF (incoming_PCN_marking_rate <> 0)
THEN
{ signaled_remarked_bytes =
((signaled_overload_rate -
incoming_PCN_marking_rate) * T) / N
}
ELSE signaled_remarked_bytes = signaled_overload_rate * T / N
}
The incoming_PCN_marking_rate can be calculated as
follows:
incoming_PCN_marking_rate =
N * input_PCN_marking_bytes / T
where input_PCN_marking_bytes represents the measured
number of bytes carried by PCN_marking encoded packets.
The signal_remarked_bytes represents also the number of the outgoing
packets (after the dropping stage) that SHOULD be remarked, during
each measurement interval T, by a node when operates in Flow
Termination state.
4.3.1.2. ECMP solutions
As discussed in Section 3.2.4, the ECMP issue that may occur during
Flow Termination operational state, could be solved by using an
additional PCN marking encoding approach, denoted as:
PCN_Affected_Marking.
In this case both the Flow Termination base and optimization modes
have to be slightly modified. When the measured PHB aggregated PCN
traffic is higher than the threshold equal to (U * configured-
admissible-rate), then the PCN-interior-node changes from the
Admission Control state to Flow Termination state, see Section 3.3.1.
Furthermore, in Flow Termination state, the PCN-interior-node marks
all PCN unmarked (i.e., not congested PCN encoded) packets that are
passing through the PCN-interior-node.
4.3.2. Operation in PCN-egress-nodes
The PCN-egress-node measures the rate of the received PHB aggregated
PCN unmarked and PCN_marking encoded packets, see Figure 4. To
activate/trigger the Flow Termination a similar functionality can be
used as discussed in [Char07], see Section 3.2.2 and 3.3.2. The
Westberg, et al. Expires August 28, 2008 [Page 28]
Internet-Draft LC-PCN February 2008
implementation of the Flow Termination algorithm can be accomplished
in the following way.
The PCN-egress-node node applies a predefined policy to solve the
flow termination situation, by selecting a number of inter-domain
(end-to-end) flows that should be terminated, or forwarded in a lower
priority queue.
Some flows, belonging to the same PHB traffic class might get other
priority than other flows belonging to the same PHB traffic class.
It is considered that this difference in priority can be notified by
a signaling protocol and that the PCN-edge-nodes can store and
maintain the priority information releted to each of the end-to-end
flows. The terminated flows are selected from the flows belonging to
the same edge-to-edge pair PCN aggregate and having the same PHB
traffic class as the PHB of the PCN_marking encoded packets (and
PCN_Affected_Marking encoded packets, when the ECMP solution is
used).
For flows associated with the same PHB traffic class the priority of
the flow plays a significant role. An example of calculating the
number of flows associated with each priority class that have to be
terminated is described below.
An example of the algorithm for the calculation of the number of
flows, belonging to the same edge-to-edge pair PCN aggregate and
associated with each priority class that have to be terminated is
described using the pseudocode below. First, when the PCN-egress-
node operates in the Flow Termination state, see Section 3.4.2, then
the total amount of PCN_marking rate, per edge-to-edge pair PCN
aggregate, associated with the PHB traffic class, say
incoming_PCN_marking_rate, is calculated. This rate represents per
edge-to-edge pair PCN aggregate, the flow termination bandwidth, that
should be terminated. The incoming_PCN_marking_rate can be
calculated as follows:
incoming_PCN_marking_rate =
N * input_PCN_marking_bytes / T
where input_PCN_marking_bytes represents the measured
number of bytes carried by PCN_marking encoded packets.
To provide reliable estimation of the encoded information several
techniques can be used, see [AtLi01], [AdCa03], [ThCo04], [AnHa06].
The term denoted as terminated_bandwidth in the below pseudocode is a
temporal variable representing the total bandwidth that have to be
terminated, belonging to the same PHB traffic class. The
Westberg, et al. Expires August 28, 2008 [Page 29]
Internet-Draft LC-PCN February 2008
terminate_flow_bandwidth(priority_class) is the total of bandwidth
associated with flows of priority class equal to priority_class. The
parameter priority_class is an integer fulfilling
0 < priority_class =< Maximum_priority.
Note that if the PCN domain does not support priority differentiation
then the variable Maximum_priority SHOULD be equal to 0.
The calculate_terminate_flows(priority_class) function determines the
flows for a given priority class and per PHB that has to be
terminated. This function also calculates the term
sum_bandwidth_terminate(priority_class), which is the sum of the
bandwith associated with the flows that will be terminated. The
constraint of finding the total number of flows that have to be
terminated is that sum_bandwidth_terminate(priority_class), should be
smaller or approximatelly equal to the variable
terminate_bandwidth(priority_class).
terminated_bandwidth = 0;
priority_class = 0;
while terminated_bandwidth < termination_PCN_marking_rate
{
terminate_bandwidth(priority_class) =
termination_PCN_marking_rate - terminated_bandwidth
calculate_terminate_flows(priority_class);
terminated_bandwidth =
sum_bandwidth_terminate(priority_class) + terminated_bandwidth;
priority_class = priority_class + 1;
}
where:
* termination_PCN_marking_rate = incoming_PCN_marking_rate -
configured-termination-rate-egress
* configured-termination-rate-egress is defined in Section 3.2.2
and in the description of event B in Section 3.3.2.
For the end-to-end flows (sessions) that have to be terminated, the
PCN-egress-node SHOULD generate and send notification message to the
PCN-ingress-node to indicate the flow termination in the
communication path. Furthermore, for the aggregated sessions that
are affected, the PCN-egress-node SHOULD send within a notify message
the to be released bandwidth, associated with the edge-to-edge pair
PCN aggregated state. When DSCP is used for PCN encoding and no
trust relationships exist between the PCN-domains, then for packets
Westberg, et al. Expires August 28, 2008 [Page 30]
Internet-Draft LC-PCN February 2008
that are forwarded outside the PCN-domain, the PCN-egress-node SHOULD
restore the original DSCP values of the PCN remarked packets,
otherwise multiple actions for the same event might occur, see
Section 3.4.
4.3.2.1. ECMP solutions
When the ECMP solution is used by the PCN-egress-node then the
following modifications are required. The rate of the PCN unmarked
(or bytes), used on the calculations of the event that triggers the
Flow Termination state, see Section 3.3.2 has to be replaced, by the
rate of PCN_Affected_Marking encoded packets (or bytes).
Furthermore, the PCN-egress-node uses the PCN_Affected_Marking to
identify which flows were affected by the exceptional/severe
congestion. In this way the PCN-egress-node, when operating in Flow
Termination state, see Section 4.3.2, is able to terminate only the
flows that received one or more PCN_Affected_Marking packets.
4.3.3. Operation in PCN-ingress-nodes
Upon receiving the notification message sent by the PCN-egress-node,
the PCN-ingress-node resolves the flow termination congestion by a
predefined policy, e.g., by refusing new incoming flows (sessions),
terminating the affected and notified flows (sessions), and blocking
their packets or shifting them to an alternative LC-PCN traffic class
(PHB). This operation is depicted in Figure 4, where the PCN-
ingress- node, for each flow (session) to be terminated, receives a
notification message.
When the PCN-ingress-node receives the notification message, it
starts the termination of the flows within the LC-PCN domain by e.g.,
sending external to PCN, release signaling messages. Furthermore,
the PCN-ingress-node SHOULD reject all new flow admission requests
that are associated with the same edge-to-edge pair PCN aggregate,
see Section 3.3.3.
When the PCN-ingress-node receives the notification message that
contains the to be released aggregation bandwidth, it can use it to
resize the size of the aggregation size accordingly. The
functionality required to resize the edge-to-edge pair PCN aggregated
state is out of the scope of PCN.
When DSCP is used for PCN encoding and no trust relationships exist
between the PCN-domains, then for packets that are forwarded outside
the PCN-domain, the PCN-ingress-node SHOULD restore the original DSCP
values of the PCN remarked packets, otherwise multiple actions for
the same event might occur, see Section 3.4. Furthermore, when the
DSCP encoding is used to encode the not congested PCN state, see
Westberg, et al. Expires August 28, 2008 [Page 31]
Internet-Draft LC-PCN February 2008
Section 3.4, then the PCN- ingress-node SHOULD remark to not
congested PCN encoding state, all incoming to PCN domain, packets
associated to flows that need to use the LC-PCN features.
4.4. Admission control based on data marking and probing for bi-
directional flows
This section describes the admission control scheme that uses the
admission control function based on datamarking and probing when bi-
directional reservations are supported.
PCN-ingress-node PCN-interior-node PCN-interior-node PCN-egress-node
user| | | | |
data| | | | |
--->| | user data | |user data |
|-------------------------------------------->S (#marked bytes)
| | | S-------------->|
| | | S(#unmarked bytes)
| | | S-------------->|
| | | S |
| | probe(re-marked DSCP) |
| | | S |
|-------------------------------------------->S |
| | | S-------------->|
| | | S |
| | response(unsuccessful) |
|<------------------------------------------------------------|
| | | S |
Figure 5: Admission control based on data marking and probing
for bi-directional admission control (pre-congestion on
path from PCN-ingress-node towards PCN-egress-node)
This procedure is similar to the admission control procedure
described in Section 4.1 and 4.2 for the situations that the
admission control with data marking and admission control with
probing are used, respectively. The main difference is related to
the location of the PCN-interior-node that operates in admission
control state, i.e., "forward" path (i.e., path between PCN-ingress-
node towards PCN- egress-node) or "reverse" path (i.e., path between
PCN- egress-node towards PCN-ingress-node). Figure 5 shows the
scenario where the pre-congested PCN-interior-node is located in the
"forward" path. The functionality of providing admission control is
the same as the one described in Section 4.1 and 4.2, Figure 3.
Figure 6 shows the scenario where the pre-congested PCN-interior-node
is located in the "reverse" path. The probe packet sent in the
Westberg, et al. Expires August 28, 2008 [Page 32]
Internet-Draft LC-PCN February 2008
"forward" direction will not be affected by the pre-congested PCN-
interior-node, while the probe packet and any packet of the "reverse"
direction flows will be PCN_marking encoded. The PCN-ingress-node is
in this way notified that a pre-congestion situation occurred in the
network and therefore it will able to reject the new initiation of
the reservation.
PCN-ingress-node PCN-interior-node PCN-interior-node PCN-egress-node
user| | | | |
data| | | | |
--->| | user data | | |
|-------------------------------------------->|user data |user
| | | |-------------->|data
| | | | |--->
| | | | |user
| | | | |data
| | | | |<---
| S | user data | |
| S user data |<--------------------------|
| user data S<---------------| | |
|<---------------S | | |
| user data S | | |
| (#marked bytes)S | | |
|<---------------S | | |
| S probe(unmarked DSCP) |
| S | | |
|----------------S------------------------------------------->|
| S probe(re-marked DSCP) |
| S<-------------------------------------------|
|<---------------S | | |
Figure 6: Admission control based on data marking and probing for
bi-directional admission control (pre-congestion on path
PCN-egress-node towards PCN-ingress-node)
4.5. Flow Termination handling for bi-directional flows
This section describes the flow termination handling operation for
bi-directional flows. This flow termination handling operation is
similar to the one described in Section 4.3.
Westberg, et al. Expires August 28, 2008 [Page 33]
Internet-Draft LC-PCN February 2008
PCN-ingress-node PCN-interior-node PCN-interior-node PCN-egress-node
user| | | | |
data| user | | | |
--->| data | user data | |user data |
|--------------->| | S |
| |--------------------------->S (#marked bytes)
| | | S-------------->|
| | | S(#unmarked bytes)
| | | S-------------->|Term
| | | S |flow?
| | notification (terminate) |YES
|<------------------------------------------------------------|
|release (forward) | S |
|------------------------------------------------------------>|
| release (reverese) | S |
|<------------------------------------------------------------|
| | | S |
Figure 7: Flow termination handling for bi-directional
reservation (congestion on path PCN-ingress-node
towards PCN-egress-node)
This procedure is similar to the flow termination handling procedure
described in Section 4.3. The main difference is related to the
location of the the PCN-interior-node that operates in Flow
Termination state, , i.e. "forward" or "reverse" path. Figure 7
shows the scenario where the severe congested node is located in the
"forward" path. This scenario is very similar to the flow
termination handling scenario described in Section 4.3. The
difference is related to the release procedure, which is accomplished
in both directions "forward" and "reverse". Figure 8 shows the
scenario where the severe congested node is located in the "reverse"
path. The main difference between this scenario and the scenario
shown in Figure 7 is that no notification messages have to be
generated by the PCN-egress-node. This is because the (#marked and
#unmarked) user data is arriving at the PCN-ingress-node. The PCN-
ingress-node will be able to calculate the number of flows that have
to be terminated or forwarded in a lower priority queue. When a flow
termination congestion occurs on e.g., in the forward path, and when
the algorithm terminates flows to solve the flow termination in the
forward path (see Figure 7), then the reserved bandwidth associated
with the terminated bidirectional flows is also released. Therefore,
a careful selection of the flows that have to be terminated should
take place. A possible method of selecting the flows belonging to
the same priority type passing through the flow termination
congestion point on a unidirectional path can be the following:
Westberg, et al. Expires August 28, 2008 [Page 34]
Internet-Draft LC-PCN February 2008
o the PCN-egress-node should select, if possible, first
unidirectional flows instead of bidirectional flows
o the PCN-egress-node should select, if possible, bidirectional
flows that reserved a relatively small amount of resources on the
path reversed to the path of congestion.
PCN-ingress-node PCN-interior-node PCN-interior-node PCN-egress-node
user| | | | |
data| user | | | |
--->| data | user data | |user data |
|--------------->| | | |
| |--------------------------->|user data |user
| | | |-------------->|data
| | | | |--->
| | | user | |<---
| user data | | data |<--------------|
| (#marked bytes)| S<----------| |
|<--------------------------------S | |
| (#unmarked bytes) S | |
Term|<--------------------------------S | |
Flow? | S | |
YES | | S | |
|release (forward) S | |
|------------------------------------------------------------>|
| release (reverse) S | |
|<------------------------------------------------------------|
| | S | |
Figure 8: Flow termination handling for
bi-directional reservation (flow termination congestion on
path PCN-egress-node towards PCN-ingress-node)
Furthermore, a special case of this operation is associated to the
Flow Termination situation occurring simultaneously on the forward
and reverse paths. An example of this operation is given below (see
Figure 9). Consider that the PCN-egress-node selects a number of bi-
directional flows to be terminated, see Figure 9. In this case the
PCN-egress- node will send for each bi-directional flows a
notification message to PCN-ingress-node. If the PCN-ingress-node
receives these notification messages and its operational state
(associated with reverse path) is in the Flow Termination state (see
Section 3.3.3), then the PCN-ingress-node operates in the following
way:
Westberg, et al. Expires August 28, 2008 [Page 35]
Internet-Draft LC-PCN February 2008
PCN-ingress-node PCN-interior-node PCN-interior-node PCN-egress-node
user| | | | |
data| user | | | |
--->| data | #unmarked bytes| | |
|--------------->S #marked bytes | | |
| S--------------------------->| |
| | | |-------------->|data
| | | | |--->
| | | | Term.?
| NOTIFY | | |Yes
|<------------------------------------------------------------|
| | | | |data
| | | user | |<---
| user data | | data |<--------------|
| (#marked bytes)| S<----------| |
|<--------------------------------S | |
| (#unmarked bytes) S | |
Term|<--------------------------------S | |
Flow? | S | |
YES | | S | |
|release (forward) S | |
|------------------------------------------------------------>|
| release (reverse) S | |
|<------------------------------------------------------------|
Figure 9: Flow termination handling for
bi-directional reservation (flow termination congestion on
both forward and reverse direction)
o For each notification message, the PCN-ingress-node should
identify the bidirectional flows that have to be terminated.
o The PCN-ingress-node then calculates the total bandwidth that
should be released in the reverse direction (thus not in forward
direction) if the bidirectional flows will be terminated
(preempted), say "notify_reverse_bandwidth". This bandwidth can
be calculated by the sum of the bandwidth values associated with
all the end-to-end flows that received a (flow termination)
notification message.
o Furthermore, using the received marked packets (from the reverse
path) the PCN-ingress-node will calculate, using the algorithm
used by an PCN-egress-node and described in Section 4.3.2, the
total bandwidth that has to be terminated in order to solve the
flow termination congestion in the reverse path direction, say
"marked_reverse_bandwidth".
Westberg, et al. Expires August 28, 2008 [Page 36]
Internet-Draft LC-PCN February 2008
o The PCN-ingress-node then calculates the bandwidth of the
additional flows that have to be terminated, say
"additional_reverse_bandwidth", in order to solve the flow
termination congestion in the reverse direction, by taking into
account:
* the bandwidth in the reverse direction of the bidirectional
flows that were appointed by the PCN-egress-node (the ones that
received a notification message) to be preempted, i.e.,
"notify_reverse_bandwidth"
* the total amount of bandwidth in the reverse direction that has
been calculated by using the received marked packets, i.e.,
"marked_reverse_bandwidth". This additional bandwidth can be
calculated using the following algorithm:
IF ("marked_reverse_bandwidth" > "notify_reverse_bandwidth") THEN
"additional_reverse_bandwidth" =
"marked_reverse_bandwidth"- "notify_reverse_bandwidth";
ELSE
"additional_reverse_bandwidth" = 0
o PCN-ingress-node terminates the flows that experienced a severe
congestion in the "forward" path and received a (flow termination)
notification message
o If possible the PCN-ingress-node should terminate unidirectional
flows that are using the same egress-ingress reverse direction
communication path to satisfy the release of a total bandiwtdh up
equal to the: "additional_reverse_bandwidth".
o If the number of required uni-directional flows (to satisfy the
above issue) is not available, then a number of bi-directional
flows that are using the same egress-ingress reverse direction
communication path may be selected for flow termination in order
to satisfy the release of a total bandiwtdh equal up to the:
"additional_reverse_bandwidth". Note that using the guidelines
given in above, first the bidirectional flows that reserved a
relatively small amount of resources on the path reversed to the
path of congestion should be selected for termination.
o Furthermore, the PCN-egress-node includes the to be released
aggregated bandwidth value in one of the notification messages.
o The PCN-ingress-node receives this notification message and reads
the value of the carried to be released aggregated bandwidth.
Westberg, et al. Expires August 28, 2008 [Page 37]
Internet-Draft LC-PCN February 2008
The size of the aggregated reservation state can be reduced in the
"forward" and "reverse" by using the received to be reduced values
the aggregated bandwidth in "forward" and "reverese" directions.
5. Security Considerations
The security considerations associated with this document are similar
to the one described in [Eard08].
6. IANA Considerations
To be Added
7. Acknowledgements
The authors express their acknowledgement to the following
colleagues: Andras Csaszar, Attila Takacs, David Partain, Zoltan
Turanyi, Geert Heijenk, Anna Charny, Philip Eardley, Kwok Chan, Bob
Briscoe, Joe Babiarz, Michael Menth.
8. Informative References
[AdCa03] Adler, M., Cai, J., Shapiro, J., and D. Towsley,
"Estimation of congestion price using probabilistic packet
marking", Proc. IEEE INFOCOM, pp. 2068-2078, 2003.
[AnHa06] Lachlan, A. and S. Hanly, "The Estimation Error of
Adaptive Deterministic Packet Marking", 44th Annual
Allerton Conference on Communication, Control and
Computing, , 2006.
[AtLi01] Athuraliya, S., Li, V., Low, S., and Q. Yin, "REM: active
queue management", IEEE Network, vol. 15, pp. 48-53, May/
June 2001.
[Babi07] Babiarz, J. and et. al., "Three State PCN Marking",
draft-babiarz-pcn-3sm-01 (work in progress), ,
November 2007.
[Bernet99]
Bernett, Y., Yavatkar, R., Ford, P., Baker, F., Zhang, L.,
Speer, M., and R. Braden, "Interoperation of RSVP/Intserv
and Diffserv Networks", Work in Progress , March 1999.
Westberg, et al. Expires August 28, 2008 [Page 38]
Internet-Draft LC-PCN February 2008
[Berson97]
Berson, S. and R. Vincent, "Aggregation of Internet
Integrated Services State", Work in Progress, ,
December 1997.
[CL-ARCH] Briscoe, B. and et. al., "An edge-to-edge Deployment model
for pre-congestion notification: Admission control over a
Diffserv region", , October 2006.
[CL-PHB] Briscoe, B. and et. al., "Pre-congestion notification
marking", , October 2006.
[Char07] Charny, A. and et. al., "Pre-Congestion Notification Using
Single Marking for Admission and Termination",
draft-charny-pcn-single-marking-03 (work in progress), ,
November 2007.
[CsTa05] Csaszar, A., Takacs, A., Szabo, R., and T. Henk,
"Resilient Reduced-State Resource Reservation", Journal of
Communication and Networks Vol. 7, Num. 4, December 2005.
[Eard08] Eardley, P., "Pre-Congestion Notification Architecture",
draft-ietf-pcn-architecture-03 (work in progress), ,
February 2008.
[RFC2475] Blake, S., Black, D., Carlson, M., Davies, E., Wang, Z.,
and W. Weiss, "An Architecture for Differentiated
Services", RFC 2475, December 1998.
[RFC3175] Baker, F., Iturralde, C., Le Faucheur, F., and B. Davie,
"Aggregation of RSVP for IPv4 and IPv6 Reservations",
RFC 3175, September 2001.
[RMD] Bader, A., "RMD-QOSM: The resource management in Diffserv
QoS Model", draft-ietf-nsis-rmd-12.txt (work in
progress), , November 2007.
[Stoica99]
Stoica, I. and et. al., "Per Hop Behaviors Based on
Dynamic Packet States", Work in Progress , February 1999.
[ThCo04] Thommes, R. and M. Coates, "Deterministic packet marking
for congestion packet estimation", Proc. IEEE Infocom ,
2004.
[Westberg00]
Westberg, L. and et. al., "Load Control of Real-Time
Traffic", IETF Work in Progress , April 2000.
Westberg, et al. Expires August 28, 2008 [Page 39]
Internet-Draft LC-PCN February 2008
Authors' Addresses
Lars Westberg
Ericsson
Torshamnsgatan 23
SE-164 80 Stockholm
Sweden
Email: Lars.westberg@ericsson.com
Anurag Bhargava
Ericsson
920 Main Campus Dr., Suite 500
Raleigh, NC 27606
USA
Phone: +1 919 472 9964
Email: anurag.bhargava@ericsson.com
Attila Bader
Ericsson
Laborc 1
Budapest
Hungary
Email: Attila.Bader@ericsson.com
Georgios Karagiannis
University of Twente
P.O. Box 217
7500 AE Enscede
Netherlands
Email: g.karagiannis@ewi.utwente.nl
Westberg, et al. Expires August 28, 2008 [Page 40]
Internet-Draft LC-PCN February 2008
Full Copyright Statement
Copyright (C) The IETF Trust (2008).
This document is subject to the rights, licenses and restrictions
contained in BCP 78, and except as set forth therein, the authors
retain all their rights.
This document and the information contained herein are provided on an
"AS IS" basis and THE CONTRIBUTOR, THE ORGANIZATION HE/SHE REPRESENTS
OR IS SPONSORED BY (IF ANY), THE INTERNET SOCIETY, THE IETF TRUST AND
THE INTERNET ENGINEERING TASK FORCE DISCLAIM 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.
Intellectual Property
The IETF takes no position regarding the validity or scope of any
Intellectual Property Rights 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; nor does it represent that it has
made any independent effort to identify any such rights. Information
on the procedures with respect to rights in RFC documents can be
found in BCP 78 and BCP 79.
Copies of IPR disclosures made to the IETF Secretariat 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 on-line IPR repository at
http://www.ietf.org/ipr.
The IETF invites any interested party to bring to its attention any
copyrights, patents or patent applications, or other proprietary
rights that may cover technology that may be required to implement
this standard. Please address the information to the IETF at
ietf-ipr@ietf.org.
Acknowledgment
Funding for the RFC Editor function is provided by the IETF
Administrative Support Activity (IASA).
Westberg, et al. Expires August 28, 2008 [Page 41]
| PAFTECH AB 2003-2026 | 2026-04-23 08:59:42 |