draft-ietf-detnet-bounded-latency-00.txt   draft-ietf-detnet-bounded-latency-01.txt 
DetNet N. Finn DetNet N. Finn
Internet-Draft Huawei Technologies Co. Ltd Internet-Draft Huawei Technologies Co. Ltd
Intended status: Informational J-Y. Le Boudec Intended status: Informational J-Y. Le Boudec
Expires: January 25, 2020 E. Mohammadpour Expires: May 7, 2020 E. Mohammadpour
EPFL EPFL
J. Zhang J. Zhang
Huawei Technologies Co. Ltd Huawei Technologies Co. Ltd
B. Varga B. Varga
J. Farkas J. Farkas
Ericsson Ericsson
July 24, 2019 November 4, 2019
DetNet Bounded Latency DetNet Bounded Latency
draft-ietf-detnet-bounded-latency-00 draft-ietf-detnet-bounded-latency-01
Abstract Abstract
This document presents a timing model for Deterministic Networking This document presents a timing model for Deterministic Networking
(DetNet), so that existing and future standards can achieve the (DetNet), so that existing and future standards can achieve the
DetNet quality of service features of bounded latency and zero DetNet quality of service features of bounded latency and zero
congestion loss. It defines requirements for resource reservation congestion loss. It defines requirements for resource reservation
protocols or servers. It calls out queuing mechanisms, defined in protocols or servers. It calls out queuing mechanisms, defined in
other documents, that can provide the DetNet quality of service. other documents, that can provide the DetNet quality of service.
skipping to change at page 1, line 42 skipping to change at page 1, line 42
Internet-Drafts are working documents of the Internet Engineering Internet-Drafts are working documents of the Internet Engineering
Task Force (IETF). Note that other groups may also distribute Task Force (IETF). Note that other groups may also distribute
working documents as Internet-Drafts. The list of current Internet- working documents as Internet-Drafts. The list of current Internet-
Drafts is at https://datatracker.ietf.org/drafts/current/. Drafts is at https://datatracker.ietf.org/drafts/current/.
Internet-Drafts are draft documents valid for a maximum of six months Internet-Drafts are draft documents valid for a maximum of six months
and may be updated, replaced, or obsoleted by other documents at any and may be updated, replaced, or obsoleted by other documents at any
time. It is inappropriate to use Internet-Drafts as reference time. It is inappropriate to use Internet-Drafts as reference
material or to cite them other than as "work in progress." material or to cite them other than as "work in progress."
This Internet-Draft will expire on January 25, 2020. This Internet-Draft will expire on May 7, 2020.
Copyright Notice Copyright Notice
Copyright (c) 2019 IETF Trust and the persons identified as the Copyright (c) 2019 IETF Trust and the persons identified as the
document authors. All rights reserved. document authors. All rights reserved.
This document is subject to BCP 78 and the IETF Trust's Legal This document is subject to BCP 78 and the IETF Trust's Legal
Provisions Relating to IETF Documents Provisions Relating to IETF Documents
(https://trustee.ietf.org/license-info) in effect on the date of (https://trustee.ietf.org/license-info) in effect on the date of
publication of this document. Please review these documents publication of this document. Please review these documents
skipping to change at page 2, line 24 skipping to change at page 2, line 24
Table of Contents Table of Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 2 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 2
2. Terminology and Definitions . . . . . . . . . . . . . . . . . 3 2. Terminology and Definitions . . . . . . . . . . . . . . . . . 3
3. DetNet bounded latency model . . . . . . . . . . . . . . . . 4 3. DetNet bounded latency model . . . . . . . . . . . . . . . . 4
3.1. Flow creation . . . . . . . . . . . . . . . . . . . . . . 4 3.1. Flow creation . . . . . . . . . . . . . . . . . . . . . . 4
3.1.1. Static flow latency calculation . . . . . . . . . . . 4 3.1.1. Static flow latency calculation . . . . . . . . . . . 4
3.1.2. Dynamic flow latency calculation . . . . . . . . . . 5 3.1.2. Dynamic flow latency calculation . . . . . . . . . . 5
3.2. Relay node model . . . . . . . . . . . . . . . . . . . . 6 3.2. Relay node model . . . . . . . . . . . . . . . . . . . . 6
4. Computing End-to-end Latency Bounds . . . . . . . . . . . . . 8 4. Computing End-to-end Delay Bounds . . . . . . . . . . . . . . 8
4.1. Non-queuing delay bound . . . . . . . . . . . . . . . . . 8 4.1. Non-queuing delay bound . . . . . . . . . . . . . . . . . 8
4.2. Queuing delay bound . . . . . . . . . . . . . . . . . . . 8 4.2. Queuing delay bound . . . . . . . . . . . . . . . . . . . 9
4.2.1. Per-flow queuing mechanisms . . . . . . . . . . . . . 9 4.2.1. Per-flow queuing mechanisms . . . . . . . . . . . . . 9
4.2.2. Per-class queuing mechanisms . . . . . . . . . . . . 9 4.2.2. Per-class queuing mechanisms . . . . . . . . . . . . 9
4.3. Ingress considerations . . . . . . . . . . . . . . . . . 10 4.3. Ingress considerations . . . . . . . . . . . . . . . . . 10
4.4. Interspersed non-DetNet transit nodes . . . . . . . . . . 11 4.4. Interspersed non-DetNet transit nodes . . . . . . . . . . 11
5. Achieving zero congestion loss . . . . . . . . . . . . . . . 11 5. Achieving zero congestion loss . . . . . . . . . . . . . . . 11
5.1. A General Formula . . . . . . . . . . . . . . . . . . . . 11 6. Queuing techniques . . . . . . . . . . . . . . . . . . . . . 13
6. Queuing techniques . . . . . . . . . . . . . . . . . . . . . 12 6.1. Queuing data model . . . . . . . . . . . . . . . . . . . 13
6.1. Queuing data model . . . . . . . . . . . . . . . . . . . 12 6.2. Preemption . . . . . . . . . . . . . . . . . . . . . . . 15
6.2. Preemption . . . . . . . . . . . . . . . . . . . . . . . 14
6.3. Time-scheduled queuing . . . . . . . . . . . . . . . . . 15 6.3. Time-scheduled queuing . . . . . . . . . . . . . . . . . 15
6.4. Credit-Based Shaper with Asynchronous Traffic Shaping . . 16 6.4. Credit-Based Shaper with Asynchronous Traffic Shaping . . 16
6.4.1. Flow Admission . . . . . . . . . . . . . . . . . . . 19 6.4.1. Delay Bound Calculation . . . . . . . . . . . . . . . 18
6.4.2. Flow Admission . . . . . . . . . . . . . . . . . . . 19
6.5. IntServ . . . . . . . . . . . . . . . . . . . . . . . . . 20 6.5. IntServ . . . . . . . . . . . . . . . . . . . . . . . . . 20
6.6. Cyclic Queuing and Forwarding . . . . . . . . . . . . . . 22 6.6. Cyclic Queuing and Forwarding . . . . . . . . . . . . . . 23
6.6.1. CQF timing sequence . . . . . . . . . . . . . . . . . 23 6.6.1. CQF timing sequence . . . . . . . . . . . . . . . . . 24
6.6.2. CQF latency calculation . . . . . . . . . . . . . . . 24 6.6.2. CQF latency calculation . . . . . . . . . . . . . . . 24
7. References . . . . . . . . . . . . . . . . . . . . . . . . . 24 7. References . . . . . . . . . . . . . . . . . . . . . . . . . 25
7.1. Normative References . . . . . . . . . . . . . . . . . . 24 7.1. Normative References . . . . . . . . . . . . . . . . . . 25
7.2. Informative References . . . . . . . . . . . . . . . . . 25 7.2. Informative References . . . . . . . . . . . . . . . . . 26
Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 26 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 27
1. Introduction 1. Introduction
The ability for IETF Deterministic Networking (DetNet) or IEEE 802.1 The ability for IETF Deterministic Networking (DetNet) or IEEE 802.1
Time-Sensitive Networking (TSN, [IEEE8021TSN]) to provide the DetNet Time-Sensitive Networking (TSN, [IEEE8021TSN]) to provide the DetNet
services of bounded latency and zero congestion loss depends upon A) services of bounded latency and zero congestion loss depends upon A)
configuring and allocating network resources for the exclusive use of configuring and allocating network resources for the exclusive use of
DetNet/TSN flows; B) identifying, in the data plane, the resources to DetNet/TSN flows; B) identifying, in the data plane, the resources to
be utilized by any given packet, and C) the detailed behavior of be utilized by any given packet, and C) the detailed behavior of
those resources, especially transmission queue selection, so that those resources, especially transmission queue selection, so that
skipping to change at page 4, line 41 skipping to change at page 4, line 41
6. Assuming that the resources are available, commit those resources 6. Assuming that the resources are available, commit those resources
to the flow. This may or may not require adjusting the to the flow. This may or may not require adjusting the
parameters that control the filtering and/or queuing mechanisms parameters that control the filtering and/or queuing mechanisms
at each hop along the flow's path. at each hop along the flow's path.
This paradigm can be implemented using peer-to-peer protocols or This paradigm can be implemented using peer-to-peer protocols or
using a central server. In some situations, a lack of resources can using a central server. In some situations, a lack of resources can
require backtracking and recursing through this list. require backtracking and recursing through this list.
Issues such as un-provisioning a DetNet flow in favor of another when Issues such as un-provisioning a DetNet flow in favor of another,
resources are scarce are not considered, here. Also not addressed is when resources are scarce, are not considered, here. Also not
the question of how to choose the path to be taken by a DetNet flow. addressed is the question of how to choose the path to be taken by a
DetNet flow.
3.1.1. Static flow latency calculation 3.1.1. Static flow latency calculation
The static problem: The static problem:
Given a network and a set of DetNet flows, compute an end-to- Given a network and a set of DetNet flows, compute an end-to-
end latency bound (if computable) for each flow, and compute end latency bound (if computable) for each flow, and compute
the resources, particularly buffer space, required in each the resources, particularly buffer space, required in each
DetNet transit node to achieve zero congestion loss. DetNet transit node to achieve zero congestion loss.
In this calculation, all of the DetNet flows are known before the In this calculation, all of the DetNet flows are known before the
skipping to change at page 5, line 47 skipping to change at page 5, line 47
particularly buffer space, required in each DetNet transit particularly buffer space, required in each DetNet transit
node to achieve zero congestion loss. node to achieve zero congestion loss.
This calculation is dynamic, in the sense that flows can be added or This calculation is dynamic, in the sense that flows can be added or
deleted at any time, with a minimum of computation effort, and deleted at any time, with a minimum of computation effort, and
without affecting the guarantees already given to other flows. without affecting the guarantees already given to other flows.
The choice of queuing methods is critical to the applicability of the The choice of queuing methods is critical to the applicability of the
dynamic calculation. Some queuing methods (e.g. CQF, Section 6.6) dynamic calculation. Some queuing methods (e.g. CQF, Section 6.6)
make it easy to configure bounds on the network's capacity, and to make it easy to configure bounds on the network's capacity, and to
make independent calculations for each flow. Other queuing methods make independent calculations for each flow. [[E:The rest of this
(e.g., transmission selection by strict priority), make this paragraph should be changed.]] Other queuing methods (e.g.,
calculation impossible, because the worst case for one flow cannot be transmission selection by strict priority), make this calculation
computed without complete knowledge of all other flows. Other impossible, because the worst case for one flow cannot be computed
queuing methods (e.g. the credit-based shaper defined in [IEEE8021Q] without complete knowledge of all other flows. Other queuing methods
section 8.6.8.2) can be used for dynamic flow creation, but yield (e.g. the credit-based shaper defined in [IEEE8021Q] section 8.6.8.2)
poorer latency and buffer space guarantees than when that same can be used for dynamic flow creation, but yield poorer latency and
queuing method is used for static flow creation (Section 3.1.1). buffer space guarantees than when that same queuing method is used
for static flow creation (Section 3.1.1).
[[E:proposed replacement: Some other queuing methods (e.g. strict
priority with the credit-based shaper defined in [IEEE8021Q] section
8.6.8.2) can be used for dynamic flow creation, but yield poorer
latency and buffer space guarantees than when that same queuing
method is used for static flow creation (Section 3.1.1).]]
3.2. Relay node model 3.2. Relay node model
A model for the operation of a DetNet transit node is required, in A model for the operation of a DetNet transit node is required, in
order to define the latency and buffer calculations. In Figure 1 we order to define the latency and buffer calculations. In Figure 1 we
see a breakdown of the per-hop latency experienced by a packet see a breakdown of the per-hop latency experienced by a packet
passing through a DetNet transit node, in terms that are suitable for passing through a DetNet transit node, in terms that are suitable for
computing both hop-by-hop latency and per-hop buffer requirements. computing both hop-by-hop latency and per-hop buffer requirements.
DetNet transit node A DetNet transit node B DetNet transit node A DetNet transit node B
skipping to change at page 6, line 34 skipping to change at page 6, line 41
|<->|<------>|<------->|<->|<---->|<->|<------>|<------>|<->|<-- |<->|<------>|<------->|<->|<---->|<->|<------>|<------>|<->|<--
2,3 4 5 6 1 2,3 4 5 6 1 2,3 2,3 4 5 6 1 2,3 4 5 6 1 2,3
1: Output delay 4: Processing delay 1: Output delay 4: Processing delay
2: Link delay 5: Regulation delay 2: Link delay 5: Regulation delay
3: Preemption delay 6: Queuing delay. 3: Preemption delay 6: Queuing delay.
Figure 1: Timing model for DetNet or TSN Figure 1: Timing model for DetNet or TSN
In Figure 1, we see two DetNet transit nodes (typically, bridges or In Figure 1, we see two DetNet transit nodes (typically, bridges or
routers), with a wired link between them. In this model, the only routers), with a wired link between them. In this model, the only
queues we deal with explicitly are attached to the output port; other queues, that we deal with explicitly, are attached to the output
queues are modeled as variations in the other delay times. (E.g., an port; other queues are modeled as variations in the other delay
input queue could be modeled as either a variation in the link delay times. (E.g., an input queue could be modeled as either a variation
[2] or the processing delay [4].) There are six delays that a packet in the link delay [2] or the processing delay [4].) There are six
can experience from hop to hop. delays that a packet can experience from hop to hop.
1. Output delay 1. Output delay
The time taken from the selection of a packet for output from a The time taken from the selection of a packet for output from a
queue to the transmission of the first bit of the packet on the queue to the transmission of the first bit of the packet on the
physical link. If the queue is directly attached to the physical physical link. If the queue is directly attached to the physical
port, output delay can be a constant. But, in many port, output delay can be a constant. But, in many
implementations, the queuing mechanism in a forwarding ASIC is implementations, the queuing mechanism in a forwarding ASIC is
separated from a multi-port MAC/PHY, in a second ASIC, by a separated from a multi-port MAC/PHY, in a second ASIC, by a
multiplexed connection. This causes variations in the output multiplexed connection. This causes variations in the output
delay that are hard for the forwarding node to predict or control. delay that are hard for the forwarding node to predict or control.
skipping to change at page 8, line 20 skipping to change at page 8, line 28
variation in the next hop, so the output delay variations of the variation in the next hop, so the output delay variations of the
previous hop (on each input port) must be known in order to previous hop (on each input port) must be known in order to
calculate the buffer space required on this hop. calculate the buffer space required on this hop.
o Variations in processing delay (4) require additional output o Variations in processing delay (4) require additional output
buffers in the queues of that same DetNet transit node. Depending buffers in the queues of that same DetNet transit node. Depending
on the details of the queueing subsystem delay (6) calculations, on the details of the queueing subsystem delay (6) calculations,
these variations need not be visible outside the DetNet transit these variations need not be visible outside the DetNet transit
node. node.
4. Computing End-to-end Latency Bounds 4. Computing End-to-end Delay Bounds
4.1. Non-queuing delay bound 4.1. Non-queuing delay bound
End-to-end latency bounds can be computed using the delay model in End-to-end delay bounds can be computed using the delay model in
Section 3.2. Here it is important to be aware that for several Section 3.2. Here, it is important to be aware that for several
queuing mechanisms, the worst-case end-to-end delay is less than the queuing mechanisms, the end-to-end delay bound is less than the sum
sum of the per-hop worst-case delays. An end-to-end latency bound of the per-hop delay bounds. An end-to-end delay bound for one
for one DetNet flow can be computed as DetNet flow can be computed as
end_to_end_latency_bound = non_queuing_latency + queuing_latency end_to_end_delay_bound = non_queuing_delay_bound +
queuing_delay_bound
The two terms in the above formula are computed as follows. First, The two terms in the above formula are computed as follows.
at the h-th hop along the path of this DetNet flow, obtain an upper
bound per-hop_non_queuing_latency[h] on the sum of delays 1,2,3,4 of
Figure 1. These upper-bounds are expected to depend on the specific
technology of the DetNet transit node at the h-th hop but not on the
T-SPEC of this DetNet flow. Then set non_queuing_latency = the sum
of per-hop_non_queuing_latency[h] over all hops h.
4.2. Queuing delay bound First, at the h-th hop along the path of this DetNet flow, obtain an
upperbound per-hop_non_queuing_delay_bound[h] on the sum of the
bounds over the delays 1,2,3,4 of Figure 1. These upper bounds are
expected to depend on the specific technology of the DetNet transit
node at the h-th hop but not on the T-SPEC of this DetNet flow. Then
set non_queuing_delay_bound = the sum of per-
hop_non_queuing_delay_bound[h] over all hops h.
Second, compute queuing_latency as an upper bound to the sum of the Second, compute queuing_delay_bound as an upper bound to the sum of
queuing delays along the path. The value of queuing_latency depends the queuing delays along the path. The value of queuing_delay_bound
on the T-SPEC of this flow and possibly of other flows in the depends on the T-SPEC of this flow and possibly of other flows in the
network, as well as the specifics of the queuing mechanisms deployed network, as well as the specifics of the queuing mechanisms deployed
along the path of this flow. along the path of this flow. The computation of queuing_delay_bound
is described in Section 4.2 as a separate section.
For several queuing mechanisms, queuing_latency is less than the sum 4.2. Queuing delay bound
of upper bounds on the queuing delays (5,6) at every hop. This
For several queuing mechanisms, queuing_delay_bound is less than the
sum of upper bounds on the queuing delays (5,6) at every hop. This
occurs with (1) per-flow queuing, and (2) per-class queuing with occurs with (1) per-flow queuing, and (2) per-class queuing with
regulators, as explained in Section 4.2.1, Section 4.2.2, and regulators, as explained in Section 4.2.1, Section 4.2.2, and
Section 6. Section 6.
For other queuing mechanisms the only available value of For other queuing mechanisms the only available value of
queuing_latency is the sum of the per-hop queuing delay bounds. In queuing_delay_bound is the sum of the per-hop queuing delay bounds.
such cases, the computation of per-hop queuing delay bounds must In such cases, the computation of per-hop queuing delay bounds must
account for the fact that the T-SPEC of a DetNet flow is no longer account for the fact that the T-SPEC of a DetNet flow is no longer
satisfied at the ingress of a hop, since burstiness increases as one satisfied at the ingress of a hop, since burstiness increases as one
flow traverses one DetNet transit node. flow traverses one DetNet transit node.
4.2.1. Per-flow queuing mechanisms 4.2.1. Per-flow queuing mechanisms
With such mechanisms, each flow uses a separate queue inside every With such mechanisms, each flow uses a separate queue inside every
node. The service for each queue is abstracted with a guaranteed node. The service for each queue is abstracted with a guaranteed
rate and a delay. For every flow the per-node delay bound as well as rate and a latency. For every flow, a per-node delay bound as well
end-to-end delay bound can be computed from the traffic specification as an end-to-end delay bound can be computed from the traffic
of this flow at its source and from the values of rates and latencies specification of this flow at its source and from the values of rates
at all nodes along its path. Details of calculation for IntServ are and latencies at all nodes along its path. The per-flow queuing is
described in Section 6.5. used in IntServ. Details of calculation for IntServ are described in
Section 6.5.
4.2.2. Per-class queuing mechanisms 4.2.2. Per-class queuing mechanisms
With such mechanisms, the flows that have the same class share the With such mechanisms, the flows that have the same class share the
same queue. A practical example is the credit-based shaper defined same queue. A practical example is the credit-based shaper defined
in section 8.6.8.2 of [IEEE8021Q]. One key issue in this context is in section 8.6.8.2 of [IEEE8021Q]. One key issue in this context is
how to deal with the burstiness cascade: individual flows that share how to deal with the burstiness cascade: individual flows that share
a resource dedicated to a class may see their burstiness increase, a resource dedicated to a class may see their burstiness increase,
which may in turn cause increased burstiness to other flows which may in turn cause increased burstiness to other flows
downstream of this resource. Computing latency upper bounds for such downstream of this resource. Computing delay upper bounds for such
cases is difficult, and in some conditions impossible cases is difficult, and in some conditions impossible
[charny2000delay][bennett2002delay]. Also, when bounds are obtained, [charny2000delay][bennett2002delay]. Also, when bounds are obtained,
they depend on the complete configuration, and must be recomputed they depend on the complete configuration, and must be recomputed
when one flow is added. (The dynamic calculation, Section 3.1.2.) when one flow is added. (The dynamic calculation, Section 3.1.2.)
A solution to deal with this issue is to reshape the flows at every A solution to deal with this issue is to reshape the flows at every
hop. This can be done with per-flow regulators (e.g. leaky bucket hop. This can be done with per-flow regulators (e.g. leaky bucket
shapers), but this requires per-flow queuing and defeats the purpose shapers), but this requires per-flow queuing and defeats the purpose
of per-class queuing. An alternative is the interleaved regulator, of per-class queuing. An alternative is the interleaved regulator,
which reshapes individual flows without per-flow queuing which reshapes individual flows without per-flow queuing
skipping to change at page 10, line 12 skipping to change at page 10, line 22
when an interleaved regulator is appended to a FIFO subsystem, it when an interleaved regulator is appended to a FIFO subsystem, it
does not increase the worst-case delay of the latter. does not increase the worst-case delay of the latter.
Figure 2 shows an example of a network with 5 nodes, per-class Figure 2 shows an example of a network with 5 nodes, per-class
queuing mechanism and interleaved regulators as in Figure 1. An end- queuing mechanism and interleaved regulators as in Figure 1. An end-
to-end delay bound for flow f, traversing nodes 1 to 5, is calculated to-end delay bound for flow f, traversing nodes 1 to 5, is calculated
as follows: as follows:
end_to_end_latency_bound_of_flow_f = C12 + C23 + C34 + S4 end_to_end_latency_bound_of_flow_f = C12 + C23 + C34 + S4
In the above formula, Cij is a bound on the aggregate response time In the above formula, Cij is a bound on the delay of the queuing
of queuing subsystem in node i and interleaved regulator of node j, subsystem in node i and interleaved regulator of node j, and S4 is a
and S4 is a bound on the response time of the queuing subsystem in bound on the delay of the queuing subsystem in node 4 for flow f. In
node 4 for flow f. In fact, using the delay definitions in fact, using the delay definitions in Section 3.2, Cij is a bound on
Section 3.2, Cij is a bound on sum of the delays 1,2,3,6 of node i sum of the delays 1,2,3,6 of node i and 4,5 of node j. Similarly, S4
and 4,5 of node j. Similarly, S4 is a bound on sum of the delays is a bound on sum of the delays 1,2,3,6 of node 4. A practical
1,2,3,6 of node 4. A practical example of queuing model and delay example of queuing model and delay calculation is presented
calculation is presented Section 6.4. Section 6.4.
f f
-----------------------------> ----------------------------->
+---+ +---+ +---+ +---+ +---+ +---+ +---+ +---+ +---+ +---+
| 1 |---| 2 |---| 3 |---| 4 |---| 5 | | 1 |---| 2 |---| 3 |---| 4 |---| 5 |
+---+ +---+ +---+ +---+ +---+ +---+ +---+ +---+ +---+ +---+
\__C12_/\__C23_/\__C34_/\_S4_/ \__C12_/\__C23_/\__C34_/\_S4_/
Figure 2: End-to-end latency computation example Figure 2: End-to-end delay computation example
REMARK: The end-to-end delay bound calculation provided here gives a REMARK: The end-to-end delay bound calculation provided here gives a
much better upper bound in comparison with end-to-end delay bound much better upper bound in comparison with end-to-end delay bound
computation by adding the delay bounds of each node in the path of a computation by adding the delay bounds of each node in the path of a
flow [TSNwithATS]. flow [TSNwithATS].
4.3. Ingress considerations 4.3. Ingress considerations
A sender can be a DetNet node which uses exactly the same queuing A sender can be a DetNet node which uses exactly the same queuing
methods as its adjacent DetNet transit node, so that the latency and methods as its adjacent DetNet transit node, so that the delay and
buffer calculations at the first hop are indistinguishable from those buffer bounds calculations at the first hop are indistinguishable
at a later hop within the DetNet domain. On the other hand, the from those at a later hop within the DetNet domain. On the other
sender may be DetNet unaware, in which case some conditioning of the hand, the sender may be DetNet unaware, in which case some
flow may be necessary at the ingress DetNet transit node. conditioning of the flow may be necessary at the ingress DetNet
transit node.
This ingress conditioning typically consists of a FIFO with an output This ingress conditioning typically consists of a FIFO with an output
regulator that is compatible with the queuing employed by the DetNet regulator that is compatible with the queuing employed by the DetNet
transit node on its output port(s). For some queuing methods, simply transit node on its output port(s). For some queuing methods, simply
requires added extra buffer space in the queuing subsystem. Ingress requires added extra buffer space in the queuing subsystem. Ingress
conditioning requirements for different queuing methods are mentioned conditioning requirements for different queuing methods are mentioned
in the sections, below, describing those queuing methods. in the sections, below, describing those queuing methods.
4.4. Interspersed non-DetNet transit nodes 4.4. Interspersed non-DetNet transit nodes
It is sometimes desirable to build a network that has both DetNet It is sometimes desirable to build a network that has both DetNet
aware transit nodes and DetNet non-aware transit nodes, and for a aware transit nodes and DetNet non-aware transit nodes, and for a
DetNet flow to traverse an island of non-DetNet transit nodes, while DetNet flow to traverse an island of non-DetNet transit nodes, while
still allowing the network to offer latency and congestion loss still allowing the network to offer delay and congestion loss
guarantees. This is possible under certain conditions. guarantees. This is possible under certain conditions.
In general, when passing through a non-DetNet island, the island In general, when passing through a non-DetNet island, the island
causes delay variation in excess of what would be caused by DetNet causes delay variation in excess of what would be caused by DetNet
nodes. That is, the DetNet flow is "lumpier" after traversing the nodes. That is, the DetNet flow is "lumpier" after traversing the
non-DetNet island. DetNet guarantees for latency and buffer non-DetNet island. DetNet guarantees for delay and buffer
requirements can still be calculated and met if and only if the requirements can still be calculated and met if and only if the
following are true: following are true:
1. The latency variation across the non-DetNet island must be 1. The latency variation across the non-DetNet island must be
bounded and calculable. bounded and calculable.
2. An ingress conditioning function (Section 4.3) may be required at 2. An ingress conditioning function (Section 4.3) may be required at
the re-entry to the DetNet-aware domain. This will, at least, the re-entry to the DetNet-aware domain. This will, at least,
require some extra buffering to accommodate the additional delay require some extra buffering to accommodate the additional delay
variation, and thus further increases the worst-case latency. variation, and thus further increases the delay bound.
The ingress conditioning is exactly the same problem as that of a The ingress conditioning is exactly the same problem as that of a
sender at the edge of the DetNet domain. The requirement for bounds sender at the edge of the DetNet domain. The requirement for bounds
on the latency variation across the non-DetNet island is typically on the latency variation across the non-DetNet island is typically
the most difficult to achieve. Without such a bound, it is obvious the most difficult to achieve. Without such a bound, it is obvious
that DetNet cannot deliver its guarantees, so a non-DetNet island that DetNet cannot deliver its guarantees, so a non-DetNet island
that cannot offer bounded latency variation cannot be used to carry a that cannot offer bounded latency variation cannot be used to carry a
DetNet flow. DetNet flow.
5. Achieving zero congestion loss 5. Achieving zero congestion loss
When the input rate to an output queue exceeds the output rate for a When the input rate to an output queue exceeds the output rate for a
sufficient length of time, the queue must overflow. This is sufficient length of time, the queue must overflow. This is
congestion loss, and this is what deterministic networking seeks to congestion loss, and this is what deterministic networking seeks to
avoid. avoid.
5.1. A General Formula
To avoid congestion losses, an upper bound on the backlog present in To avoid congestion losses, an upper bound on the backlog present in
the regulator and queuing subsystem of Figure 1 must be computed the regulator and queuing subsystem of Figure 1 must be computed
during resource reservation. This bound depends on the set of flows during resource reservation. This bound depends on the set of flows
that use these queues, the details of the specific queuing mechanism that use these queues, the details of the specific queuing mechanism
and an upper bound on the processing delay (4). The queue must and an upper bound on the processing delay (4). The queue must
contain the packet in transmission plus all other packets that are contain the packet in transmission plus all other packets that are
waiting to be selected for output. waiting to be selected for output.
A conservative backlog bound, that applies to all systems, can be A conservative backlog bound, that applies to all systems, can be
derived as follows. derived as follows.
skipping to change at page 12, line 18 skipping to change at page 12, line 26
The backlog bound is counted in data units (bytes, or words of The backlog bound is counted in data units (bytes, or words of
multiple bytes) that are relevant for buffer allocation. For every multiple bytes) that are relevant for buffer allocation. For every
class we need one buffer space for the packet in transmission, plus class we need one buffer space for the packet in transmission, plus
space for the packets that are waiting to be selected for output. space for the packets that are waiting to be selected for output.
Excluding transmission and preemption times, the packets are waiting Excluding transmission and preemption times, the packets are waiting
in the queue since reception of the last bit, for a duration equal to in the queue since reception of the last bit, for a duration equal to
the processing delay (4) plus the queuing delays (5,6). the processing delay (4) plus the queuing delays (5,6).
Let Let
o nb_classes be the number of classes of traffic that may use this
output port
o total_in_rate be the sum of the line rates of all input ports that o total_in_rate be the sum of the line rates of all input ports that
send traffic of any class to this output port. The value of send traffic of any class to this output port. The value of
total_in_rate is in data units (e.g. bytes) per second. total_in_rate is in data units (e.g. bytes) per second.
o nb_input_ports be the number input ports that send traffic of any o nb_input_ports be the number input ports that send traffic of any
class to this output port class to this output port
o max_packet_length be the maximum packet size for packets of any o max_packet_length be the maximum packet size for packets of any
class that may be sent to this output port. This is counted in class that may be sent to this output port. This is counted in
data units. data units.
o max_delay45 be an upper bound, in seconds, on the sum of the o max_delay456 be an upper bound, in seconds, on the sum of the
processing delay (4) and the queuing delays (5,6) for a packet of processing delay (4) and the queuing delays (5,6) for a packet of
any class at this output port. any class at this output port.
Then a bound on the backlog of traffic of all classes in the queue at Then a bound on the backlog of traffic of all classes in the queue at
this output port is this output port is
[[E: The formula is not right; why do we need nb_classes to compute
backlog bound?]]
backlog_bound = ( nb_classes + nb_input_ports ) * backlog_bound = ( nb_classes + nb_input_ports ) *
max_packet_length + total_in_rate* max_delay45 max_packet_length + total_in_rate* max_delay456
[[E: proposed general backlog bound:]]
backlog_bound = nb_input_ports * max_packet_length +
total_in_rate* max_delay456
6. Queuing techniques 6. Queuing techniques
6.1. Queuing data model 6.1. Queuing data model
Sophisticated queuing mechanisms are available in Layer 3 (L3, see, Sophisticated queuing mechanisms are available in Layer 3 (L3, see,
e.g., [RFC7806] for an overview). In general, we assume that "Layer e.g., [RFC7806] for an overview). In general, we assume that "Layer
3" queues, shapers, meters, etc., are precisely the "regulators" 3" queues, shapers, meters, etc., are precisely the "regulators"
shown in Figure 1. The "queuing subsystems" in this figure are not shown in Figure 1. The "queuing subsystems" in this figure are not
the province solely of bridges; they are an essential part of any the province solely of bridges; they are an essential part of any
skipping to change at page 15, line 14 skipping to change at page 15, line 47
6.3. Time-scheduled queuing 6.3. Time-scheduled queuing
In [IEEE8021Q], the notion of time-scheduling queue gates is In [IEEE8021Q], the notion of time-scheduling queue gates is
described in section 8.6.8.4. Below every output queue (the lower described in section 8.6.8.4. Below every output queue (the lower
row of queues in Figure 3) is a gate that permits or denies the queue row of queues in Figure 3) is a gate that permits or denies the queue
to present data for transmission selection. The gates are controlled to present data for transmission selection. The gates are controlled
by a rotating schedule that can be locked to a clock that is by a rotating schedule that can be locked to a clock that is
synchronized with other DetNet transit nodes. The DetNet class of synchronized with other DetNet transit nodes. The DetNet class of
service can be supplied by queuing mechanisms based on time, rather service can be supplied by queuing mechanisms based on time, rather
than the regulator model in Figure 3. Generally speacking, this than the regulator model in Figure 3. Generally speaking, this time-
time-aware scheduling can be used as a layer 2 time division aware scheduling can be used as a layer 2 time division multiplexing
multiplexing (TDM) technique. (TDM) technique.
Consider the static configuration of a deterministic network. To Consider the static configuration of a deterministic network. To
provide end-to-end latency guaranteed service, network nodes can provide end-to-end latency guaranteed service, network nodes can
support time-based behavior, which is determined by gate control list support time-based behavior, which is determined by gate control list
(GCL). GCL defines the gate operation, in open or closed state, with (GCL). GCL defines the gate operation, in open or closed state, with
associated timing for each traffic class queue. A time slice with associated timing for each traffic class queue. A time slice with
gate state "open" is called transmission window. The time-based gate state "open" is called transmission window. The time-based
traffic scheduling must be coordinated among the DetNet transit nodes traffic scheduling must be coordinated among the DetNet transit nodes
along the path from sender to receiver, to control the transmission along the path from sender to receiver, to control the transmission
of time-sensitive traffic. of time-sensitive traffic.
skipping to change at page 16, line 11 skipping to change at page 16, line 47
synchronized network and coordinated GCL configuration. Synthesis of synchronized network and coordinated GCL configuration. Synthesis of
GCL on multiple nodes in network is a scheduling problem considering GCL on multiple nodes in network is a scheduling problem considering
all TSN/DetNet flows traversing the network, which is a non- all TSN/DetNet flows traversing the network, which is a non-
deterministic polynomial-time hard (NP-hard) problem. Also, at this deterministic polynomial-time hard (NP-hard) problem. Also, at this
writing, scheduled traffic service supports no more than eight writing, scheduled traffic service supports no more than eight
traffic classes, typically using up to seven priority classes and at traffic classes, typically using up to seven priority classes and at
least one best effort class. least one best effort class.
6.4. Credit-Based Shaper with Asynchronous Traffic Shaping 6.4. Credit-Based Shaper with Asynchronous Traffic Shaping
Consider a network with a set of nodes (DetNet transit nodes and In the cosidered queuing model, there are four types of flows,
hosts) along with a set of flows between hosts. Hosts are sources or namely, control-data traffic (CDT), class A, class B, and best effort
destinations of flows. There are four types of flows, namely, (BE) in decreasing order of priority. Flows of classes A and B are
control-data traffic (CDT), class A, class B, and best effort (BE) in together referred to AVB flows. This model is a subset of Time-
decreasing order of priority. Flows of classes A and B are together Sensitive Networking as described next.
referred to AVB flows. It is assumed a subset of TSN functions as
described next.
It is also assumed that contention occurs only at the output port of Based on the timing model described in Figure 1, the contention
a TSN node. Each node output port performs per-class scheduling with occurs only at the output port of a relay node; therefore, the focus
eight classes: one for CDT, one for class A traffic, one for class B of the rest of this subsection is on the regulator and queuing
traffic, and five for BE traffic denoted as BE0-BE4 (according to TSN subsystem in the output port of a relay node. The output port
standard). In addition, each node output port also performs per-flow performs per-class scheduling with eight classes (queuing
regulation for AVB flows using an interleaved regulator (IR), called subsystems): one for CDT, one for class A traffic, one for class B
Asynchronous Traffic Shaper (ATS) in TSN. Thus, at each output port traffic, and five for BE traffic denoted as BE0-BE4. The queuing
of a node, there is one interleaved regulator per-input port and per- policy for each queuing subsystem is FIFO. In addition, each node
class. The detailed picture of scheduling and regulation output port also performs per-flow regulation for AVB flows using an
architecture at a node output port is given by Figure 4. The packets interleaved regulator (IR), called Asynchronous Traffic Shaper
received at a node input port for a given class are enqueued in the [IEEE8021Qcr]. Thus, at each output port of a node, there is one
respective interleaved regulator at the output port. Then, the interleaved regulator per-input port and per-class; the interleaved
packets from all the flows, including CDT and BE flows, are enqueued regulator is mapped to the regulator depicted in Figure 1. The
in a class based FIFO system (CBFS) [TSNwithATS]. detailed picture of scheduling and regulation architecture at a node
output port is given by Figure 4. The packets received at a node
input port for a given class are enqueued in the respective
interleaved regulator at the output port. Then, the packets from all
the flows, including CDT and BE flows, are enqueued in queuing
subsytem; there is no regulator for such classes.
+--+ +--+ +--+ +--+ +--+ +--+ +--+ +--+
| | | | | | | | | | | | | | | |
|IR| |IR| |IR| |IR| |IR| |IR| |IR| |IR|
| | | | | | | | | | | | | | | |
+-++XXX++-+ +-++XXX++-+ +-++XXX++-+ +-++XXX++-+
| | | | | | | |
| | | | | | | |
+---+ +-v-XXX-v-+ +-v-XXX-v-+ +-----+ +-----+ +-----+ +-----+ +-----+ +---+ +-v-XXX-v-+ +-v-XXX-v-+ +-----+ +-----+ +-----+ +-----+ +-----+
| | | | | | |Class| |Class| |Class| |Class| |Class| | | | | | | |Class| |Class| |Class| |Class| |Class|
skipping to change at page 17, line 28 skipping to change at page 17, line 48
| +-v-+ +-v-+ | | | | | | +-v-+ +-v-+ | | | | |
| |CBS| |CBS| | | | | | | |CBS| |CBS| | | | | |
| +-+-+ +-+-+ | | | | | | +-+-+ +-+-+ | | | | |
| | | | | | | | | | | | | | | |
+-v--------v-----------v---------v-------V-------v-------v-------v--+ +-v--------v-----------v---------v-------V-------v-------v-------v--+
| Strict Priority selection | | Strict Priority selection |
+--------------------------------+----------------------------------+ +--------------------------------+----------------------------------+
| |
V V
Figure 4: Architecture of a TSN node output port with interleaved Figure 4: The architecture of an output port inside a relay node with
regulators (IRs) interleaved regulators (IRs) and credit-based shaper (CBS)
The CBFS includes two Credit-Based Shaper (CBS) subsystems, one for Each of the queuing subsystems for class A and B, contains Credit-
each class A and B. The CBS serves a packet from a class according Based Shaper (CBS). The CBS serves a packet from a class according
to the available credit for that class. The credit for each class A to the available credit for that class. The credit for each class A
or B increases based on the idle slope, and decreases based on the or B increases based on the idle slope, and decreases based on the
send slope, both of which are parameters of the CBS. The CDT and send slope, both of which are parameters of the CBS (Section 8.6.8.2
BE0-BE4 flows in the CBFS are served by separate FIFO subsystems. of [IEEE8021Q]). The CDT and BE0-BE4 flows are served by separate
Then, packets from all flows are served by a transmission selection queuing subsystems. Then, packets from all flows are served by a
subsystem that serves packets from each class based on its priority. transmission selection subsystem that serves packets from each class
All subsystems are non-preemptive. Guarantees for AVB traffic can be based on its priority. All subsystems are non-preemptive.
provided only if CDT traffic is bounded; it is assumed that the CDT Guarantees for AVB traffic can be provided only if CDT traffic is
traffic has leaky bucket arrival curve with two parameters r_h as bounded; it is assumed that the CDT traffic has leaky bucket arrival
rate and b_h as bucket size, i.e., the amount of bits entering a node curve with two parameters r_h as rate and b_h as bucket size, i.e.,
within a time interval t is bounded by r_h t + b_h. the amount of bits entering a node within a time interval t is
bounded by r_h t + b_h.
Additionally, it is assumed that the AVB flows are also regulated at Additionally, it is assumed that the AVB flows are also regulated at
their source according to leaky bucket arrival curve. At the source their source according to leaky bucket arrival curve. At the source,
hosts, the traffic satisfies its regulation constraint, i.e. the the traffic satisfies its regulation constraint, i.e. the delay due
delay due to interleaved regulator at hosts is ignored. to interleaved regulator at source is ignored.
At each DetNet transit node implementing an interleaved regulator, At each DetNet transit node implementing an interleaved regulator,
packets of multiple flows are processed in one FIFO queue; the packet packets of multiple flows are processed in one FIFO queue; the packet
at the head of the queue is regulated based on its leaky bucket at the head of the queue is regulated based on its leaky bucket
parameters; it is released at the earliest time at which this is parameters; it is released at the earliest time at which this is
possible without violating the constraint. The regulation parameters possible without violating the constraint. The regulation parameters
for a flow (leaky bucket rate and bucket size) are the same at its for a flow (leaky bucket rate and bucket size) are the same at its
source and at all DetNet transit nodes along its path. A delay bound source and at all DetNet transit nodes along its path.
of CBFS for an AVB flow f of class A or B can be computed if the
following condition holds:
sum of leaky bucket rates of all flows of this class at this node 6.4.1. Delay Bound Calculation
<= R, where R is given below for every class.
If the condition holds, the delay bound is: A delay bound of the queuing subsystem ([4] in Figure 1) for an AVB
flow of class A or B can be computed if the following condition
holds:
d_f = T + (b_t-L_min_f)/R - L_min_f/c sum of leaky bucket rates of all flows of this class at this
transit node <= R, where R is given below for every class.
where L_min_f is the minimum packet length of flow f; c is the output If the condition holds, the delay bounds for a flow of class X (A or
link transmission rate; b_t is the sum of the b term (bucket size) B) is d_X and calculated as:
for all the flows having the same class as flow f at this node.
Parameters R and T are calculated as follows for class A and class B,
separately:
If f is of class A: d_X = T_X + (b_t_X-L_min_X)/R_X - L_min_X/c
R = I_A (c-r_h)/ c where L_min_X is the minimum packet lengths of class X (A or B); c is
the output link transmission rate; b_t_X is the sum of the b term
(bucket size) for all the flows of the class X. Parameters R_X and
T_X are calculated as follows for class A and class B, separately:
T = L_nA + b_h + r_h L_n/c)/(c-r_h) If the flow is of class A:
R_A = I_A (c-r_h)/ c
T_A = L_nA + b_h + r_h L_n/c)/(c-r_h)
where L_nA is the maximum packet length of class B and BE packets; where L_nA is the maximum packet length of class B and BE packets;
L_n is the maximum packet length of classes A,B, and BE. L_n is the maximum packet length of classes A,B, and BE.
If f is of class B: If the flow is of class B:
R = I_B (c-r_h)/ c R_B = I_B (c-r_h)/ c
T = (L_BE + L_A + L_nA I_A/(c_h-I_A) + b_h + r_h L_n/c)/(c-r_h) T_B = (L_BE + L_A + L_nA I_A/(c_h-I_A) + b_h + r_h L_n/c)/(c-r_h)
where L_A is the maximum packet length of class A; L_BE is the where L_A is the maximum packet length of class A; L_BE is the
maximum packet length of class BE. maximum packet length of class BE.
Then, an end-to-end delay bound is calculated by the formula Then, an end-to-end delay bound of class X (A or B)is calculated by
Section 4.2.2, where for Cij: the formula Section 4.2.2, where for Cij:
Cij = max(d_f')
where f' is any flow that shares the same CBFS class with flow f at Cij = d_X
node i and the same interleaved regulator as flow f at node j.
More information of delay analysis in such a DetNet transit node is More information of delay analysis in such a DetNet transit node is
described in [TSNwithATS]. described in [TSNwithATS].
6.4.1. Flow Admission 6.4.2. Flow Admission
The delay calculation requires some information about each node. For The delay bound calculation requires some information about each
each node, it is required to know the idle slope of CBS for each node. For each node, it is required to know the idle slope of CBS
class A and B (I_A and I_B), as well as the transmission rate of the for each class A and B (I_A and I_B), as well as the transmission
output link (c). Besides, it is necessary to have the information on rate of the output link (c). Besides, it is necessary to have the
each class, i.e. maximum packet length of classes A, B, and BE. information on each class, i.e. maximum packet length of classes A,
Moreover, the leaky bucket parameters of CDT (r_h,b_h) should be B, and BE. Moreover, the leaky bucket parameters of CDT (r_h,b_h)
known. To admit a flow/flows, their delay requirements should be should be known. To admit a flow/flows, their delay requirements
guaranteed not to be violated. As described in Section 3.1, the two should be guaranteed not to be violated. As described in
problems static and dynamic are addressed separately. In either of Section 3.1, the two problems, static and dynamic, are addressed
the problems, the rate and delay should be guaranteed. Thus, separately. In either of the problems, the rate and delay should be
guaranteed. Thus,
The static admission control: The static admission control:
The leaky bucket parameters of all flows are known, The leaky bucket parameters of all flows are known,
therefore, for each flow a delay bound can be calculated. therefore, for each flow f, a delay bound can be calculated.
The computed delay bound for every flow should not be more The computed delay bound for every flow should not be more
than its delay requirement. Moreover, the sum of the rate of than its delay requirement. Moreover, the sum of the rate of
each flow (r_f) should not be more than the rate allocated to each flow (r_f) should not be more than the rate allocated to
each class (R). If these two conditions hold, the each class (R). If these two conditions hold, the
configuration is declared admissible. configuration is declared admissible.
The dynamic admission control: The dynamic admission control:
For dynamic admission control, we allocate to every node and For dynamic admission control, we allocate to every node and
class A or B, static value for rate (R) and maximum class A or B, static value for rate (R) and maximum
burstiness (b_t). In addition, for every node and every burstiness (b_t). In addition, for every node and every
class A and B, two counters are maintained: class A and B, two counters are maintained:
R_acc is equal to the sum of the leaky-bucket rates of all R_acc is equal to the sum of the leaky-bucket rates of all
flows of this class already admitted at this node; At all flows of this class already admitted at this node; At all
times, we must have: times, we must have:
R_acc <=R, (Eq. 1) R_acc <=R, (Eq. 1)
skipping to change at page 20, line 21 skipping to change at page 20, line 41
The choice of the static values of R and b_t at all nodes and classes The choice of the static values of R and b_t at all nodes and classes
must be done in a prior configuration phase; R controls the bandwidth must be done in a prior configuration phase; R controls the bandwidth
allocated to this class at this node, b_t affects the delay bound and allocated to this class at this node, b_t affects the delay bound and
the buffer requirement. R must satisfy the constraints given in the buffer requirement. R must satisfy the constraints given in
Annex L.1 of [IEEE8021Q]. Annex L.1 of [IEEE8021Q].
6.5. IntServ 6.5. IntServ
Integrated service (IntServ) is an architecture that specifies the Integrated service (IntServ) is an architecture that specifies the
elements to guarantee quality of service (QoS) on networks. To elements to guarantee quality of service (QoS) on networks. [[E: The
satisfied guaranteed service, a flow must conform to a traffic rest of this paragraph is better not to be placed here; these should
specification (T-spec), and reservation is made along a path, only if be mentioned (is mentioned) in the introduction.]] To satisfied
routers are able to guarantee the required bandwidth and buffer. guaranteed service, a flow must conform to a traffic specification
(T-spec), and reservation is made along a path, only if routers are
able to guarantee the required bandwidth and buffer.
[[E: The information about arrival and service curves can be shorter
with less detail. I put a proposed text after description of
these.]]
Consider the traffic model which conforms to token bucket regulator Consider the traffic model which conforms to token bucket regulator
(r, b), with (r, b), with
o Token bucket depth (b). o Token bucket depth (b).
o Token bucket rate (r). o Token bucket rate (r).
The traffic specification can be described as an arrival curve: The traffic specification can be described as an arrival curve:
skipping to change at page 20, line 48 skipping to change at page 21, line 31
the number of bit for the flow is limited by alpha(t) = b + rt. the number of bit for the flow is limited by alpha(t) = b + rt.
If resource reservation on a path is applied, IntServ model of a If resource reservation on a path is applied, IntServ model of a
router can be described as a rate-latency service curve beta(t). router can be described as a rate-latency service curve beta(t).
beta(t) = max(0, R(t-T)) beta(t) = max(0, R(t-T))
It describes that bits might have to wait up to T before being served It describes that bits might have to wait up to T before being served
with a rate greater or equal to R. with a rate greater or equal to R.
It should be noted that, the guaranteed service rate R is a share of [[E: proposed text:
link's bandwidth. The choice of R is related to the specification of
flows which will transmit on this node. For example, in strict The flow, at the source, has a leaky bucket arrival curve with two
priority policy, considering a flow with priority j, its share of parameters r as rate and b as bucket size, i.e., the amount of bits
bandwidth may be R=c-sum(r_i), i<j, where c is the link bandwidth, entering a node within a time interval t is bounded by r t + b.
r_i is the token bucket rate for the flows with priority higher than
j. The choice of T is also related to the specification of all the If a resource reservation on a path is applied, a node provides a
flows traversing this node. For example, in a generalized processor guaranteed rate R and maximum service latency of T. This can be
interpreted in a way that the bits might have to wait up to T before
being served with a rate greater or equal to R. ]]
It should be noted that the guaranteed service rate R is a portion of
link's bandwidth. The selection of R is related to the specification
of flows traversing through the current node. For example, in strict
priority policy, considering a flow with priority i, its guaranteed
rate is R=c-sum(r_j), j<i, where c is the link bandwidth, r_j is the
token bucket rate for a flow j with priority higher than flow i. The
choice of T is also related to the specification of all the flows
traversing this node. For example, in a generalized processor
sharing (GPS) node, T = L / R + L_max/c, where L is the maximum sharing (GPS) node, T = L / R + L_max/c, where L is the maximum
packet size for the flow, L_max is the maximum packet size in the packet size for the flow, L_max is the maximum packet size in the
node across all flows. Other choice of R and T are also supported, node across all flows. Other choice of R and T are also supported,
according to the specific scheduling of the node and flows traversing according to the specific scheduling of the node and flows traversing
this node. this node.
As mentioned previously in this section, delay bound and backlog As mentioned previously in this section, a delay bound and a buffer
bound can be easily obtained by comparing arrival curve and service size bound can be easily obtained by comparing arrival curve and
curve. Backlog bound, or buffer bound, is the maximum vertical service curve. Backlog bound, or buffer bound, is the maximum
derivation between curves alpha(t) and beta(t), which is v=b+rT. vertical derivation between curves alpha(t) and beta(t), which is
Delay bound is the maximum horizontal derivation between curves v=b+rT. Delay bound is the maximum horizontal derivation between
alpha(t) and beta(t), which is h = T+b/R. Graphical illustration of curves alpha(t) and beta(t), which is h = T+b/R. Graphical
the IntServ model is shown in Figure 5. illustration of the IntServ model is shown in Figure 5.
+ bit . * + bit . *
| . * | . *
| . * | . *
| * | *
| * . | * .
| * . | * .
| * | . .. Service curve | * | . .. Service curve
*-----h-|---. ** Arrival curve *-----h-|---. ** Arrival curve
| v . h Delay_bound | v . h Delay_bound
skipping to change at page 22, line 21 skipping to change at page 23, line 15
Similarly, delay bound, backlog bound and output bound can be Similarly, delay bound, backlog bound and output bound can be
computed by using the original arrival curve alpha(t) and computed by using the original arrival curve alpha(t) and
concatenated service curve beta_e2e(t). concatenated service curve beta_e2e(t).
6.6. Cyclic Queuing and Forwarding 6.6. Cyclic Queuing and Forwarding
Annex T of [IEEE8021Q] describes Cyclic Queuing and Forwarding (CQF), Annex T of [IEEE8021Q] describes Cyclic Queuing and Forwarding (CQF),
which provides bounded latency and zero congestion loss using the which provides bounded latency and zero congestion loss using the
time-scheduled gates of [IEEE8021Q] section 8.6.8.4. For a given time-scheduled gates of [IEEE8021Q] section 8.6.8.4. For a given
DetNet class of service, a set of two or three buffers is provided at DetNet class of service, a set of two or three buffers is provided at
the output queue layer of Figure 3. A cycle time Tc is configured the output queue layer of Figure 3. A cycle time T_c is configured
for each class c, and all of the buffer sets in a class swap buffers for each class c, and all of the buffer sets in a class swap buffers
simultaneously throughout the DetNet domain at that cycle rate, all simultaneously throughout the DetNet domain at that cycle rate, all
in phase. in phase.
0 time --> 0.7 1 (units of Tc) 2 3 0 time --> 0.7 1 (units of T_c) 2 3
DetNet transit node A out port 1 DetNet transit node A out port 1
| a <-DT->| b | c | d | a <-DT->| b | c | d
+------------+------+-------------------+-------------------+-------- +------------+------+-------------------+-------------------+--------
\_____ \_____ \_____ \_____
\_____ \_____ queue-to-queue delay = 1.3 Tc \_____ \_____ queue-to-queue delay = 1.3 T_c
\_____ \_____ \_____ \_____
\_____ \_____ DetNet transit node B \_____ \_____ DetNet transit node B
\_ \_ queue assignment, in \_ \_ queue assignment, in
| | |<-DT->| port 2 to out 3 | | | |<-DT->| port 2 to out 3 |
-------+-------------------+------------+------+-------------------+- -------+-------------------+------------+------+-------------------+-
0.3 time--> 1.3 2.0 2.3 3.3 0.3 time--> 1.3 2.0 2.3 3.3
window to transfer window to transfer
to buffer c ---> VVVVVVVVVVVV to buffer c ---> VVVVVVVVVVVV
if dead time not window to transfer if dead time not window to transfer
skipping to change at page 23, line 20 skipping to change at page 24, line 14
3. The output queues on port 2 of node B. 3. The output queues on port 2 of node B.
In this figure, the output ports on the two nodes are synchronized, In this figure, the output ports on the two nodes are synchronized,
and a new buffer starts transmitting at each tick, shown as 0, 1, 2, and a new buffer starts transmitting at each tick, shown as 0, 1, 2,
... The output times shown for timelines 1 and 3 are the times at ... The output times shown for timelines 1 and 3 are the times at
which packets are selected for output, which is the start point of which packets are selected for output, which is the start point of
the output time (1) of Figure 1. The queue assignments times on the output time (1) of Figure 1. The queue assignments times on
timeline 3 take place at the beginning of the queuing delay (6) of timeline 3 take place at the beginning of the queuing delay (6) of
Figure 1. Time-based CQF, as described here, does not require any Figure 1. Time-based CQF, as described here, does not require any
regulator queues. In the shown in the figure, the total time for regulator queues. In the shown in the figure, the total time [[E:
delays 1 through 6 of Figure 1 is 1.3Tc. Of course, any value is what is meant by total time? Does it mean a delay bound is 1.3
possible. T_C?]] for delays (1) through (6) of Figure 1, is 1.3T_c. Of course,
any value is possible.
6.6.1. CQF timing sequence 6.6.1. CQF timing sequence
In general, as shown in Figure 6, the windows for buffer assignment In general, as shown in Figure 6, the windows for buffer assignment
do not align perfectly with the windows for buffer transmission. The do not align perfectly with the windows for buffer transmission. The
input gates (the center timeline in Figure 6) must switch from using input gates (the center timeline in Figure 6) must switch from using
one buffer to using another buffer in sync with the (delayed) one buffer to using another buffer in sync with the (delayed)
received data, at times offset by the dead time from the output received data, at times offset by the dead time from the output
buffer switching (the bottom timeline in Figure 6). buffer switching (the bottom timeline in Figure 6).
skipping to change at page 24, line 10 skipping to change at page 24, line 51
from node A buffer a into node B buffer d between the times 1.3 and from node A buffer a into node B buffer d between the times 1.3 and
2.3 in Figure 6. Buffer b starts outputting at time = 2.0, while 2.3 in Figure 6. Buffer b starts outputting at time = 2.0, while
buffer d is filling. Thus, three buffers are in use, one filling, buffer d is filling. Thus, three buffers are in use, one filling,
one waiting, and one emptying. one waiting, and one emptying.
6.6.2. CQF latency calculation 6.6.2. CQF latency calculation
The per-hop latency is trivially determined by the wire delay and the The per-hop latency is trivially determined by the wire delay and the
queuing delay. Since the wire delay is either absorbed into the queuing delay. Since the wire delay is either absorbed into the
queueing delay (dead time is small and two buffers are used) or queueing delay (dead time is small and two buffers are used) or
padded out to a whole cycle time Tc (three buffers are used) the per- padded out to a whole cycle time T_c (three buffers are used) the
hop latency is always an integral number of cycle times Tc, with a per-hop latency is always an integral number of cycle times T_c, with
latency variation at the output of the final hop of Tc. a latency variation at the output of the final hop of T_c.
Ingress conditioning (Section 4.3) may be required if the source of a Ingress conditioning (Section 4.3) may be required if the source of a
DetNet flow does not, itself, employ CQF. DetNet flow does not, itself, employ CQF.
Note that there are no per-flow parameters in the CQF technique. Note that there are no per-flow parameters in the CQF technique.
Therefore, there is no requirement for per-hop configuration when a Therefore, there is no requirement for per-hop configuration when a
new DetNet flow is added to a network, except perhaps for ingress new DetNet flow is added to a network, except perhaps for ingress
checks to see that the transmitter does not exceed the contracted checks to see that the transmitter does not exceed the contracted
bandwidth. bandwidth.
 End of changes. 67 change blocks. 
190 lines changed or deleted 231 lines changed or added

This html diff was produced by rfcdiff 1.47. The latest version is available from http://tools.ietf.org/tools/rfcdiff/