Internet DRAFT - draft-peng-lsr-flex-algo-deterministic-routing
draft-peng-lsr-flex-algo-deterministic-routing
LSR S. Peng
Internet-Draft ZTE
Intended status: Standards Track T. Li
Expires: 26 February 2023 Juniper Networks
25 August 2022
IGP Flexible Algorithm with Deterministic Routing
draft-peng-lsr-flex-algo-deterministic-routing-03
Abstract
IGP Flexible Algorithm proposes a solution that allows IGPs to
compute constraint based paths over the network and specifies a way
of using Segment Routing (SR) Prefix-SIDs, SRv6 locators, or pure IP
prefixes to steer packets along the constraint-based paths. This
document describes how to compute deterministic delay paths within a
Flex-Algorithm topology.
Status of This Memo
This Internet-Draft is submitted in full conformance with the
provisions of BCP 78 and BCP 79.
Internet-Drafts are working documents of the Internet Engineering
Task Force (IETF). Note that other groups may also distribute
working documents as Internet-Drafts. The list of current Internet-
Drafts is at https://datatracker.ietf.org/drafts/current/.
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."
This Internet-Draft will expire on 26 February 2023.
Copyright Notice
Copyright (c) 2022 IETF Trust and the persons identified as the
document authors. All rights reserved.
Peng & Li Expires 26 February 2023 [Page 1]
Internet-Draft Deterministic Routing August 2022
This document is subject to BCP 78 and the IETF Trust's Legal
Provisions Relating to IETF Documents (https://trustee.ietf.org/
license-info) in effect on the date of publication of this document.
Please review these documents carefully, as they describe your rights
and restrictions with respect to this document. Code Components
extracted from this document must include Revised BSD License text as
described in Section 4.e of the Trust Legal Provisions and are
provided without warranty as described in the Revised BSD License.
Table of Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3
2. Requirements Language . . . . . . . . . . . . . . . . . . . . 4
3. Determinisitc Links . . . . . . . . . . . . . . . . . . . . . 4
3.1. Deterministic Link Bounds with CQF . . . . . . . . . . . 6
3.2. Deterministic Link Bounds with Deadlines . . . . . . . . 7
4. Deterministic Delay Metric Extension to IS-IS . . . . . . . . 8
4.1. CQF Intra-Node Scheduling Delay Sub-Sub-TLV . . . . . . . 9
4.2. Deadline Intra-Node Scheduling Delay Sub-Sub-TLV . . . . 11
4.2.1. Simplified Deadline Intra-Node Scheduling Delay
TLV . . . . . . . . . . . . . . . . . . . . . . . . . 12
5. Deterministic Delay Metric Extension to OSPF . . . . . . . . 13
6. Announcement Suppression . . . . . . . . . . . . . . . . . . 13
7. Computing Deterministic Routes . . . . . . . . . . . . . . . 13
7.1. Advertising CQF Scheduling Parameters in a FAD . . . . . 14
7.1.1. IS-IS Advertisement of CQF Scheduling Parameters . . 14
7.1.2. OSPF Advertisement of Flex-algo Binding CQF . . . . . 14
7.2. Advertising Deadline Scheduling Parameters in a FAD . . . 15
7.2.1. IS-IS Advertisement of Deadline Scheduling
Parameters . . . . . . . . . . . . . . . . . . . . . 15
7.2.2. OSPF Advertisement of Flex-algo Binding Deadline . . 16
7.3. CQF Deterministic Route Computation . . . . . . . . . . . 16
7.4. Deadline Deterministic Route Computation . . . . . . . . 17
8. Routing Convergence and Redundancy Considerations . . . . . . 18
9. Examples of Deterministic Delay SPF . . . . . . . . . . . . . 20
9.1. CQF Deterministic Delay SPF . . . . . . . . . . . . . . . 20
9.2. Deadline Deterministic Delay SPF . . . . . . . . . . . . 21
10. Use Cases . . . . . . . . . . . . . . . . . . . . . . . . . . 23
11. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 24
11.1. IS-IS Deterministic Delay Metric Sub-TLV . . . . . . . . 24
11.2. Registry for Sub-Sub-TLVs for the Deterministic Delay
Metric Sub-TLV . . . . . . . . . . . . . . . . . . . . . 24
11.3. IGP Metric-Type Registration . . . . . . . . . . . . . . 25
11.4. IS-IS Sub-Sub-TLVs for Flexible Algorithm Definition
Sub-TLV . . . . . . . . . . . . . . . . . . . . . . . . 25
11.5. OSPF IANA considerations . . . . . . . . . . . . . . . . 25
12. Security Considerations . . . . . . . . . . . . . . . . . . . 25
13. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . 25
Peng & Li Expires 26 February 2023 [Page 2]
Internet-Draft Deterministic Routing August 2022
14. References . . . . . . . . . . . . . . . . . . . . . . . . . 26
14.1. Normative References . . . . . . . . . . . . . . . . . . 26
14.2. Informative References . . . . . . . . . . . . . . . . . 28
Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 28
1. Introduction
IGP Flexible Algorithm [I-D.ietf-lsr-flex-algo] (Flex-Algorithm)
proposes a solution that allows IGPs to compute constraint based
paths over the network and specifies a way of using Segment Routing
[RFC8402] Prefix-SIDs, SRv6 locators, or pure IP prefixes
[I-D.ietf-lsr-ip-flexalgo] to steer packets along the constraint-
based paths. It specifies a set of extensions to IS-IS, OSPFv2 and
OSPFv3 that enable a router to send TLVs that identify (a)
calculation-type, (b) specify a metric-type, and (c )describe a set
of constraints on the topology, that are to be used to compute the
best paths along the constrained topology. A given combination of
calculation-type, metric-type, and constraints is known as a FAD
(Flexible Algorithm Definition).
[RFC8655] describes the architecture of a deterministic network and
defines the QoS goals of deterministic forwarding:
* Minimum and maximum end-to-end latency from source to destination,
timely delivery, and bounded jitter (packet delay variation)
* A bounded packet loss ratio under various assumptions about the
operational states of the nodes and links
* An upper bound on out-of-order packet delivery.
In order to achieve these goals, deterministic networks use resource
reservation, explicit routing, and service protection, as well as
other means. A deterministic path is typically (but not necessarily)
an explicit route so that it does not suffer temporary interruptions
caused by the convergence of routing or bridging protocols.
Flexible-Algorithm has the characteristic mentioned in [RFC8655] that
it operates under a single administrative control or within a closed
group of administrative control. Flexible-Algorithm also supports
Min Unidirectional Link Delay (defined in [RFC8570]) metric type to
compute shortest paths with minimum delay, however, the cumulative
delay is essentially the accumulation of propagation delay of all
links, excluding node delay. In order to compensate for this gap, it
is necessary to enhance Flexible-Algorithm to compute the path with
deterministic delay, i.e., including deterministic node delay as
well.
Peng & Li Expires 26 February 2023 [Page 3]
Internet-Draft Deterministic Routing August 2022
This document describes how to compute distributed shortest paths
with a deterministic delay metric within a Flexible-Algorithm
topology, as the basis of the whole distributed deterministic scheme.
It should be noted that relying on this enhancement alone does not
guarantee complete determinacy, it needs to be used in conjunction
with other tools, such as creating additional redundant deterministic
delay paths with consistent delay metric for PREOF (Packet
Replication, Elimination, and Ordering Functions), smoothing the
delay jitter during route convergence, providing a deterministic
forwarding mechanism, admission control, etc.
2. Requirements Language
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
"SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and
"OPTIONAL" in this document are to be interpreted as described in BCP
14 [RFC2119] [RFC8174] when, and only when, they appear in all
capitals, as shown here.
3. Determinisitc Links
When a packet is forwarded to a link, the delay produced includes two
parts: the first part is the dwell delay of the packet in the node,
and the second part is the propagation delay of the packet on the
link. In packet switching networks, a priority based queuing scheme
is sometimes used. It may give better average latency, but may have
unacceptable worst case latency. [SP-LATENCY] analyzes the
guaranteed latency with the traditional strict priority scheme, and
shows that low bounded latency is achievable when high priority
traffic is constrained in low utilization, but deteriorates quickly
with increasing amounts of high priority traffic. DiffServ [RFC2475]
with strict priority has been deployed in the network and existing
non-deterministic service flows may elect the highest priority, so it
is difficult to support deterministic services based on DiffServ
without any modification.
We call a link with a queuing mechanism that does not guarantee a
bounded delay a non-determinisitc link and a link with a queuing
mechanism that can provide deterministic node delay is called a
deterministic link. To achieve a deterministic network, other new
scheduling mechanisms need to be introduced, and their scheduling
priority must be higher than that of the traditional strict priority
queue. The typical queuing mechanisms are as follows:
* IEEE 802.1 WG has specified IEEE802.1Qav [CBS] which uses a
credit-based shaper mechanism to assign packets to different
queues based on a credit value that is proportional to the flow's
reserved bandwidth. The credit values of different transmission
Peng & Li Expires 26 February 2023 [Page 4]
Internet-Draft Deterministic Routing August 2022
queues will automatically change with the packet transmission
process, which will ensure that packets with lower priority will
also get transmitted. The CBS shaper mechanism is similar to
Weighted Fair Queuing (WFQ) scheme, and they both control the
scheduling of packets based on the reserved bandwidth. The worst-
case delay calculation of class A CBS traffic is relatively
simple, but other classes are complex. For class A traffic, the
queuing delay equals to the maximum size of the interference frame
(such as 2000 octets) divided by the port bandwidth.
* IEEE 802.1 WG has specified IEEE802.1Qch [CQF] which uses a cyclic
queuing and forwarding (CQF) mechanism and relies on time
synchronization. Under CQF, the maximum delay experienced by a
given packet is (H+1)*D, the minimum delay experienced by a given
packet is (H-1)*D, and the delay jitter is 2*D, where H is the
number of hops and D is cycle duration. Other variants based on
CQF can avoid relying on time synchronization, but require the
same cycle duration for all nodes. Basically, if a packet
received in the current sending window (i.e., cycle) can always be
sent in the next sending window, then the deterministic node
delay, on average, is one cycle duration, or several cycle
durations if the intra-node forwarding delay (from incoming port
to outgoing port) can't be ignored.
* [I-D.peng-detnet-deadline-based-forwarding] introduced a deadline
based forwarding mechanism that allows a packet to control its
expected dwell time in the node according to a planned deadline.
There are two policies for scheduling packets in deadline queuing.
For the in-time policy, the end-to-end delay is H*(F~D) and the
jitter is H*Q, where H is the number of hops, F is the intra-node
forwarding delay, D is the planned deadline, and Q is the
scheduling delay. For the on-time policy, the end-to-end delay is
H*D, and the jitter is 0 (however there may be a one authorization
time delay due to the granularity of queue scheduling). That is,
a packet received can be sent within F~D or D respectively for
these two policies.
We define authorization time to be the time period that a given
deadline queue is allowed to transmit the stored packets.
Deadline queue participates in priority based scheduling. A
deadline queue with the highest priority will maintain its highest
priority for the predefined authorization time. During this
period, if it contains packets, it may obtain one or more
authorization size (e.g, an authorization size of 2000 bytes which
is generally a chip configurable fixed value) from the scheduling
decision module, to clear the queue. When it becomes empty, it
will withdraw the request and the scheduling opportunity will be
given to other queues with lower priority. For a deadline queue
Peng & Li Expires 26 February 2023 [Page 5]
Internet-Draft Deterministic Routing August 2022
with in-time mode, it always sends a request to the scheduling
decision module as long as it contains packets, regardless of its
priority. Instead, for a deadline queue with on-time mode, if it
contains packets, the request is sent only when its priority
becomes the highest.
This document discusses deterministic links based on CQF or the
deadline algorithm. Other algorithms will be described in the
future. Note that a deadline or CQF queue has the highest resource
access priority. In extreme cases, this kind of queue can access the
whole port bandwidth, but the queue does no really reserve bandwidth
and is independent of any specific service flow. However, the
operator can create and maintain the reserved bandwidth of the
service flow from the perspective of the service, and take admission
control on the ingress node.
3.1. Deterministic Link Bounds with CQF
A node may be configured with the parameters for CQF based packet
scheduling for its local links, including enabling CQF scheduling and
cycle durations, which in turn determine the node delay and jitter
attributes for each link. The meanings of these parameters are:
* CQF scheduling enabled: if the CQF scheduling algorithm is enabled
for a link, then the packets sent to that link will be scheduled
by the CQF scheduling algorithm.
* Cycle duration: the cycle duration, also called cycle_size. One
or more durations of different values can be configured for a
link, such as 10us, 20us, 30us, and so on.
* Node delay/jitter:
- According to classical CQF, for a given cycle_size, it can be
deduced that the minimum delay in the node for a packet is 0,
the maximum delay in the node is 2*cycle_size, the average
delay in the node is one cycle_size, and the delay jitter in
the node is 2*cycle_size. The detailed reasons for this are as
follows: if a node receives a packet at the tail end of cycle i
and sends that packet at the head end of cycle i+1, the
resulting node delay, i.e., the minimum node delay, is 0; if a
node receives a packet at the head end of cycle i and sends
that packet at the tail end of cycle i+1, the resulting node
delay, i.e., the maximum node delay, is 2*cycle_ size; the
average node delay is one cycle_size, and the node delay jitter
is 2*cycle_size. Each cycle_size corresponds to a different
set of delay/jitter attributes.
Peng & Li Expires 26 February 2023 [Page 6]
Internet-Draft Deterministic Routing August 2022
- However, for some variants based on CQF, if the intra-node
forwarding delay can't be ignored, e.g, wasting 2 cycle
duration, then the minimum node delay, the maximum node delay,
and the average node delay need to add 2 cycle_size
respectively, but the node delay jitter is still 2*cycle_size.
3.2. Deterministic Link Bounds with Deadlines
A node may be configured for deadline based packet scheduling for its
local links, including enabling deadline scheduling, one or more
deadline scheduling delays, and the scheduling policy supported for
each deadline scheduling delay. The deadline scheduling delay will
determine the node delay and jitter attributes of the link. The
meanings of these parameters are:
* Deadline scheduling enabled: if the deadline scheduling algorithm
is enabled for a link, then a packet forwarded to the link will be
scheduled by the deadline based packet scheduling algorithm. The
dwell time of the packet in the node will not exceed the maximum
allowable dwell time D, where, D = intra-node forwarding delay (F)
+ specific deadline scheduling delay (Q).
* Deadline scheduling delays: a set composed of one or more deadline
scheduling delays <Q1, Q2, ..., Qn>, assuming that Q1 is the
minimum and Qn is the maximum in the set. Generally, the
difference between two adjacent elements in the set is fixed, for
example, a fixed interval (I).
* Scheduling policy: for each scheduling delay Q, there are two
possible scheduling policies: the in-time policy and the on-time
policy. In the case of the in-time policy, the packet may be sent
to the outgoing port even when it's scheduling delay has not
reached Q. For the on-time policy, the packet is sent to the
outgoing port only when the scheduling delay of the packet is
equal to Q. Therefore, for in-time policy, the actual dwell time
of the packet in the node is within the range [F, F+Q], i.e., the
minimum node delay is F, the maximum node delay is F+Q, and the
node delay jitter is Q; For on-time policy, the actual dwell time
of the packet in the node is equal to F+Q, i.e., the minimum node
delay is F+Q, the maximum node delay is also F+Q, and the node
delay jitter is 0 (however, there may be one authorization time
delay due to the granularity of queue scheduling).
Peng & Li Expires 26 February 2023 [Page 7]
Internet-Draft Deterministic Routing August 2022
4. Deterministic Delay Metric Extension to IS-IS
This document defines a new sub-TLV, the Deterministic Delay Metric,
in the "IS-IS Sub-TLVs for TLVs Advertising Neighbor Information"
registry, to distribute deterministic delay information. The
deterministic delay advertised by this sub-TLV MUST be the delay from
the local neighbor to the remote neighbor (i.e., the forward-path
latency). The format of this sub-TLV is:
0 1 2 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Type | Length |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|A| Reserved | Link propagation delay |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Intra-Node Forwarding Delay |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
// sub-sub-TLVs for Intra-Node Scheduling Delay //
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Figure 1
where:
Type: TBA1
Length: Variable
A bit: This field represents the Anomalous (A) bit. The A bit is
set when one or more measured values of link propagation delay
exceed a configured maximum threshold. The A bit is cleared when
the measured value falls below its configured reuse threshold. If
the A bit is cleared, the sub-TLV represents steady-state link
propagation delay.
Reserved: This field is reserved for future use. It MUST be set
to 0 when sent and MUST be ignored when received.
Link propagation delay: This 24-bit field carries the average
measured link propagation delay value (in microseconds) over a
configurable interval, encoded as an integer value.
Implementations MAY also permit the configuration of an offset
value (in microseconds) to be added to the measured delay value,
to facilitate the communication of operator-specific delay
constraints. When the delay value is set to the maximum value
16,777,215 (16.777215 seconds), then the delay is at least that
value and may be larger. Note that [RFC8570] also specified the
Peng & Li Expires 26 February 2023 [Page 8]
Internet-Draft Deterministic Routing August 2022
Unidirectional Link Delay Sub-TLV and the Min/Max Unidirectional
Link Delay Sub-TLV to advertise the average and min/max link delay
respectively. For simplicity, the Deterministic Delay Metric only
contains the average link delay. Although the exising TE or
Flexalgo path may be calculated with delay metric, which is
exactly Min Unidirectional Link Delay metric, to reduce
information flooding, for these applications, only Min/Max
Unidirectional Link Delay Sub-TLV, but not Unidirectional Link
Delay Sub-TLV, need to be advertised by IGP. Then, as
deterministic routing introduced, IGP should also advertise
Deterministic Delay Sub-TLV (which contains average link delay),
and Unidirectional Link Delay Sub-TLV still doesn't need to be
advertised.
Intra-Node Forwarding Delay: This 16-bit field carries the intra-
node forwarding delay value (in microseconds). It represents the
latency of a packet from reception on the incoming port (or
generated from control plane) to queuing on the outgoing port. If
the forwarding delay can be ignored, it is set to 0.
Note: the forwarding delay for a node may be identical for all
links on that node. In a future version of this document, we will
consider advertising a single forwarding delay for the node.
sub-sub-TLVs for Intra-Node Scheduling Delay: Optional sub-sub-
TLVs are included to indicate the scheduling delays that are
related to the scheduling algorithm such as CQF, deadline, etc.
If this field is absent, the scheduling delay is unknown.
Typically, a link may enable a single scheduling algorithm to get
deterministic scheduling delay, so that a single sub-sub-TLV is
included. However, it is possible for a link to enable multiple
different scheduling algorithms, as long as these algorithms can
coordinate the forwarding resources, in this case, multiple sub-
sub-TLVs are included. Supported Sub-sub-TLVs are specified in
the following sub-sections.
4.1. CQF Intra-Node Scheduling Delay Sub-Sub-TLV
The CQF Intra-Node Scheduling Delay Sub-Sub-TLV is an optional Sub-
Sub-TLV of the Deterministic Delay Metric Sub-TLV. Only one CQF
Intra-Node Scheduling Delay Sub-Sub-TLV MAY be included.
The CQF Intra-Node Scheduling Delay Sub-Sub-TLV has the format:
Peng & Li Expires 26 February 2023 [Page 9]
Internet-Draft Deterministic Routing August 2022
0 1 2 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Type | Length | Cycle_size 1 |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Cycle_size 2 | ... ... |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Cycle_size N |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Figure 2
where:
Type: 1
Length: 2*N, depending on the number of the cycle_sizes included
Cycle_size: The length of cycle duration, in units of
microseconds. A link can support multiple cycle durations, for
example, 10us, 20us, 30us, etc, each for a specific service
requirement.
Only those links that enable the CQF scheduling algorithm need to
advertise the Intra-Node CQF Scheduling Delay Sub-Sub-TLV.
Note that the advertised cycle_size must be consistent with the CQF
queue scheduling mechanism actually instantiated by the link in the
forwarding plane. If the forwarding plane does not instantiate a CQF
queue scheduling supporting a certain cycle_size, that is advertised
in the CQF Intra-Node Scheduling Delay Sub-Sub-TLV, the subsequent
route computation may get the wrong results.
For a given cycle_size, other nodes can deduce the corresponding node
delay and jitter attributes, so these attributes can no longer be
explicitly included in the CQF Intra-Node Scheduling Delay Sub-Sub-
TLV.
As mentioned earlier, if the intra-node forwarding delay, F, is not
0, then the minimum node delay, the maximum node delay, and the
average node delay need to take F into account respectively. F is
replaced by ((F/cycle_size)+1)*cycle_size for computation. That is:
* If F is 0 for a given cycle_size, the minimum node delay is 0, the
maximum node delay is 2*cycle_size, the average node delay is
cycle_size, and the node delay jitter is 2*cycle_size.
Peng & Li Expires 26 February 2023 [Page 10]
Internet-Draft Deterministic Routing August 2022
* If F is not 0, for a given cycle_size, the minimum node delay is
((F/cycle_size)+1)*cycle_size, the maximum node delay is ((F/
cycle_size)+3)*cycle_size, the average node delay is ((F/
cycle_size)+2)*cycle_size, and the node delay jitter is
2*cycle_size.
4.2. Deadline Intra-Node Scheduling Delay Sub-Sub-TLV
The Deadline Intra-Node Scheduling Delay Sub-Sub-TLV is an optional
Sub-Sub-TLV of the Deterministic Delay Metric Sub-TLV. Only one
Deadline Intra-Node Scheduling Delay Sub-Sub-TLV MAY be included.
The format of the Deadline Intra-Node Scheduling Delay Sub-Sub-TLV
is:
0 1 2 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Type | Length |P1 | Q1 |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|P2 | Q2 |P3 | Q3 |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| ... ... |Pn | Qn |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Figure 3
where:
Type: 2
Length: 2*n, for n deadline scheduling delay pairs, (Pn, Qn).
Pn: Two bits indicating the scheduling policy associated with the
nth scheduling delay. The value of Pn indicates:
0 reserved
1 the in-time policy
2 the on-time policy
3 both the in-time policy and the on-time policy
Qn: Fourteen bits indicating the nth scheduling delay, in units of
microseconds.
Peng & Li Expires 26 February 2023 [Page 11]
Internet-Draft Deterministic Routing August 2022
As mentioned earlier, given a scheduling delay Qn and its scheduling
policy Pn, combined with the intra-node forwarding delay F, the
corresponding delay and jitter attributes for the node can be
calculated. Therefore, the delay and jitter associated with the
queue are not explicitly included in the Deadline Intra-Node
Scheduling Delay Sub-Sub-TLV.
4.2.1. Simplified Deadline Intra-Node Scheduling Delay TLV
Frequently in deadline scheduling, the queuing delays are all
separated by a fixed interval and have a common policy. In this
case, it is more efficient to encode the set of queuing delays as a
policy, starting value, ending value, and interval. The format of
the Simplified Deadline Intra-Node Scheduling Delay Sub-Sub-TLV is:
0 1 2 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Type | Length | Q1 |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Qn | P | I |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Figure 4
where:
Type: TBD
Length: 6
Q1: the minimum scheduling delay, in units of microseconds.
Qn: the maximum scheduling delay, in units of microseconds.
P: Two bits indicating the scheduling policy for all delays. The
value of P indicates:
0 reserved
1 the in-time policy
2 the on-time policy
3 both the in-time policy and the on-time policy
I: the fixed interval between any two adjacent elements in the
set, in units of microseconds.
Peng & Li Expires 26 February 2023 [Page 12]
Internet-Draft Deterministic Routing August 2022
The values of Q1, Qn, and I MUST form a sequence. That is, Qn = Q1 +
I*m for some integer value of m greater than or equal to 0.
5. Deterministic Delay Metric Extension to OSPF
To be defined in next version.
6. Announcement Suppression
The value of the Deterministic Delay Metric contains both the node
delay and the link propagation delay. The node delay portion is
constant for the implementation of the scheduling algorithm.
However, the link propagation delay is provided by a dynamic
measurement mechanism. To prevent oscillations and unnecessary
advertisements, implementations MUST comply with the requirements
found in sections 5 and 6 of [RFC8570] regarding announcement
thresholds, filters, and suppression.
7. Computing Deterministic Routes
In order to use the deterministic link resources to compute a
deterministic delay path, a corresponding Flexible Algorithm
Definition needs to be created. To dinstinguish between a
traditional low latency path (based on metric type "Min
Unidirectional Link Delay") and the deterministic low latency path
introduced in this document, a new metric type, the Deterministic
Delay Metric, will be defined and used in the Flexible Algorithm
Definition (FAD).
* Metric-Type: TBA4: Deterministic Delay Metric, as defined in this
document, to indicate the calculation of deterministic low latency
paths.
Additional constraints are also necessary in the FAD, to specify the
queuing parameters that route computation should use. These are
specified below.
It is possible to create multiple Flexible Algorithm definitions,
each using different scheduling delays, for different service
requirements.
By convention, during a deterministic low latency SPF path
calculation, a node MUST consider its own node delay to be 0.
However, the SPF path calculation MUST always take transit node
delays into account.
Peng & Li Expires 26 February 2023 [Page 13]
Internet-Draft Deterministic Routing August 2022
7.1. Advertising CQF Scheduling Parameters in a FAD
If CQF is to be used during the path computation for a given FAD,
then the cycle_size parameter is advertised as part of the FAD. All
links that are within the FAD constraints MUST have CQF scheduling
enabled and use the advertised cycle_size.
7.1.1. IS-IS Advertisement of CQF Scheduling Parameters
A new IS-IS sub-sub-TLV is defined to indicate the use of CQF for a
FAD and advertise the CQF cycle_size: the FAD CQF Sub-Sub-TLV, which
is advertised within the IS-IS Flexible Algorithm Definition Sub-TLV.
Only one FAD CQF Sub-Sub-TLV MAY be included in the FAD. The FAD
MUST use the Metric Type for the Deterministic Delay Metric.
The FAD CQF Sub-Sub-TLV has the format:
0 1 2 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Type | Length | cycle_size |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Figure 5
where:
Type: TBA5
Length: 2
cycle_size: The cycle_size parameter for CQF scheduling, in units
of microseconds.
Inclusion of this sub-sub-TLV within a FAD indicates that nodes using
this FAD should compute paths using the CQF scheduling algorithm with
the advertised cycle_size.
7.1.2. OSPF Advertisement of Flex-algo Binding CQF
To be defined in next version.
Peng & Li Expires 26 February 2023 [Page 14]
Internet-Draft Deterministic Routing August 2022
7.2. Advertising Deadline Scheduling Parameters in a FAD
If deadline scheduling is to be used during the path computation for
a given FAD, then the scheduling delay and scheduling policy is
advertised as part of the FAD. All links that are within the FAD
constraints MUST have deadline scheduling enabled and use the
advertised parameters.
7.2.1. IS-IS Advertisement of Deadline Scheduling Parameters
The FAD Deadline Sub-Sub-TLV is advertised within the IS-IS Flexible
Algorithm Definition Sub-TLV and indicates the use of deadline
scheduling for a FAD and advertises the scheduling delay and the
scheduling policy. Only one FAD Deadline Sub-Sub-TLV MAY be included
in the FAD. The FAD MUST use the Metric Type for the Deterministic
Delay Metric.
The FAD Deadline Sub-Sub-TLV has the format:
0 1 2 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Type | Length | Flags |U| P | Q //
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
// Q |
+-+-+-+-+-+-+-+-+
Figure 6
where:
Type: TBD
Length: 3
Flags: Two flags are currently defined. The remaining bits are
reserved for future use. They must be zero when transmitted and
ignored upon receipt.
U: 1 bit, indicates whether the value of scheduling delay Q is
known or unknown. 0 indicates known and 1 indicates unknown.
P: 2 bits, indicating scheduling policy. The value can be:
0 reserved
1 the in-time policy
Peng & Li Expires 26 February 2023 [Page 15]
Internet-Draft Deterministic Routing August 2022
2 the on-time policy
3 reserved
Q: 2 octets. Indicates the deadline scheduling delay Q in units
of microseconds. Note that if the U flag is 1, the value of Q
MUST be transmitted as 0 and ignored upon receipt.
If a FAD specifies the Deterministic Delay Metric as its Metric type
and FAD does not contain either the FAD CQF Sub-Sub-TLV or the FAD
Deadline Sub-Sub-TLV, then all nodes participating in the FAD should
behave as if the Min Unidirectional Link Delay Metric Type had been
advertised instead.
A FAD MUST NOT contain both the FAD CQF Sub-Sub-TLV and the FAD
Deadline Sub-Sub-TLV. If both do appear in appear in a FAD at the
same time, the first one is used, and subsequent ones are ignored.
7.2.2. OSPF Advertisement of Flex-algo Binding Deadline
To be defined in next version.
7.3. CQF Deterministic Route Computation
This document uses the Deterministic Delay Metric Type and the FAD
CQF Sub-Sub-TLV to signal that participating nodes should compute CQF
paths with the minimum deterministic end-to-end delay, which includes
the accumulated node delay provided by CQF and the accumulated link
propagation delay.
For a FAD that is using a specific CQF cycle_size, the delay metric
of a candidate path equals H * N + L, where H is the number of hops,
N is the node delay, and L is the accumulated link propagation delay.
N can be computed from the cycle_size and the intra-node forwarding
delay as described in Section 4.1.
The best path from the source node to the destination node is the
candidate path with minimum deterministic delay. This result may be
different from the traditional minimum delay paths that result from
considering only link propagation delay and ignoring node delay.
The deterministic delay jitter of a candidate path computed with this
technique should be the node delay jitter, which is 2*cycle_size,
plus the accumulated link delay jitter, which is effectively 0.
Peng & Li Expires 26 February 2023 [Page 16]
Internet-Draft Deterministic Routing August 2022
7.4. Deadline Deterministic Route Computation
This document uses the Deterministic Delay Metric Type and the FAD
Deadline Sub-Sub-TLV to compute deadline based shortest paths with
the minimum deterministic end-to-end delay, which contains the
accumulated node delay provided by deadline scheduling and the
accumulated link propagation delay.
For a FAD using deadline scheduling, the delay metric of a candidate
path equals H * N + L, where H is the number of hops, N is the node
delay, and L is the accumulated link propagation delay. N can be
computed based on the scheduling delay, scheduling policy and
forwarding delay intra node as described in Section 4.2.
Assuming that the scheduling delay Q and scheduling policy P are
obtained from the FAD Deadline Sub-Sub-TLV (treating Q as 0 if the U
bit is set), the node delay contributed by any intermediate node i
is:
* For the in-time policy, the node delay is in the range of [F(i),
F(i)+Q], where F(i) represents the intra-node forwarding delay i.
Because the node delay value in this case is a range, and we need
to have a specific value for SPF computation, there are several
options, i.e., we could select F(i), F(i)+Q, or the average of
F(i) and F(i)+Q. This document establishes F(i)+Q as the
conventional value that all nodes MUST use.
* For the on-time policy, the node delay is equal to F(i)+Q.
It should be noted that the above calculation process selects the
optimal deterministic delay path from multiple candidate paths.
However, once the deterministic delay SPF path is obtained, the
computed delay should closely reflect the actual delay. Especially:
* When the scheduling delay Q is an unknown value, the resulting
deterministic delay metric is a formula containing the variable
quantity Q. In this case, the value of scheduling delay Q needs
to be learned through other methods, such as timestamps carried in
data packets. This allows the same path to provide different
delays for different services.
* For the in-time policy, the minimum delay of the SPF path is H*F,
and the maximum delay is H*(F+Q), so that delay jitter is H*Q.
The deterministic delay jitter of a candidate path equals the
accumulated node delay jitter, which is H*Q for the in-time policy
and 0 for the on-time policy, plus the accumulated link delay jitter,
which is effectively 0.
Peng & Li Expires 26 February 2023 [Page 17]
Internet-Draft Deterministic Routing August 2022
8. Routing Convergence and Redundancy Considerations
As described in [I-D.ietf-lsr-flex-algo], Loop Free Alternate (LFA)
paths for a given Flexible Algorithm MUST be computed using the same
constraints as the calculation of the primary paths for that Flexible
Algorithm. The Segment Routing framework,
[I-D.ietf-rtgwg-segment-routing-ti-lfa], can provide TI-LFA paths, as
the expected post-convergence paths from the point of local repair in
any biconnected network using a link-state IGP. However, the
ordinary IGP convergence and FRR protection results may not meet the
needs of deterministic services. The main reasons include:
* IGP convergence may cause a considerable packet loss rate, even if
FRR switching is implemented on the basis of rapid fault
detection.
* The cumulative deterministic delay of the LFA path may be very
different from that of the primary path, which may not meet the
strict requirements for delay jitter.
Thus, according to the Service Protection function defined in
[RFC8655], packets can be spread over multiple disjoint forwarding
paths to mitigate or eliminate the packet loss rate. In the context
of Flexible Algorithm, an additional redundant deterministic delay
path different from the FRR path needs to be created, if the PLR
enables the Packet Replication Function (PRF) and the destination
enables the Packet Elimination Function (PEF). In this case, the
data packets are sent along the primary deterministic delay SPF path
and the redundant deterministic delay path at the same time, with
almost the same cumulative delay.
The additional redundant deterministic delay path within is often a
traffic engineering path that is calculated by the PLR based on the
constraints contained in the FAD and the following constraints:
* the number of nodes intersecting the primary and redundant
deterministic delay paths shall be minimized,
* the difference between the number of hops of the primary and
redundant deterministic delay paths shall be minimized, and
* the difference between the cumulative link propagation delay of
the primary and redundant deterministic delay paths shall be
minimized.
Unlike a LFA FRR path, more scheduling parameters learned from the
link-state database can be used in the redundant deterministic delay
path to obtain the delay equal or close to the primary path.
Peng & Li Expires 26 February 2023 [Page 18]
Internet-Draft Deterministic Routing August 2022
Take a deadline based path as an example. Suppose that the number of
hops in the primary path is m, and the intermediate nodes on the path
are A1, A2, ..., Am, the intra-node forwarding delay for each hop is
Fa, the intra-node scheduling delay for each hop is Qa, and the
cumulative link propagation delay is La. Then the cumulative delay
of the path is given by:
* Delay(primary) = m*Fa + m*Qa + La
Similarly, suppose that the number of hops in the redundant
deterministic delay path is n, the intermediate nodes are B1, B2,
..., Bn, the intra-node forwarding delay for each hop is Fb, the
intra-node scheduling delay for each hop is Qb, and the cumulative
link propagation delay is Lb. Then the cumulative delay of the
redundant path is given by:
* Delay(redundant) = n*Fb + n*Qb + Lb
The value of Delay(primary) can be calculated based on the known
value of Qa taken from the FAD. Then, an appropriate Qb is selected
to make Delay(redundant) equal to Delay(primary). Qb is likely to be
different from Qa, and SHOULD be carried in the packets sent along
the redundant deterministic delay path to get the expected latency.
If the value of Qa in the FAD is unknown, per-service state should be
maintained at the ingress node to determine the specific value of Qa
according to the SLA of the services sent along the primary
deterministic delay SPF path. Based on this, Qb can be calculated.
In this case, both Qa and Qb SHOULD be carried in the packets to get
the expected latency.
If the Packet Replication Function is implemented on an intermediate
node of the network, the node may regard itself as the head node of a
new protection sub-domain, and can still adopt the above scheme. The
intermediate node can also compute the value of Delay(primary) based
on the advertised Qa (or this can be obtained by observing deadlines
in packets). Based on this, Qb can be calculated.
It should be noted that both packets sent along the primary
deterministic delay SPF path and the redundant deterministic delay
path MUST use SIDs or prefixes related to the Flexible Algorithm.
The RIB entries for the Flexible Algorithm, such as SID entries,
contain specific deterministic scheduling parameters to enable the
packet to excecute the corresponding scheduling function. However,
if the packet also carries scheduling parameters, the ones in the
packet must be preferred.
Peng & Li Expires 26 February 2023 [Page 19]
Internet-Draft Deterministic Routing August 2022
9. Examples of Deterministic Delay SPF
As shown in Figure 7, a topology contains five nodes with five
bidirectional links. The figure shows the propagation delay of each
link, e.g, the propagation delay of the link between node R1 and node
R2 is 10us.
20us 30us
+----- R3 -----+
| |
10us | |
R1 ----- R2 R5
| |
| 10us 20us |
+----- R4 -----+
Figure 7
Next, we will demonstrate how the node delay affects the calculation
results of the SPF path. It should be noted that when the link
propagation delay is much greater than the node delay, the SPF
calculation result at this time is actually similar to metric type
"Min Unidirectional Link Delay", but it can still provide a certain
delay jitter.
9.1. CQF Deterministic Delay SPF
All links are configured with consistent CQF scheduling parameters
and have the node delay and delay jitter attributes:
Intra-node forwarding delay = 0us
CQF enable/disable = ON
Supported cycle_size set = <10 us>
Node R1 starts with itself as the root and calculates the
deterministic delay SPT shown in Figure 8. In the figure, the sum of
the node delay and the link transmssion delay is shown on each link.
For example, the delay of link from node R2 to R3 is 10+20, where 10
is the node delay and 20 is the link propagation delay.
Peng & Li Expires 26 February 2023 [Page 20]
Internet-Draft Deterministic Routing August 2022
R1
10+10 /
R2
10+20 / \ 10+10
R3 R4
\ 10+20
R5
Figure 8
The resulting path from R1 to R5 takes the path (R1-R2-R4-R5) with a
delay of 70us, and jitter of 20us.
Assuming that node R5 advertised a SID, the following RIB entry will
be created on node R1.
next_hop = R2
interface = link(R1-R2)
metric_type = Deterministic Delay
scheduling algorithm = CQF with cycle_size 10 us
total_metric = 70 us
total_metric_variation = 20 us
9.2. Deadline Deterministic Delay SPF
For this example, we will use the following deadline scheduling
paramters:
Intra-node forwarding delay = 5us
Deadline enable/disable = ON
Supported scheduling delay set = <10us, 20us, 30us, 40us, 50us,
60us>, and each item in the set supports both the in-time and the
on-time policy
The FAD for the Flexible Algorithm specifies deadline scheduling with
the in-time policy and a scheduling delay (Q) of 10us.
R1 starts with itself as the root and calculates the SPT as shown in
Figure 9. In the figure, the sum of the node delay and the link
transmssion delay is shown on each link. Note that this document
Peng & Li Expires 26 February 2023 [Page 21]
Internet-Draft Deterministic Routing August 2022
suggests taking F+Q as the node delay during calculations even for
the in-time policy. For example, the delay of link from node R2 to
R3 is 15+20, where 15 is node delay (intra-node forwarding delay of
5us and queuing delay of 10us) and 20 is link propagation delay.
R1
15+10 /
R2
15+20 / \ 15+10
R3 R4
\ 15+20
R5
Figure 9
The path from R1 to R5 has a cumulative delay of 85us, and the
cumulative jitter is 30us.
Assuming that node R5 advertised a SID, the following RIB entry will
be created on node R1.
next_hop = R2
interface = link(R1-R2)
metric_type = Deterministic Delay
scheduling algorithm = Deadline with Q=10 us with the in-time
policy
total_metric = 85 us
total_metric_variation = 30 us
Similarly, if the FAD specifies the on-time policy, the cumulative
deterministic delay would be is 85us, but the cumulative delay jitter
is 0. The RIB entry would instead be:
next_hop = R2
interface = link(R1-R2)
metric_type = Deterministic Delay
scheduling algorithm = Deadline with Q=10 us with the on-time
policy
Peng & Li Expires 26 February 2023 [Page 22]
Internet-Draft Deterministic Routing August 2022
total_metric = 85 us
total_metric_variation = 0
10. Use Cases
[RFC8578] described various deterministic routing use cases from
multiple industries, including: Pro Audio and Video, Electrical
Utilities, Building Automation Systems (BAS), Wireless for Industrial
Applications, Cellular Radio, Industrial Machine to Machine (M2M),
Mining Industry, Private Blockchain, Network Slicing, etc. Among
them, some industries are now transitioning to packet based
infrastructure, and some industries have already linked their
different subsystems through networks (intra-domain or inter-domain).
These industries have put forward their requirements for delay and
jitter bounds. For example, BAS requires low delay (10ms ~ 100ms)
and low jitter (1ms), M2M requires that the underlying network
infrastructure must ensure that the maximum end-to-end delay is
between 100 us and 50 ms, and the mining industry requires a
predictable time delay to implement real-time monitoring.
Another application lies in the Financial sector, specifically in
High Frequency Trading (HFT). This sector uses traffic engineering
today to direct trading traffic and best effort traffic over non-
overlapping paths. This can also be addressed through the use of
Flexible Algorithm, but in either case, the cost of the redundant
topologies is high and in neither case does it account for the
queueing delays that could happen.
Applying deterministic routing to HFT would allow a network to
provide bounded delays and jitter for trading traffic and
signficantly reduce costs by reducing the amount of topology that
needed to be reserved for the trading traffic. If a network chooses
to retain a separate topology for trading traffic, then deterministic
routing could be used to allow the best effort topology as a backup
path for trading traffic. In this scheme, the best effort topology
would be a proper subset of the trading topology, operating with two
independent FADs. The primary trading topology links would not
appear in the best effort topology, so the trading traffic would be
protected if there were no failures. However, in the event of a
failure, deterministic routing could then route around the failure
using links shared with the best effort topology. The queueing on
the links could still provide deterministic bounds, even in the face
of congestion.
The mechanism introduced in this document can compute a SPF path with
determinstic delay, but more importantly, with deterministic jitter.
The magnitude of the path delay actually depends on the network
Peng & Li Expires 26 February 2023 [Page 23]
Internet-Draft Deterministic Routing August 2022
scale. It can be large or small, but it can be guaranteed to be the
smallest of all candidate paths. The jitter is also bounded and may
be a cumulative value related to the number of hops or a value
independent of the number of hops. SPF Paths with such
characteristics will benefit multiple applications as mentioned
above.
11. IANA Considerations
11.1. IS-IS Deterministic Delay Metric Sub-TLV
This document requests that IANA allocate a code point, TBA1, from
the "IS-IS Sub-TLVs for TLVs Advertising Neighbor Information"
registry. The registry entry should appear as follows:
+------+--------------------+----+----+----+-----+-----+-----+
| Type | Description | 22 | 23 | 25 | 141 | 222 | 223 |
+======+====================+====+====+====+=====+=====+=====+
| | Deterministic Delay| | | | | | |
| TBA1 | Metric | y | y | y | y | y | y |
+------+--------------------+----+----+----+-----+-----+-----+
Figure 10
The reference for this entry should be this document.
11.2. Registry for Sub-Sub-TLVs for the Deterministic Delay Metric Sub-
TLV
This document requests that IANA create a new registry, "IS-IS Sub-
Sub-TLVs for the IS-IS Deterministic Delay Metric Sub-TLV". The
registration procedure for this registry is Expert Review. The
initial experts are Shaofu Peng and Tony Li. The description should
read "Sub-Sub-TLVs for the Deterministic Delay Metric Sub-TLV". The
initial population for the registry should be as follows:
+------+----------------------+---------------------------+
| Type | Description | Reference |
+======+======================+===========================+
| | CQF Intra-Node | This document Section 4.1 |
| 1 | Scheduling Delay | |
+------+----------------------+---------------------------+
| | Deadline Intra-Node | This document Section 4.2 |
| 2 | Scheduling Delay | |
+------+----------------------+---------------------------+
Figure 11
Peng & Li Expires 26 February 2023 [Page 24]
Internet-Draft Deterministic Routing August 2022
11.3. IGP Metric-Type Registration
This document requests that IANA allocate a code point for the
"Deterministic Delay Metric" in the forthcoming "IGP Metric-Type
Registry":
Type: TBA4 (suggested 4)
Description: Deterministic Delay Metric as defined in this
document
Reference: This document (Section 4)
11.4. IS-IS Sub-Sub-TLVs for Flexible Algorithm Definition Sub-TLV
This document allocate two code points for the following Sub-Sub-TLVs
in the forthcoming "Sub-Sub-TLVs for Flexible Algorithm Definition
Sub-TLV" registry:
+------+----------------------+-----------------------------+
| Type | Description | Reference |
+======+======================+=============================+
| TBA5 | FAD CQF | This document Section 7.1.1 |
+------+----------------------+-----------------------------+
| TBA6 | FAD Deadline | This document Section 7.2.1 |
+------+----------------------+-----------------------------+
Figure 12
11.5. OSPF IANA considerations
TBD.
12. Security Considerations
This document introduces no new security issues. Security of routing
within a domain is already addressed as part of the routing protocols
themselves. This document proposes no changes to those security
architectures.
The authentication methods described in [RFC5304] and [RFC5310] for
IS-IS, [RFC2328] and [RFC7474] for OSPFv2 and [RFC5340] and [RFC4552]
for OSPFv3 SHOULD be used to prevent attacks on the IGPs.
13. Acknowledgements
The authors would like to acknowledge the review and inputs from
Peter Psenak, Bin Tan, Quan Xiong.
Peng & Li Expires 26 February 2023 [Page 25]
Internet-Draft Deterministic Routing August 2022
14. References
14.1. Normative References
[I-D.ietf-lsr-flex-algo]
Psenak, P., Hegde, S., Filsfils, C., Talaulikar, K., and
A. Gulko, "IGP Flexible Algorithm", Work in Progress,
Internet-Draft, draft-ietf-lsr-flex-algo-20, 18 May 2022,
<https://www.ietf.org/archive/id/draft-ietf-lsr-flex-algo-
20.txt>.
[I-D.ietf-lsr-ip-flexalgo]
Britto, W., Hegde, S., Kaneriya, P., Shetty, R., Bonica,
R., and P. Psenak, "IGP Flexible Algorithms (Flex-
Algorithm) In IP Networks", Work in Progress, Internet-
Draft, draft-ietf-lsr-ip-flexalgo-06, 16 May 2022,
<https://datatracker.ietf.org/api/v1/doc/document/draft-
ietf-lsr-ip-flexalgo/>.
[I-D.ietf-rtgwg-segment-routing-ti-lfa]
Litkowski, S., Bashandy, A., Filsfils, C., Francois, P.,
Decraene, B., and D. Voyer, "Topology Independent Fast
Reroute using Segment Routing", Work in Progress,
Internet-Draft, draft-ietf-rtgwg-segment-routing-ti-lfa-
08, 21 January 2022,
<https://datatracker.ietf.org/api/v1/doc/document/draft-
ietf-rtgwg-segment-routing-ti-lfa/>.
[I-D.peng-detnet-deadline-based-forwarding]
Peng, S., Tan, B., and P. Liu, "Deadline Based
Deterministic Forwarding", Work in Progress, Internet-
Draft, draft-peng-detnet-deadline-based-forwarding-02, 8
July 2022,
<https://datatracker.ietf.org/api/v1/doc/document/draft-
peng-detnet-deadline-based-forwarding/>.
[RFC2119] Bradner, S., "Key words for use in RFCs to Indicate
Requirement Levels", BCP 14, RFC 2119,
DOI 10.17487/RFC2119, March 1997,
<https://www.rfc-editor.org/info/rfc2119>.
[RFC2328] Moy, J., "OSPF Version 2", STD 54, RFC 2328,
DOI 10.17487/RFC2328, April 1998,
<https://www.rfc-editor.org/info/rfc2328>.
Peng & Li Expires 26 February 2023 [Page 26]
Internet-Draft Deterministic Routing August 2022
[RFC2475] Blake, S., Black, D., Carlson, M., Davies, E., Wang, Z.,
and W. Weiss, "An Architecture for Differentiated
Services", RFC 2475, DOI 10.17487/RFC2475, December 1998,
<https://www.rfc-editor.org/info/rfc2475>.
[RFC4552] Gupta, M. and N. Melam, "Authentication/Confidentiality
for OSPFv3", RFC 4552, DOI 10.17487/RFC4552, June 2006,
<https://www.rfc-editor.org/info/rfc4552>.
[RFC5304] Li, T. and R. Atkinson, "IS-IS Cryptographic
Authentication", RFC 5304, DOI 10.17487/RFC5304, October
2008, <https://www.rfc-editor.org/info/rfc5304>.
[RFC5310] Bhatia, M., Manral, V., Li, T., Atkinson, R., White, R.,
and M. Fanto, "IS-IS Generic Cryptographic
Authentication", RFC 5310, DOI 10.17487/RFC5310, February
2009, <https://www.rfc-editor.org/info/rfc5310>.
[RFC5340] Coltun, R., Ferguson, D., Moy, J., and A. Lindem, "OSPF
for IPv6", RFC 5340, DOI 10.17487/RFC5340, July 2008,
<https://www.rfc-editor.org/info/rfc5340>.
[RFC7474] Bhatia, M., Hartman, S., Zhang, D., and A. Lindem, Ed.,
"Security Extension for OSPFv2 When Using Manual Key
Management", RFC 7474, DOI 10.17487/RFC7474, April 2015,
<https://www.rfc-editor.org/info/rfc7474>.
[RFC8174] Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC
2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174,
May 2017, <https://www.rfc-editor.org/info/rfc8174>.
[RFC8402] Filsfils, C., Ed., Previdi, S., Ed., Ginsberg, L.,
Decraene, B., Litkowski, S., and R. Shakir, "Segment
Routing Architecture", RFC 8402, DOI 10.17487/RFC8402,
July 2018, <https://www.rfc-editor.org/info/rfc8402>.
[RFC8570] Ginsberg, L., Ed., Previdi, S., Ed., Giacalone, S., Ward,
D., Drake, J., and Q. Wu, "IS-IS Traffic Engineering (TE)
Metric Extensions", RFC 8570, DOI 10.17487/RFC8570, March
2019, <https://www.rfc-editor.org/info/rfc8570>.
[RFC8578] Grossman, E., Ed., "Deterministic Networking Use Cases",
RFC 8578, DOI 10.17487/RFC8578, May 2019,
<https://www.rfc-editor.org/info/rfc8578>.
Peng & Li Expires 26 February 2023 [Page 27]
Internet-Draft Deterministic Routing August 2022
[RFC8655] Finn, N., Thubert, P., Varga, B., and J. Farkas,
"Deterministic Networking Architecture", RFC 8655,
DOI 10.17487/RFC8655, October 2019,
<https://www.rfc-editor.org/info/rfc8655>.
14.2. Informative References
[CBS] "IEEE802.1Qav", 2009,
<https://ieeexplore.ieee.org/document/8684664>.
[CQF] "IEEE802.1Qch", 2017,
<https://ieeexplore.ieee.org/document/7961303>.
[SP-LATENCY]
"Guaranteed Latency with SP", 2020,
<https://www.ieee802.org/1/files/public/docs2020/dd-
grigorjew-strict-priority-latency-0320-v02.pdf>.
Authors' Addresses
Shaofu Peng
ZTE
Nanjing
China
Email: peng.shaofu@zte.com.cn
Tony Li
Juniper Networks
1133 Innovation Way
Sunnyvale, California 94089
United States of America
Email: tony.li@tony.li
Peng & Li Expires 26 February 2023 [Page 28]
