draft-ietf-mpls-spring-entropy-label-05.txt   draft-ietf-mpls-spring-entropy-label-06.txt 
Network Working Group S. Kini Network Working Group S. Kini
Internet-Draft Internet-Draft
Intended status: Informational K. Kompella Intended status: Informational K. Kompella
Expires: October 26, 2017 Juniper Expires: November 4, 2017 Juniper
S. Sivabalan S. Sivabalan
Cisco Cisco
S. Litkowski S. Litkowski
Orange Orange
R. Shakir R. Shakir
Google Google
J. Tantsura J. Tantsura
April 24, 2017 May 3, 2017
Entropy label for SPRING tunnels Entropy label for SPRING tunnels
draft-ietf-mpls-spring-entropy-label-05 draft-ietf-mpls-spring-entropy-label-06
Abstract Abstract
Source routed tunnels with label stacking is a technique that can be Source routed tunnels with label stacking is a technique that can be
leveraged to provide a method to steer a packet through a controlled leveraged to steer a packet through a controlled set of segments.
set of segments. This can be applied to the Multi Protocol Label This can be applied to the Multi Protocol Label Switching (MPLS) data
Switching (MPLS) data plane. Entropy label (EL) is a technique used plane. Entropy label (EL) is a technique used in MPLS to improve
in MPLS to improve load balancing. This document examines and load-balancing. This document examines and describes how ELs are to
describes how ELs are to be applied to source routed tunnels with be applied to source routed tunnels with label stacks.
label stacks.
Status of This Memo Status of This Memo
This Internet-Draft is submitted in full conformance with the This Internet-Draft is submitted in full conformance with the
provisions of BCP 78 and BCP 79. provisions of BCP 78 and BCP 79.
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 http://datatracker.ietf.org/drafts/current/. Drafts is at http://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 October 26, 2017. This Internet-Draft will expire on November 4, 2017.
Copyright Notice Copyright Notice
Copyright (c) 2017 IETF Trust and the persons identified as the Copyright (c) 2017 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
(http://trustee.ietf.org/license-info) in effect on the date of (http://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
carefully, as they describe your rights and restrictions with respect carefully, as they describe your rights and restrictions with respect
to this document. Code Components extracted from this document must to this document. Code Components extracted from this document must
include Simplified BSD License text as described in Section 4.e of include Simplified BSD License text as described in Section 4.e of
the Trust Legal Provisions and are provided without warranty as the Trust Legal Provisions and are provided without warranty as
described in the Simplified BSD License. described in the Simplified BSD License.
Table of Contents Table of Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3
1.1. Requirements Language . . . . . . . . . . . . . . . . . . 3 1.1. Requirements Language . . . . . . . . . . . . . . . . . . 3
2. Abbreviations and Terminology . . . . . . . . . . . . . . . . 4 2. Abbreviations and Terminology . . . . . . . . . . . . . . . . 3
3. Use-case requiring multipath load balancing . . . . . . . . . 4 3. Use-case requiring multipath load-balancing . . . . . . . . . 4
4. Entropy Readable Label Depth . . . . . . . . . . . . . . . . 5 4. Entropy Readable Label Depth . . . . . . . . . . . . . . . . 5
5. Maximum SID Depth . . . . . . . . . . . . . . . . . . . . . . 7 5. Maximum SID Depth . . . . . . . . . . . . . . . . . . . . . . 7
6. LSP stitching using the binding SID . . . . . . . . . . . . . 8 6. LSP stitching using the binding SID . . . . . . . . . . . . . 8
7. Insertion of entropy labels for SPRING path . . . . . . . . . 10 7. Insertion of entropy labels for SPRING path . . . . . . . . . 10
7.1. Overview . . . . . . . . . . . . . . . . . . . . . . . . 10 7.1. Overview . . . . . . . . . . . . . . . . . . . . . . . . 10
7.1.1. Example 1 . . . . . . . . . . . . . . . . . . . . . . 11 7.1.1. Example 1 . . . . . . . . . . . . . . . . . . . . . . 11
7.1.2. Example 2 . . . . . . . . . . . . . . . . . . . . . . 12 7.1.2. Example 2 . . . . . . . . . . . . . . . . . . . . . . 12
7.2. Considerations for the placement of entropy labels . . . 12 7.2. Considerations for the placement of entropy labels . . . 12
7.2.1. ERLD value . . . . . . . . . . . . . . . . . . . . . 13 7.2.1. ERLD value . . . . . . . . . . . . . . . . . . . . . 13
7.2.2. Segment type . . . . . . . . . . . . . . . . . . . . 14 7.2.2. Segment type . . . . . . . . . . . . . . . . . . . . 14
7.2.2.1. Node-SID . . . . . . . . . . . . . . . . . . . . 14 7.2.2.1. Node-SID . . . . . . . . . . . . . . . . . . . . 14
7.2.2.2. Adjacency-SID representing an ECMP bundle . . . . 14 7.2.2.2. Adjacency-SID representing an ECMP bundle . . . . 14
7.2.2.3. Adjacency-SID representing a single IP link . . . 15 7.2.2.3. Adjacency-SID representing a single IP link . . . 15
7.2.2.4. Adjacency-SID representing a single link within a 7.2.2.4. Adjacency-SID representing a single link within
L2 bundle . . . . . . . . . . . . . . . . . . . . 15 an L2 bundle . . . . . . . . . . . . . . . . . . 15
7.2.2.5. Adjacency-SID representing a L2 bundle . . . . . 15 7.2.2.5. Adjacency-SID representing an L2 bundle . . . . . 15
7.2.3. Maximizing number of LSRs that will loadbalance . . . 15 7.2.3. Maximizing number of LSRs that will load-balance . . 15
7.2.4. Preference for a part of the path . . . . . . . . . . 16 7.2.4. Preference for a part of the path . . . . . . . . . . 16
7.2.5. Combining criteria . . . . . . . . . . . . . . . . . 16 7.2.5. Combining criteria . . . . . . . . . . . . . . . . . 16
8. A simple algorithm example . . . . . . . . . . . . . . . . . 16 8. A simple algorithm example . . . . . . . . . . . . . . . . . 16
9. Deployment Considerations . . . . . . . . . . . . . . . . . . 17 9. Deployment Considerations . . . . . . . . . . . . . . . . . . 17
10. Options considered . . . . . . . . . . . . . . . . . . . . . 18 10. Options considered . . . . . . . . . . . . . . . . . . . . . 18
10.1. Single EL at the bottom of the stack of tunnels . . . . 18 10.1. Single EL at the bottom of the stack of tunnels . . . . 18
10.2. An EL per tunnel in the stack . . . . . . . . . . . . . 18 10.2. An EL per tunnel in the stack . . . . . . . . . . . . . 18
10.3. A re-usable EL for a stack of tunnels . . . . . . . . . 19 10.3. A re-usable EL for a stack of tunnels . . . . . . . . . 19
10.4. EL at top of stack . . . . . . . . . . . . . . . . . . . 20 10.4. EL at top of stack . . . . . . . . . . . . . . . . . . . 20
10.5. ELs at readable label stack depths . . . . . . . . . . . 20 10.5. ELs at readable label stack depths . . . . . . . . . . . 20
skipping to change at page 3, line 29 skipping to change at page 3, line 23
segments. This can be directly applied to the MPLS data plane, but segments. This can be directly applied to the MPLS data plane, but
it has implications on the label stack depth. it has implications on the label stack depth.
Clarifying statements on label stack depth have been provided in Clarifying statements on label stack depth have been provided in
[RFC7325] but the RFC does not address the case of source routed [RFC7325] but the RFC does not address the case of source routed
stacked MPLS tunnels as described in stacked MPLS tunnels as described in
[I-D.ietf-spring-segment-routing] where deeper label stacks are more [I-D.ietf-spring-segment-routing] where deeper label stacks are more
prevalent. prevalent.
Entropy label (EL) [RFC6790] is a technique used in the MPLS data Entropy label (EL) [RFC6790] is a technique used in the MPLS data
plane to provide entropy for load balancing. When using LSP plane to provide entropy for load-balancing. When using LSP
hierarchies there are implications on how [RFC6790] should be hierarchies, there are implications on how [RFC6790] should be
applied. The current document addresses the case where the hierarchy applied. The current document addresses the case where the hierarchy
is created at a single LSR as required by source routed tunnels with is created at a single LSR as required by source routed tunnels with
label stacks. label stacks.
A use-case requiring load balancing with source routed tunnels with A use-case requiring load-balancing with source routed tunnels with
label stacks is given in Section 3. A recommended solution is label stacks is given in Section 3. A recommended solution is
described in Section 7 keeping in consideration the limitations of described in Section 7 keeping in consideration the limitations of
implementations when applying [RFC6790] to deeper label stacks. implementations when applying [RFC6790] to deeper label stacks.
Options that were considered to arrive at the recommended solution Options that were considered to arrive at the recommended solution
are documented for historical purposes in Section 10. are documented for historical purposes in Section 10.
1.1. Requirements Language 1.1. Requirements Language
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
"SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
skipping to change at page 4, line 10 skipping to change at page 4, line 4
Although this document is not a protocol specification, the use of Although this document is not a protocol specification, the use of
this language clarifies the instructions to protocol designers this language clarifies the instructions to protocol designers
producing solutions that satisfy the requirements set out in this producing solutions that satisfy the requirements set out in this
document. document.
2. Abbreviations and Terminology 2. Abbreviations and Terminology
EL - Entropy Label EL - Entropy Label
ELI - Entropy Label Identifier ELI - Entropy Label Identifier
ELC - Entropy Label Capability ELC - Entropy Label Capability
ERLD - Entropy Readable Label Depth ERLD - Entropy Readable Label Depth
SR - Segment Routing SR - Segment Routing
ECMP - Equal Cost Multi Paths ECMP - Equal Cost Multi Path
LSR - Label Switch Router LSR - Label Switch Router
MPLS - Multiprotocol Label Switching MPLS - Multiprotocol Label Switching
MSD - Maximum SID Depth MSD - Maximum SID Depth
SID - Segment Identifier SID - Segment Identifier
RLD - Readable Label Depth RLD - Readable Label Depth
OAM - Operation, Administration and Maintenance OAM - Operation, Administration and Maintenance
3. Use-case requiring multipath load balancing 3. Use-case requiring multipath load-balancing
+------+ +------+
| | | |
+-------| P3 |-----+ +-------| P3 |-----+
| +-----| |---+ | | +-----| |---+ |
L3| |L4 +------+ L1| |L2 +----+ L3| |L4 +------+ L1| |L2 +----+
| | | | +--| P4 |--+ | | | | +--| P4 |--+
+-----+ +-----+ +-----+ | +----+ | +-----+ +-----+ +-----+ +-----+ | +----+ | +-----+
| S |-----| P1 |------------| P2 |--+ +--| D | | S |-----| P1 |------------| P2 |--+ +--| D |
| | | | | |--+ +--| | | | | | | |--+ +--| |
skipping to change at page 5, line 4 skipping to change at page 4, line 45
+-----+ +-----+ +-----+ | +----+ | +-----+ +-----+ +-----+ +-----+ | +----+ | +-----+
+--| P5 |--+ +--| P5 |--+
+----+ +----+
S=Source LSR, D=Destination LSR, P1,P2,P3,P4,P5=Transit LSRs, S=Source LSR, D=Destination LSR, P1,P2,P3,P4,P5=Transit LSRs,
L1,L2,L3,L4=Links L1,L2,L3,L4=Links
Figure 1: Traffic engineering use-case Figure 1: Traffic engineering use-case
Traffic-engineering (TE) is one of the applications of MPLS and is Traffic-engineering (TE) is one of the applications of MPLS and is
also a requirement for source routed tunnels with label stacks also a requirement for source routed tunnels with label stacks
[RFC7855]. Consider the topology shown in Figure 1. The LSR S [RFC7855]. Consider the topology shown in Figure 1. The LSR S
requires data to be sent to LSR D along a traffic- engineered path requires data to be sent to LSR D along a traffic-engineered path
that goes over the link L1. Good load balancing is also required that goes over the link L1. Good load-balancing is also required
across equal cost paths (including parallel links). To engineer across equal cost paths (including parallel links). To engineer
traffic along a path that takes link L1, the label stack that LSR S traffic along a path that takes link L1, the label stack that LSR S
creates consists of a label to the node SID of LSR P3, stacked over creates consists of a label to the node SID of LSR P3, stacked over
the label for the adjacency SID of link L1 and that in turn is the label for the adjacency SID of link L1 and that in turn is
stacked over the label to the node SID of LSR D. For simplicity lets stacked over the label to the node SID of LSR D. For simplicity lets
assume that all LSRs use the same label space (SRGB) for source assume that all LSRs use the same label space (SRGB) for source
routed label stacks. Let L_N-Px denote the label to be used to reach routed label stacks. Let L_N-Px denote the label to be used to reach
the node SID of LSR Px. Let L_A-Ln denote the label used for the the node SID of LSR Px. Let L_A-Ln denote the label used for the
adjacency SID for link Ln. The LSR S must use the label stack <L_N- adjacency SID for link Ln. The LSR S must use the label stack <L_N-
P3, L_A-L1, L_N-D> for traffic-engineering. However to achieve good P3, L_A-L1, L_N-D> for traffic-engineering. However to achieve good
load balancing over the equal cost paths P2-P4-D, P2-P5-D and the load-balancing over the equal cost paths P2-P4-D, P2-P5-D and the
parallel links L3, L4, a mechanism such as Entropy labels [RFC6790] parallel links L3, L4, a mechanism such as Entropy labels [RFC6790]
should be adapted for source routed label stacks. Indeed, the SPRING should be adapted for source routed label stacks. Indeed, the SPRING
architecture with the MPLS dataplane uses nested MPLS LSPs composing architecture with the MPLS dataplane uses nested MPLS LSPs composing
the source routed label stacks. As each MPLS node may have the source routed label stacks. As each MPLS node may have
limitations in the number of labels it can push when it is ingress or limitations in the number of labels it can push when it is ingress or
inspect when doing loadbalancing, entropy labels insertion strategy inspect when doing load-balancing, an entropy label insertion
becomes important to keep benefit of the loadbalancing. Multiple strategy becomes important to keep the benefit of the load-balancing.
ways to apply entropy labels were considered and are documented in Multiple ways to apply entropy labels were considered and are
Section 10 along with their tradeoffs. A recommended solution is documented in Section 10 along with their trade-offs. A recommended
described in Section 7. solution is described in Section 7.
4. Entropy Readable Label Depth 4. Entropy Readable Label Depth
The Entropy Readable Label Depth (ERLD) is defined as the number of The Entropy Readable Label Depth (ERLD) is defined as the number of
labels a router can: labels a router can both:
a. read in an MPLS packet received on its incoming interface a. Read in an MPLS packet received on its incoming interface(s)
(starting from the top of the stack) and (starting from the top of the stack).
b. use in its loadbalancing function. b. Use in its load-balancing function.
The ERLD means that the router will perform load-balancing using the The ERLD means that the router will perform load-balancing using the
EL label if the EL is placed within the ERLD first labels. EL label if the EL is placed within the ERLD first labels.
A router capable of reading N labels but not using an EL located A router capable of reading N labels but not using an EL located
within those N labels MUST consider its ERLD to be 0. In a within those N labels MUST consider its ERLD to be 0. In a
distributed switching architecture, each linecard may have a distributed switching architecture, each linecard may have a
different capability in term of ERLD. For simplicity reason, an different capability in terms of ERLD. For simplicity, an
implementation MAY use the minimum ERLD between each linecard as the implementation MAY use the minimum ERLD between each linecard as the
ERLD value for the system. ERLD value for the system.
Examples: Examples:
| Payload | | Payload |
+----------+ +----------+
| Payload | | EL | P7 | Payload | | EL | P7
+----------+ +----------+ +----------+ +----------+
| Payload | | EL | | ELI | | Payload | | EL | | ELI |
skipping to change at page 6, line 29 skipping to change at page 6, line 29
| Label 16 | | Label 16 | | Label 16 | | Label 16 | | Label 16 | P1 | Label 16 | | Label 16 | | Label 16 | | Label 16 | | Label 16 | P1
+----------+ +----------+ +----------+ +----------+ +----------+ +----------+ +----------+ +----------+ +----------+ +----------+
Packet 1 Packet 2 Packet 3 Packet 4 Packet 5 Packet 1 Packet 2 Packet 3 Packet 4 Packet 5
Figure 2: Label stacks with ELI/EL Figure 2: Label stacks with ELI/EL
In the figure below, we consider the displayed packets received on a In the figure below, we consider the displayed packets received on a
router interface. We consider also a single ERLD value for the router interface. We consider also a single ERLD value for the
router. router.
o If the router has an ERLD of 3, it will be able to loadbalance the o If the router has an ERLD of 3, it will be able to load-balance
Packet 1 displayed in the Figure 2 using the EL as part of the Packet 1 displayed in Figure 2 using the EL as part of the load-
loadbalancing keys. The ERLD value of 3 means that the router can balancing keys. The ERLD value of 3 means that the router can
read and take into account the entropy label for loadbalancing if read and take into account the entropy label for load-balancing if
it is placed between position 1 (top) and position 3. it is placed between position 1 (top) and position 3.
o If the router has an ERLD of 5, it will be able to loadbalance the o If the router has an ERLD of 5, it will be able to load-balance
Packet 1 to 3 in the Figure 2 using the EL as part of the Packets 1 to 3 in Figure 2 using the EL as part of the load-
loadbalancing keys. The Packet 4 and 5 have the EL placed at a balancing keys. Packets 4 and 5 have the EL placed at a position
position greater than 5, so the router is not able to read it and greater than 5, so the router is not able to read it and use as
take it into account during the hashing. part of the load-balancing keys.
o If the router has an ERLD of 10, it will be able to loadbalance o If the router has an ERLD of 10, it will be able to load-balance
all the packets displayed in the Figure 2 using the EL as part of all the packets displayed in Figure 2 using the EL as part of the
the loadbalancing keys. load-balancing keys.
To allow an efficient loadbalancing based on entropy labels, a router To allow an efficient load-balancing based on entropy labels, a
running SPRING SHOULD advertise its ERLD (or ERLDs), so all the other router running SPRING SHOULD advertise its ERLD (or ERLDs), so all
SPRING routers in the network are aware of its capability. How this the other SPRING routers in the network are aware of its capability.
advertisement is done is out of scope of this document. How this advertisement is done is outside the scope of this document.
To advertise an ERLD value, a SPRING router: To advertise an ERLD value, a SPRING router:
o MUST be entropy label capable and as a consequence MUST apply all o MUST be entropy label capable and, as a consequence, MUST apply
the procedures defined in [RFC6790]. all the procedures defined in [RFC6790].
o MUST be able to read an ELI/EL which is located within its ERLD o MUST be able to read an ELI/EL which is located within its ERLD
value value.
o MUST take into account this EL in its load balancing function o MUST take into account this EL in its load-balancing function.
5. Maximum SID Depth 5. Maximum SID Depth
The Maximum SID Depth defines the maximum number of labels that a The Maximum SID Depth defines the maximum number of labels that a
particular node can impose on a packet. This includes any kind of particular node can impose on a packet. This includes any kind of
labels (service, entropy, transport...). In an MPLS network, the MSD labels (service, entropy, transport...). In an MPLS network, the MSD
is a limit of the Ingress LSR (I-LSR) or any stitching node that is a limit of the Ingress LSR (I-LSR) or any stitching node that
would perform an imposition of additional labels on an existing label would perform an imposition of additional labels on an existing label
stack. stack.
Depending of the number of MPLS operations (POP, SWAP...) to be Depending of the number of MPLS operations (POP, SWAP...) to be
performed before the PUSH, the MSD may vary due to the hardware or performed before the PUSH, the MSD may vary due to the hardware or
software limitations. As for the ERLD, there may also be different software limitations. As for the ERLD, there may also be different
MSD limits based on the linecard type used in a distributed switching MSD limits based on the linecard type used in a distributed switching
system. system.
When an external controller is used to program a label stack on a When an external controller is used to program a label stack on a
particular node, this node MAY advertise its MSD value or a subset of particular node, this node MAY advertise its MSD value or a subset of
its MSD value to the controller. How this advertisement is done is its MSD value to the controller. How this advertisement is done is
out of scope of this document. As the controller does not have the outside the scope of this document. As the controller does not have
knowledge of the entire label stack to be pushed by the node, the the knowledge of the entire label stack to be pushed by the node, the
node may advertise an MSD value which is lower than its real limit. node may advertise an MSD value which is lower than its actual limit.
This gives the ability for the controller to program a label stack up This gives the ability for the controller to program a label stack up
to the advertised MSD value while leaving room for the local node to to the advertised MSD value while leaving room for the local node to
add more labels (e.g. service, entropy, transport...) without add more labels (e.g., service, entropy, transport...) without
reaching the hardware/software limit. reaching the hardware/software limit.
P7 ---- P8 ---- P9 P7 ---- P8 ---- P9
/ \ / \
PE1 --- P1 --- P2 --- P3 --- P4 --- P5 --- P6 --- PE2 PE1 --- P1 --- P2 --- P3 --- P4 --- P5 --- P6 --- PE2
| \ | | \ |
----> P10 \ | ----> P10 \ |
IP Pkt | \ | IP Pkt | \ |
P11 --- P12 --- P13 P11 --- P12 --- P13
100 10000 100 10000
Figure 3 Figure 3
In the Figure 3, an IP packet comes in the MPLS network at PE1. All In the Figure 3, an IP packet comes in the MPLS network at PE1. All
metrics are considered equal to 1 except P12-P13 which is 10000 and metrics are considered equal to 1 except P12-P13 which is 10000 and
P11-P12 which is 100. PE1 wants to steer the traffic using a SPRING P11-P12 which is 100. PE1 wants to steer the traffic using a SPRING
path to PE2 along path to PE2 along
PE1->P1->P7->P8->P9->P4->P5->P10->P11->P12->P13->PE2. By using PE1->P1->P7->P8->P9->P4->P5->P10->P11->P12->P13->PE2. By using
Adjacency SIDs only, PE1 will be required to push (as an I-LSR) 10 Adjacency SIDs only, PE1 will be required to push (as an I-LSR) 10
labels on the IP packet received, it so requires an MSD of 10. If labels on the IP packet received and so requires an MSD of 10. If
the IP packet should be carried over an MPLS service like a regular the IP packet should be carried over an MPLS service like a regular
layer 3 VPN, an additional service label will be imposed, requiring layer 3 VPN, an additional service label will be imposed, requiring
an MSD of 11 for PE1. In addition, if PE1 wants to insert an ELI/EL an MSD of 11 for PE1. In addition, if PE1 wants to insert an ELI/EL
for loadbalancing purpose, PE1 will need to push 13 labels on the IP for load-balancing purpose, PE1 will need to push 13 labels on the IP
packet requiring an MSD of 13. packet requiring an MSD of 13.
In the SPRING architecture, Node SIDs or Binding SIDs can be used to In the SPRING architecture, Node SIDs or Binding SIDs can be used to
reduce the label stack size. As an example, to steer the traffic on reduce the label stack size. As an example, to steer the traffic on
the same path as before, PE1 may be able to use the following label the same path as before, PE1 may be able to use the following label
stack: <Node_P9, Node_P5, Binding_P5, Node_PE2>. In this example we stack: <Node_P9, Node_P5, Binding_P5, Node_PE2>. In this example we
consider a combination of Node SIDs and a Binding SID advertised by consider a combination of Node SIDs and a Binding SID advertised by
P5 that will stitch the traffic along the path P10->P11->P12->P13. P5 that will stitch the traffic along the path P10->P11->P12->P13.
The instruction associated with the binding SID at P5 is thus to swap The instruction associated with the binding SID at P5 is thus to swap
Binding_P5 to Adj_P12-P13 and then push <Adj_P11-P12, Node_P11>. P5 Binding_P5 to Adj_P12-P13 and then push <Adj_P11-P12, Node_P11>. P5
acts as a stitching node that pushes additional labels on an existing acts as a stitching node that pushes additional labels on an existing
label stack, P5 MSD needs also to be taken into account and may limit label stack, P5's MSD needs also to be taken into account and may
the number of labels that could be imposed. limit the number of labels that could be imposed.
6. LSP stitching using the binding SID 6. LSP stitching using the binding SID
The binding SID allows to bind a segment identifier to an existing The binding SID allows binding a segment identifier to an existing
LSP. As examples, the binding SID can represent an RSVP-TE tunnel, LSP. As examples, the binding SID can represent an RSVP-TE tunnel,
an LDP path (through the mapping server advertisement), a SPRING an LDP path (through the mapping server advertisement), a SPRING
path... Each LSP associated with a binding SID has its own entropy path... Each LSP associated with a binding SID has its own entropy
label capability. label capability.
In the figure 3, if we consider that: In the figure 3, if we consider that:
o P6, PE2, P10, P11, P12 are pure LDP routers. o P6, PE2, P10, P11, P12 are pure LDP routers.
o PE1, P1, P2, P3, P4, P7, P8, P9 are pure SPRING routers. o PE1, P1, P2, P3, P4, P7, P8, P9 are pure SPRING routers.
skipping to change at page 9, line 33 skipping to change at page 9, line 33
the tail-end of the LSP associated with the binding (PE2) is entropy the tail-end of the LSP associated with the binding (PE2) is entropy
label capable. label capable.
To accomodate the mix of signalling protocols involved during the To accomodate the mix of signalling protocols involved during the
stitching, the entropy label capability SHOULD be propagated between stitching, the entropy label capability SHOULD be propagated between
the signalling protocols. Each binding SID SHOULD have its own the signalling protocols. Each binding SID SHOULD have its own
entropy label capability that MUST be inherited from the entropy entropy label capability that MUST be inherited from the entropy
label capability of the associated LSP. If the router advertising label capability of the associated LSP. If the router advertising
the binding SID does not know the ELC state of the target FEC, it the binding SID does not know the ELC state of the target FEC, it
MUST NOT set the ELC for the binding SID. An ingress node MUST NOT MUST NOT set the ELC for the binding SID. An ingress node MUST NOT
push an ELI/EL associated to a binding SID unless this binding SID push an ELI/EL associated with a binding SID unless this binding SID
has the entropy label capability. How the entropy label capability has the entropy label capability. How the entropy label capability
is advertised for a binding SID is out of scope of this document. is advertised for a binding SID is outside the scope of this
document.
In our example, if PE2 is LDP entropy label capable, it will add the In our example, if PE2 is LDP entropy label capable, it will add the
entropy label capability in its LDP advertisement. When P5 receives entropy label capability in its LDP advertisement. When P5 receives
the FEC/label binding for PE2, it learns about the ELC and can set the FEC/label binding for PE2, it learns about the ELC and can set
the ELC in the mapping server advertisement. Thus PE1 learns about the ELC in the mapping server advertisement. Thus PE1 learns about
the ELC of PE2 and may push an ELI/EL associated with the binding the ELC of PE2 and may push an ELI/EL associated with the binding
SID. SID.
The proposed solution works only if the SPRING router advertising the The proposed solution only works if the SPRING router advertising the
binding SID is also performing the dataplane LSP stitching. In our binding SID is also performing the dataplane LSP stitching. In our
example, if the mapping server function is hosted on P8 instead of example, if the mapping server function is hosted on P8 instead of
P5, P8 does not know about the ELC state of PE2 LDP FEC. As a P5, P8 does not know about the ELC state of PE2's LDP FEC. As a
consequence, it does not set the ELC on the associated binding SID. consequence, it does not set the ELC for the associated binding SID.
7. Insertion of entropy labels for SPRING path 7. Insertion of entropy labels for SPRING path
7.1. Overview 7.1. Overview
The solution described in this section follows [RFC6790]. Within a The solution described in this section follows [RFC6790]. Within a
SPRING path, a node may be ingress, egress, transit (regarding the SPRING path, a node may be ingress, egress, transit (regarding the
entropy label processing described in [RFC6790] or it can be any entropy label processing described in [RFC6790]), or it can be any
combination of those. For example: combination of those. For example:
o The ingress node of a SPRING domain may be an ingress node from an o The ingress node of a SPRING domain may be an ingress node from an
entropy label perspective. entropy label perspective.
o Any LSR terminating a segment of the SPRING path is an egress node o Any LSR terminating a segment of the SPRING path is an egress node
(because it terminates the segment) but may also be a transit node (because it terminates the segment) but may also be a transit node
if the SPRING path is not ended here because there is a subsequent if the SPRING path is not terminated because there is a subsequent
SPRING MPLS label in the stack. SPRING MPLS label in the stack.
o Any LSR processing a binding SID may be a transit node and an o Any LSR processing a binding SID may be a transit node and an
ingress node (because it may push additional labels when ingress node (because it may push additional labels when
processing the binding SID). processing the binding SID).
As described earlier, an LSR may have a limitation on the depth of As described earlier, an LSR may have a limitation, ERLD, on the
the label stack that it can read and process in order to do multipath depth of the label stack that it can read and process in order to do
load balancing based on entropy labels: we called it the ERLD. multipath load-balancing based on entropy labels.
If an EL does not occur within the ERLD of an LSR in the label stack If an EL does not occur within the ERLD of an LSR in the label stack
of the MPLS packet that it receives, then it would lead to poor load of an MPLS packet that it receives, then it would lead to poor load-
balancing at that LSR. Hence an ELI/EL pair MUST be within the ERLD balancing at that LSR. Hence an ELI/EL pair MUST be within the ERLD
of the LSR in order for the LSR to use the EL during load balancing. of the LSR in order for the LSR to use the EL during load-balancing.
Adding a single ELI/EL pair for the entire SPRING path may lead also Adding a single ELI/EL pair for the entire SPRING path may lead also
to poor loadbalancing as well because the EL/ELI may not occur within to poor load-balancing as well because the EL/ELI may not occur
the ERLD of some LSR on the path (if too deep) or may not be present within the ERLD of some LSR on the path (if too deep) or may not be
anymore in the stack for some LSRs if too shallow. present in the stack when it reaches some LSRs if it is too shallow.
In order for the EL to occur within the ERLD of LSRs along the path In order for the EL to occur within the ERLD of LSRs along the path
corresponding to a SPRING label stack, multiple <ELI, EL> pairs MAY corresponding to a SPRING label stack, multiple <ELI, EL> pairs MAY
be inserted in this label stack. be inserted in this label stack.
The insertion of the ELI/EL SHOULD occur only with a SPRING label The insertion of the ELI/EL SHOULD occur only with a SPRING label
advertised by an LSR that advertised an ERLD (the LSR is entropy advertised by an LSR that advertised an ERLD (the LSR is entropy
label capable) or with a SPRING label associated with a binding SID label capable) or with a SPRING label associated with a binding SID
that has the ELC set. that has the ELC set.
The ELs among multiple <ELI, EL> pairs inserted in the stack MAY be The ELs among multiple <ELI, EL> pairs inserted in the stack MAY be
same or different. The LSR that inserts <ELI, EL> pairs MAY have the same or different. The LSR that inserts <ELI, EL> pairs MAY have
limitations on the number of such pairs that it can insert and also limitations on the number of such pairs that it can insert and also
the depth at which it can insert them. If due to any limitation, the the depth at which it can insert them. If due to limitations, the
inserted ELs are at positions such that an LSR along the path inserted ELs are at positions such that an LSR along the path
receives an MPLS packet without an EL in the label stack within that receives an MPLS packet without an EL in the label stack within that
LSR's ERLD, then the load balancing performed by that LSR would be LSR's ERLD, then the load-balancing performed by that LSR would be
poor. An implementation MAY consider multiple criterias when poor. An implementation MAY consider multiple criteria when
inserting <ELI, EL> pairs. inserting <ELI, EL> pairs.
7.1.1. Example 1 7.1.1. Example 1
ECMP LAG LAG ECMP LAG LAG
PE1 --- P1 --- P2 --- P3 --- P4 --- P5 --- P6 --- PE2 PE1 --- P1 --- P2 --- P3 --- P4 --- P5 --- P6 --- PE2
Figure 4 Figure 4
In the Figure 4, PE1 wants to forward some MPLS VPN traffic over an In the Figure 4, PE1 wants to forward some MPLS VPN traffic over an
explicit path to PE2 resulting in the following label stack to be explicit path to PE2 resulting in the following label stack to be
pushed onto the received IP header: {VPN_label, Adj_P6PE2, Adj_P5P6, pushed onto the received IP header: {VPN_label, Adj_P6PE2, Adj_P5P6,
Adj_P4P5, Adj_P3P4, Adj_Bundle_P2P3, Adj_P1P2}. PE1 is limited to Adj_P4P5, Adj_P3P4, Adj_Bundle_P2P3, Adj_P1P2}. PE1 is limited to
push a maximum of 11 labels (MSD=11). P2, P3 and P6 have an ERLD of push a maximum of 11 labels (MSD=11). P2, P3 and P6 have an ERLD of
3 while others have an ERLD of 10. 3 while others have an ERLD of 10.
PE1 can only add two ELI/EL pairs in the label stack due to its MSD PE1 can only add two ELI/EL pairs in the label stack due to its MSD
limitation. It should place them in a smart way to benefit of load limitation. It should insert them strategically to benefit load-
balancing along the longest part of the path. balancing along the longest part of the path.
PE1 may take into account multiple parameters when placing the ELs, PE1 may take into account multiple parameters when inserting ELs, as
as examples: examples:
o the ERLD value advertised by transit nodes. o The ERLD value advertised by transit nodes.
o the requirement of load balancing for a particular label value. o The requirement of load-balancing for a particular label value.
o any service provider preference: favor beginning of the path or o Any service provider preference: favor beginning of the path or
end of the path. end of the path.
In the Figure 4, a good strategy may be to use the following stack In the Figure 4, a good strategy may be to use the following stack
{VPN_label, ELI2,EL2, Adj_P6PE2, Adj_P5P6, Adj_P4P5, Adj_P3P4, ELI1, {VPN_label, ELI2,EL2, Adj_P6PE2, Adj_P5P6, Adj_P4P5, Adj_P3P4, ELI1,
EL1, Adj_Bundle_P2P3, Adj_P1P2}. The original stack requests P2 to EL1, Adj_Bundle_P2P3, Adj_P1P2}. The original stack requests P2 to
forward based on a bundle Adjacency segment that will require load forward based on a bundle Adjacency segment that will require load-
balancing. Therefore it is important to ensure that P2 can balancing. Therefore it is important to ensure that P2 can load-
loadbalance correctly. As P2 has a limited ERLD of 3, ELI/EL must be balance correctly. As P2 has a limited ERLD of 3, ELI/EL must be
inserted just next to the label P2 will use to forward. On the path inserted just next to the label that P2 will use to forward. On the
to PE2, P3 has also a limited ERLD, but P3 will forward based on a path to PE2, P3 has also a limited ERLD, but P3 will forward based on
basic adjacency segment that may require no load balancing. a basic adjacency segment that may require no load-balancing.
Therefore it does not seem important to ensure that P3 can do load Therefore it does not seem important to ensure that P3 can do load-
balancing despite of its limited ERLD. The next nodes along the balancing despite of its limited ERLD. The next nodes along the
forwarding path have a high ERLD that does not cause any issue, forwarding path have a high ERLD that does not cause any issue,
except P6, moreover P6 is using some LAGs to PE2 and is so expected except P6, moreover P6 is using some LAGs to PE2 and so is expected
to loadbalance. It becomes important to insert a new ELI/EL just to load-balance. It becomes important to insert a new ELI/EL just
next to P6 forwarding label. next to P6 forwarding label.
In the case above, the ingress node had enough label push capacity to In the case above, the ingress node had enough label push capacity to
ensure end to end load balancing taking into the path attributes. ensure end-to-end load-balancing taking into the path attributes.
There might be some cases, where the ingress node may not have the There might be some cases, where the ingress node may not have the
necessary label push capacity. necessary label imposition capacity.
7.1.2. Example 2 7.1.2. Example 2
ECMP LAG ECMP ECMP ECMP LAG ECMP ECMP
PE1 --- P1 --- P2 --- P3 --- P4 --- P5 --- P6 --- P7 --- P8 --- PE2 PE1 --- P1 --- P2 --- P3 --- P4 --- P5 --- P6 --- P7 --- P8 --- PE2
Figure 5 Figure 5
In the Figure 5, PE1 wants to forward MPLS VPN traffic over an In the Figure 5, PE1 wants to forward MPLS VPN traffic over an
explicit path to PE2 resulting in the following label stack to be explicit path to PE2 resulting in the following label stack to be
pushed onto the IP header : {VPN_label, Adj_Bundle_P8PE2, Adj_P7P8, pushed onto the IP header: {VPN_label, Adj_Bundle_P8PE2, Adj_P7P8,
Adj_Bundle_P6P7, Adj_P5P6, Adj_P4P5, Adj_P3P4, Adj_Bundle_P2P3, Adj_Bundle_P6P7, Adj_P5P6, Adj_P4P5, Adj_P3P4, Adj_Bundle_P2P3,
Adj_P1P2}. PE1 is limited to push a maximum of 11 labels, P2, P3 and Adj_P1P2}. PE1 is limited to push a maximum of 11 labels, P2, P3 and
P6 have a ERLD of 3 while others have a ERLD of 15. P6 have an ERLD of 3 while others have an ERLD of 15.
Using a similar strategy as the previous case may lead to a dilemma, Using a similar strategy as the previous case may lead to a dilemma,
as PE1 can only push a single ELI/EL while we may need a minimum of as PE1 can only push a single ELI/EL while we may need a minimum of
three to loadbalance the end to end path. An optimized stack that three to load-balance the end-to-end path. An optimized stack that
would enable end-to-end load balancing may be: {VPN_label, ELI3, EL3, would enable end-to-end load-balancing may be: {VPN_label, ELI3, EL3,
Adj_Bundle_P8PE2, Adj_P7P8, ELI2, EL2, Adj_Bundle_P6P7, Adj_P5P6, Adj_Bundle_P8PE2, Adj_P7P8, ELI2, EL2, Adj_Bundle_P6P7, Adj_P5P6,
Adj_P4P5, Adj_P3P4, ELI1, EL1, Adj_Bundle_P2P3, Adj_P1P2}. Adj_P4P5, Adj_P3P4, ELI1, EL1, Adj_Bundle_P2P3, Adj_P1P2}.
A decision needs to be taken to favor some part of the path for load A decision needs to be taken to favor some part of the path for load-
balancing considering that load balancing may not work on the other balancing considering that load-balancing may not work on the other
part. A service provider may decide to place the ELI/EL after P6 part. A service provider may decide to place the ELI/EL after the P6
forwarding label as it will allow P4 and P6 to loadbalance. Placing forwarding label as it will allow P4 and P6 to load-balance. Placing
the ELI/EL at bottom of the stack is also a possibility enabling load the ELI/EL at bottom of the stack is also a possibility enabling
balancing for P4 and P8. load-balancing for P4 and P8.
7.2. Considerations for the placement of entropy labels 7.2. Considerations for the placement of entropy labels
The sample cases described in the previous section shown that placing The sample cases described in the previous section showed that
the ELI/EL when the maximum number of labels to be pushed is limited placing the ELI/EL when the maximum number of labels to be pushed is
is not an easy decision and multiple criteria may be taken into limited is not an easy decision and multiple criteria may be taken
account. into account.
This section describes some considerations that could be taken into This section describes some considerations that could be taken into
account when placing ELI/ELs. This list of criteria is not account when placing ELI/ELs. This list of criteria is not
considered as exhaustive and an implementation MAY take into account considered as exhaustive and an implementation MAY take into account
additional criteria or tie breakers that are not documented here. additional criteria or tie-breakers that are not documented here.
An implementation SHOULD try to maximize the load-balancing where An implementation SHOULD try to maximize the load-balancing where
multiple ECMP paths are available and minimize the number of EL/ELIs multiple ECMP paths are available and minimize the number of EL/ELIs
that need to be inserted. In case of trade-off, an implementation that need to be inserted. In case of trade-off, an implementation
MAY provide flexibility to operator to select the criteria to be MAY provide flexibility to the operator to select the criteria to be
taken into account when placing EL/ELIs or the sub-objective to be considered when placing EL/ELIs or the sub-objective for which to
optimized for. optimize.
PE1 -- P1 -- P2 -- P3 -- P4 -- P5 -- ... -- P8 -- P9 -- PE2 PE1 -- P1 -- P2 -- P3 -- P4 -- P5 -- ... -- P8 -- P9 -- PE2
| | | |
P3'--- P4'--- P5' P3'--- P4'--- P5'
Figure 6 Figure 6
The figure above will be used as reference in the following sub The figure above will be used as reference in the following
sections. subsections.
7.2.1. ERLD value 7.2.1. ERLD value
As mentioned in Section 7.1, the ERLD value is an important parameter As mentioned in Section 7.1, the ERLD value is an important parameter
to take into account when inserting ELI/EL as if an ELI/EL does not to consider when inserting ELI/EL as if an ELI/EL does not fall
fall within the ERLD of a node on the path, the node will not be able within the ERLD of a node on the path, the node will not be able to
to loadbalance the traffic in an efficient way. load-balance the traffic efficiently.
The ERLD value can be advertised via protocols and those extensions The ERLD value can be advertised via protocols and those extensions
are described in separate documents [I-D.ietf-isis-mpls-elc] and are described in separate documents [I-D.ietf-isis-mpls-elc] and
[I-D.ietf-ospf-mpls-elc]. [I-D.ietf-ospf-mpls-elc].
Let's consider a path from PE1 to PE2 using the following stack Let's consider a path from PE1 to PE2 using the following stack
pushed by PE1: {Service_label, Adj_PE2P9, Node_P9, Adj_P1P2}. pushed by PE1: {Service_label, Adj_PE2P9, Node_P9, Adj_P1P2}.
Using the ERLD as an input parameter may help to minimize the number Using the ERLD as an input parameter may help to minimize the number
of required ELI/EL pairs to be inserted. An ERLD value must be of required ELI/EL pairs to be inserted. An ERLD value must be
retrieved for each SPRING label in the label stack. retrieved for each SPRING label in the label stack.
For a label bound to an adjacency segment, the ERLD is the ERLD of For a label bound to an adjacency segment, the ERLD is the ERLD of
the node that advertised the adjacency segment. In the example the node that advertised the adjacency segment. In the example
above, the ERLD associated with Adj_P1P2 would be the ERLD of router above, the ERLD associated with Adj_P1P2 would be the ERLD of router
P1 as P1 will perform the forwarding based on Adj_P1P2 label. P1 as P1 will perform the forwarding based on the Adj_P1P2 label.
For a label bound to a node segment, multiple strategies MAY be For a label bound to a node segment, multiple strategies MAY be
implemented. An implementation may try to evaluate the minimum ERLD implemented. An implementation may try to evaluate the minimum ERLD
value along the node segment path. If the implementation cannot find value along the node segment path. If an implementation cannot find
the minimum ERLD along the path of the segment, it can use the ERLD the minimum ERLD along the path of the segment, it can use the ERLD
of the starting node instead. In the example above, if the of the starting node instead. In the example above, if the
implementation supports computation of minimum ERLD along the path, implementation supports computation of minimum ERLD along the path,
the ERLD associated to label Node_P9 would be the minimum ERLD the ERLD associated to label Node_P9 would be the minimum ERLD
between nodes {P2,P3,P4 ..., P8}. If the implementation does not between nodes {P2,P3,P4 ..., P8}. If an implementation does not
support the computation of minimum ERLD, it should consider the ERLD support the computation of minimum ERLD, it should consider the ERLD
of P2 (starting node that will forward based on Node_P9 label). of P2 (starting node that will forward based on the Node_P9 label).
For a label bound to a binding segment, if the binding segment For a label bound to a binding segment, if the binding segment
describes a path, an implementation may also try to evaluate the describes a path, an implementation may also try to evaluate the
minimum ERLD along this path. If the implementation cannot find the minimum ERLD along this path. If the implementation cannot find the
minimum ERLD along the path of the segment, it can use the ERLD of minimum ERLD along the path of the segment, it can use the ERLD of
the starting node instead. the starting node instead.
7.2.2. Segment type 7.2.2. Segment type
Depending of the type of segment a particular label is bound to, an Depending of the type of segment a particular label is bound to, an
implementation may deduce that this particular label will be subject implementation may deduce that this particular label will be subject
to load balancing on the path. to load-balancing on the path.
7.2.2.1. Node-SID 7.2.2.1. Node-SID
An MPLS label bound to a Node-SID represents a path that may cross An MPLS label bound to a Node-SID represents a path that may cross
multiple hops. Load balancing may be needed on the node starting multiple hops. Load-balancing may be needed on the node starting
this path but also on any node along the path. this path but also on any node along the path.
Let's consider a path from PE1 to PE2 using the following stack Let's consider a path from PE1 to PE2 using the following stack
pushed by PE1: {Service_label, Adj_PE2P9, Node_P9, Adj_P1P2}. pushed by PE1: {Service_label, Adj_PE2P9, Node_P9, Adj_P1P2}.
If, for example, PE1 is limited to 6 labels to be pushed, it can add If, for example, PE1 is limited to pushing 6 labels, it can add a
a single ELI/EL within the label stack. An operator may want to single ELI/EL within the label stack. An operator may want to favor
favor a placement that would allow load balancing along the node-SID a placement that would allow load-balancing along the Node-SID path.
path. In the figure above, P3 which is along the node-SID path In the figure above, P3 which is along the Node-SID path requires
requires load balancing on two equal cost paths. load-balancing on two equal-cost paths.
An implementation may try to evaluate if load balancing is really An implementation may try to evaluate if load-balancing is really
required within a node segment path: this could be done by running required within a node segment path. This could be done by running
additional SPT computation and by analysis the node segment path. So an additional SPT computation and analysis of the node segment path
a node segment that does not really require load balancing may not be to prevent a node segment that does not really require load-balancing
preferred when placing EL/ELIs. Such inspection may be time from being preferred when placing EL/ELIs. Such inspection may be
consuming for implementations without 100% guarantee, as a node time consuming for implementations and without a 100% guarantee, as a
segment path may use LAG that could be invisible from the IP node segment path may use LAG that could be invisible from the IP
topology. A simpler approach would be to consider that a label bound topology. A simpler approach would be to consider that a label bound
to a Node-SID will be subject to load balancing and so requires an to a Node-SID will be subject to load-balancing and requires an EL/
EL/ELI. ELI.
7.2.2.2. Adjacency-SID representing an ECMP bundle 7.2.2.2. Adjacency-SID representing an ECMP bundle
When an adjacency segment representing an ECMP bundle is used within When an adjacency segment representing an ECMP bundle is used within
a label stack, an implementation can deduce that load balancing is a label stack, an implementation can deduce that load-balancing is
expected at the node that advertised this adjacency segment. An expected at the node that advertised this adjacency segment. An
implementation could then favor this particular label value when implementation could then favor this particular label value when
placing ELI/ELs. placing ELI/ELs.
7.2.2.3. Adjacency-SID representing a single IP link 7.2.2.3. Adjacency-SID representing a single IP link
When an adjacency segment representing a single IP link is used When an adjacency segment representing a single IP link is used
within a label stack, an implementation can deduce that load within a label stack, an implementation can deduce that load-
balancing may not be expected at the node that advertised this balancing may not be expected at the node that advertised this
adjacency segment. adjacency segment.
The implementation could then decide to place ELI/ELs to favor other The implementation could then decide to place ELI/ELs to favor other
LSRs than the one advertising this adjacency segment. LSRs than the one advertising this adjacency segment.
Readers should note that an adjacency segment representing a single Readers should note that an adjacency segment representing a single
IP link may require load balancing. This is the case when a LAG (L2 IP link may require load-balancing. This is the case when a LAG (L2
bundle) is implemented between two IP nodes and L2 bundle SR bundle) is implemented between two IP nodes and the L2 bundle SR
extensions [I-D.ietf-isis-l2bundles] are not implemented. In such extensions [I-D.ietf-isis-l2bundles] are not implemented. In such
case, it may be interesting to keep the possibility to insert an EL/ case, it may be interesting to keep the possibility to insert an EL/
ELI in a readable position for the LSR advertising the label ELI in a readable position for the LSR advertising the label
associated to the adjacency segment. associated with the adjacency segment.
7.2.2.4. Adjacency-SID representing a single link within a L2 bundle 7.2.2.4. Adjacency-SID representing a single link within an L2 bundle
When L2 bundle SR extensions [I-D.ietf-isis-l2bundles] are used, When L2 bundle SR extensions [I-D.ietf-isis-l2bundles] are used,
adjacency segments may be advertised for each member of the bundle. adjacency segments may be advertised for each member of the bundle.
In this case, an implementation can deduce that load balancing is not In this case, an implementation can deduce that load-balancing is not
expected on the LSR advertising this segment and could then decide to expected on the LSR advertising this segment and could then decide to
place ELI/ELs to favor other LSRs than the one advertising this place ELI/ELs to favor other LSRs than the one advertising this
adjacency segment. adjacency segment.
7.2.2.5. Adjacency-SID representing a L2 bundle 7.2.2.5. Adjacency-SID representing an L2 bundle
When L2 bundle SR extensions [I-D.ietf-isis-l2bundles] are used, an When L2 bundle SR extensions [I-D.ietf-isis-l2bundles] are used, an
adjacency segment may be advertised to represent the bundle. In this adjacency segment may be advertised to represent the bundle. In this
case, an implementation can deduce that load balancing is expected on case, an implementation can deduce that load-balancing is expected on
the LSR advertising this segment and could then decide to place ELI/ the LSR advertising this segment and could then decide to place ELI/
ELs to favor this LSR. ELs to favor this LSR.
7.2.3. Maximizing number of LSRs that will loadbalance 7.2.3. Maximizing number of LSRs that will load-balance
When placing ELI/ELs, an implementation may try to maximize the When placing ELI/ELs, an implementation may try to maximize the
number of LSRs that LSRs that both needs to load balance (i.e. have number of LSRs that both need to load-balance (i.e., have ECMP paths)
ECMP paths) and that will be able to perform load balancing (i.e. EL and that will be able to perform load-balancing (i.e., the EL label
label is within their ERLD). is within their ERLD).
Let's consider a path from PE1 to PE2 using the following stack Let's consider a path from PE1 to PE2 using the following stack
pushed by PE1: {Service_label, Adj_PE2P9, Node_P9, Adj_P1P2}. All pushed by PE1: {Service_label, Adj_PE2P9, Node_P9, Adj_P1P2}. All
routers have ERLD of 10, expect P1 and P2 which have a ERLD of 4. routers have an ERLD of 10, expect P1 and P2 which have an ERLD of 4.
PE1 is able to push 6 labels, so only a single ELI/EL can be added. PE1 is able to push 6 labels, so only a single ELI/EL can be added.
In the example above, adding ELI/EL next to Adj_P1P2 will only allow In the example above, adding ELI/EL next to Adj_P1P2 will only allow
load balancing at P1 while inserting it next to Adj_PE2P9, will allow load-balancing at P1 while inserting it next to Adj_PE2P9, will allow
load balancing at P2,P3 ... P9 so maximizing the number of LSRs that load-balancing at P2,P3 ... P9 and maximizing the number of LSRs that
could perform load balancing. could perform load-balancing.
7.2.4. Preference for a part of the path 7.2.4. Preference for a part of the path
An implementation may propose to favor a part of the end-to-end path An implementation may propose to favor a part of the end-to-end path
when the number of EL/ELI that can be pushed is not enough to cover when the number of EL/ELI that can be pushed is not enough to cover
the entire path. As example, a service provider may want to favor the entire path. As example, a service provider may want to favor
load balancing at the beginning of the path or at the end of path, so load-balancing at the beginning of the path or at the end of path, so
the implementation should prefer putting the ELI/ELs near the top or the implementation should prefer putting the ELI/ELs near the top or
near of the bottom of the stack. near of the bottom of the stack.
7.2.5. Combining criteria 7.2.5. Combining criteria
An implementation can combine multiple criteria to determine the best An implementation can combine multiple criteria to determine the best
EL/ELIs placement. But combining too much criteria may lead to EL/ELIs placement. But combining too much criteria may lead to
implementation complexity and high control plane resource implementation complexity and high control plane resource
consumption. Each time, the network topology will change, a new consumption. Each time the network topology changes, a new
evaluation of the EL/ELI placement will be necessary. evaluation of the EL/ELI placement will be necessary for each
impacted LSPs.
8. A simple algorithm example 8. A simple algorithm example
A simple implementation can only take into account ERLD when placing A simple implementation can only take into account ERLD when placing
ELI/EL while keep minimizing the number of EL/ELIs inserted and ELI/EL while keep minimizing the number of EL/ELIs inserted and
maximizing the number of LSRs that can loadbalance. maximizing the number of LSRs that can load-balance.
The algorithm example is based on the following considerations: The algorithm example is based on the following considerations:
o An LSR that is limited in the number of <ELI, EL> pairs that it o An LSR that is limited in the number of <ELI, EL> pairs that it
can insert SHOULD insert such pairs deeper in the stack. can insert SHOULD insert such pairs deeper in the stack.
o An LSR should try to insert <ELI, EL> pairs at positions so that o An LSR should try to insert <ELI, EL> pairs at positions so that
for the maximum number of transit LSRs, the EL occurs within the for the maximum number of transit LSRs, the EL occurs within the
ERLD of the incoming packet to that LSR. ERLD of those LSRs.
o An LSR should try to insert the minimum number of such pairs while o An LSR should try to insert the minimum number of such pairs while
trying to satisfy the above criteria. trying to satisfy the above criteria.
The pseudocode of the example is shown below. The pseudocode of the example is shown below.
Initialize the current EL insertion point to the Initialize the current EL insertion point to the
bottommost label in the stack that is EL-capable bottommost label in the stack that is EL-capable
while (local-node can push more <ELI,EL> pairs OR while (local-node can push more <ELI,EL> pairs OR
insertion point is not above label stack) { insertion point is not above label stack) {
skipping to change at page 17, line 20 skipping to change at page 17, line 20
move new insertion point up from current insertion point until move new insertion point up from current insertion point until
((last inserted EL is below the ERLD) AND (ERLD > 2) ((last inserted EL is below the ERLD) AND (ERLD > 2)
AND AND
(new insertion point is EL-capable)) (new insertion point is EL-capable))
set current insertion point to new insertion point set current insertion point to new insertion point
} }
Figure 7: Example algorithm to insert <ELI, EL> pairs in a label Figure 7: Example algorithm to insert <ELI, EL> pairs in a label
stack stack
When this algorithm is applied to the example described in Section 3 When this algorithm is applied to the example described in Section 3,
it will result in ELs being inserted in two positions, one below the it will result in ELs being inserted in two positions, one below the
label L_N-D and another below L_N-P3. Thus the resulting label stack label L_N-D and another below L_N-P3. Thus the resulting label stack
would be {L_N-P3, ELI, EL, L_A-L1, L_N-D, ELI, EL} would be {L_N-P3, ELI, EL, L_A-L1, L_N-D, ELI, EL}
9. Deployment Considerations 9. Deployment Considerations
As long as LSR node dataplane capabilities with be limited (number of As long as LSR node dataplane capabilities with be limited (number of
labels that can be pushed, or number of labels that can be labels that can be pushed, or number of labels that can be
inspected), hop-by-hop load balancing of SPRING encapsulated flows inspected), hop-by-hop load-balancing of SPRING encapsulated flows
will require trade-offs. will require trade-offs.
Entropy label is still a good and usable solution as it allows load Entropy label is still a good and usable solution as it allows load-
balancing without having to perform a deep packet inspection on each balancing without having to perform a deep packet inspection on each
LSR: it does not sound reasonable to have a LSR inspecting UDP port LSR: it does not seem reasonable to have an LSR inspecting UDP ports
within a GRE tunnel carried over a 15 labels SPRING tunnel. within a GRE tunnel carried over a 15 label SPRING tunnel.
Due to the limited capacity of reading a deep stack of MPLS labels, Due to the limited capacity of reading a deep stack of MPLS labels,
multiple EL/ELIs may be required within the stack which directly multiple EL/ELIs may be required within the stack which directly
impacts the capacity of the head-end to push a deep stack: each EL/ impacts the capacity of the head-end to push a deep stack: each EL/
ELI inserted requiring two additional labels to be pushed. ELI inserted requires two additional labels to be pushed.
Placement strategies of EL/ELIs are so required to find the best Placement strategies of EL/ELIs are required to find the best trade-
trade-off. Multiple criteria may be taken into account and some off. Multiple criteria may be taken into account and some level of
level of customization (by user) may be required to accommodate the customization (by the user) may be required to accommodate the
different deployments. Analyzing the path of each destination to different deployments. Analyzing the path of each destination to
determine the best EL/ELI placement may be time consuming for the determine the best EL/ELI placement may be time consuming for the
control plane, we encourage implementations to find the best trade- control plane, we encourage implementations to find the best trade-
off between simplicity, resource consumption and load balancing off between simplicity, resource consumption, and load-balancing
efficiency. efficiency.
In future, hardware and software capacity may increase dataplane In future, hardware and software capacity may increase dataplane
capabilities and may be remove some of those limits: this may capabilities and may be remove some of these limits, increasing load-
increase the capacity of load balancing using entropy labels. balancing capability using entropy labels.
10. Options considered 10. Options considered
Different options that were considered to arrive at the recommended Different options that were considered to arrive at the recommended
solution are documented in this section. solution are documented in this section.
10.1. Single EL at the bottom of the stack of tunnels 10.1. Single EL at the bottom of the stack of tunnels
In this option a single EL is used for the entire label stack. The In this option, a single EL is used for the entire label stack. The
source LSR S encodes the entropy label (EL) at the bottom of the source LSR S encodes the entropy label (EL) at the bottom of the
label stack. In the example described in Section 3, it will result label stack. In the example described in Section 3, it will result
in the label stack at LSR S to look like {L_N-P3, L_A-L1, L_N-D, ELI, in the label stack at LSR S to look like {L_N-P3, L_A-L1, L_N-D, ELI,
EL} {remaining packet header}. Note that the notation in [RFC6790] EL} {remaining packet header}. Note that the notation in [RFC6790]
is used to describe the label stack. An issue with this approach is is used to describe the label stack. An issue with this approach is
that as the label stack grows due an increase in the number of SIDs, that as the label stack grows due an increase in the number of SIDs,
the EL goes correspondingly deeper in the label stack. Hence transit the EL goes correspondingly deeper in the label stack. Hence,
LSRs have to access a larger number of bytes in the packet header transit LSRs have to access a larger number of bytes in the packet
when making forwarding decisions. In the example described in header when making forwarding decisions. In the example described in
Section 3, the LSR P1 would poorly load-balance traffic on the Section 3, the LSR P1 would load-balance traffic poorly on the
parallel links L3, L4 since the EL is below the RLD of the packet parallel links L3 and L4 since the EL is below the ERLD of the packet
received by P1. A load balanced network design using this approach received by P1. A load-balanced network design using this approach
must ensure that all intermediate LSRs have the capability to must ensure that all intermediate LSRs have the capability to
traverse the maximum label stack depth as required for that traverse the maximum label stack depth as required for the
application that uses source routed stacking. application that uses source routed stacking.
In the case where the hardware is capable of pushing a single <ELI, In the case where the hardware is capable of pushing a single <ELI,
EL> pair at any depth, this option is the same as the recommended EL> pair at any depth, this option is the same as the recommended
solution in Section 7. solution in Section 7.
This option was rejected since there exist a number of hardware This option was rejected since there exist a number of hardware
implementations which have a low maximum readable label depth. implementations which have a low maximum readable label depth.
Choosing this option can lead to a loss of load-balancing using EL in Choosing this option can lead to a loss of load-balancing using EL in
a significant part of the network but that is a critical requirement a significant part of the network when that is a critical requirement
in a service provider network. in a service-provider network.
10.2. An EL per tunnel in the stack 10.2. An EL per tunnel in the stack
In this option each tunnel in the stack can be given its own EL. The In this option, each tunnel in the stack can be given its own EL.
source LSR pushes an <ELI, EL> before pushing a tunnel label when The source LSR pushes an <ELI, EL> before pushing a tunnel label when
load balancing is required to direct traffic on that tunnel. In the load-balancing is required to direct traffic on that tunnel. In the
example described in Section 3, the source LSR S encoded label stack example described in Section 3, the source LSR S encoded label stack
would be {L_N-P3, ELI, EL, L_A-L1, L_N-D, ELI, EL} where all the ELs would be {L_N-P3, ELI, EL, L_A-L1, L_N-D, ELI, EL} where all the ELs
can be the same. Accessing the EL at an intermediate LSR is can be the same. Accessing the EL at an intermediate LSR is
independent of the depth of the label stack and hence independent of independent of the depth of the label stack and hence independent of
the specific application that uses source routed tunnels with label the specific application that uses source routed tunnels with label
stacking in that network. A drawback is that the depth of the label stacking. A drawback is that the depth of the label stack grows
stack grows significantly, almost 3 times as the number of labels in significantly, almost 3 times as the number of labels in the label
the label stack. The network design should ensure that source LSRs stack. The network design should ensure that source LSRs have the
should have the capability to push such a deep label stack. Also, capability to push such a deep label stack. Also, the bandwidth
the bandwidth overhead and potential MTU issues of deep label stacks overhead and potential MTU issues of deep label stacks should be
should be accounted for in the network design. considered in the network design.
In the case where the RLD is the minimum value (3) for all LSRs, all In the case where the RLD is the minimum value (3) for all LSRs, all
LSRs are EL capable and the LSR that is inserting <ELI, EL> pairs has LSRs are EL capable and the LSR that is inserting <ELI, EL> pairs has
no limit on how many it can insert then this option is the same as no limit on how many it can insert then this option is the same as
the recommended solution in Section 7. the recommended solution in Section 7.
This option was rejected due to the existence of hardware This option was rejected due to the existence of hardware
implementations that can push a limited number of labels on the label implementations that can push a limited number of labels on the label
stack. Choosing this option would result in a hardware requirement stack. Choosing this option would result in a hardware requirement
to push two additional labels per tunnel label. Hence it would to push two additional labels per tunnel label. Hence it would
skipping to change at page 19, line 38 skipping to change at page 19, line 38
the EL from the outer tunnel when that tunnel is terminated and re- the EL from the outer tunnel when that tunnel is terminated and re-
inserting it below the next inner tunnel label during the label swap inserting it below the next inner tunnel label during the label swap
operation. The LSR that stacks tunnels should insert an EL below the operation. The LSR that stacks tunnels should insert an EL below the
outermost tunnel. It should not insert ELs for any inner tunnels. outermost tunnel. It should not insert ELs for any inner tunnels.
Also, the penultimate hop LSR of a segment must not pop the ELI and Also, the penultimate hop LSR of a segment must not pop the ELI and
EL even though they are exposed as the top labels since the EL even though they are exposed as the top labels since the
terminating LSR of that segment would re-use the EL for the next terminating LSR of that segment would re-use the EL for the next
segment. segment.
In Section 3 above, the source LSR S encoded label stack would be In Section 3 above, the source LSR S encoded label stack would be
{L_N-P3, ELI, EL, L_A-L1, L_N-D}. At P1 the outgoing label stack {L_N-P3, ELI, EL, L_A-L1, L_N-D}. At P1, the outgoing label stack
would be {L_N-P3, ELI, EL, L_A-L1, L_N-D} after it has load balanced would be {L_N-P3, ELI, EL, L_A-L1, L_N-D} after it has load-balanced
to one of the links L3 or L4. At P3 the outgoing label stack would to one of the links L3 or L4. At P3 the outgoing label stack would
be {L_N-D, ELI, EL}. At P2 the outgoing label stack would be {L_N-D, be {L_N-D, ELI, EL}. At P2, the outgoing label stack would be {L_N-
ELI, EL} and it would load balance to one of the nexthop LSRs P4 or D, ELI, EL} and it would load-balance to one of the nexthop LSRs P4
P5. Accessing the EL at an intermediate LSR (e.g. P1) is or P5. Accessing the EL at an intermediate LSR (e.g., P1) is
independent of the depth of the label stack and hence independent of independent of the depth of the label stack and hence independent of
the specific use-case to which the label stack is applied. the specific use-case to which the label stack is applied.
This option was rejected due to the significant change in label swap This option was rejected due to the significant change in label swap
operations that would be required for existing hardware. operations that would be required for existing hardware.
10.4. EL at top of stack 10.4. EL at top of stack
A slight variant of the re-usable EL option is to keep the EL at the A slight variant of the re-usable EL option is to keep the EL at the
top of the stack rather than below the tunnel label. In this case top of the stack rather than below the tunnel label. In this case,
each LSR that is not terminating a segment should continue to keep each LSR that is not terminating a segment should continue to keep
the received EL at the top of the stack when forwarding the packet the received EL at the top of the stack when forwarding the packet
along the segment. An LSR that terminates a segment should use the along the segment. An LSR that terminates a segment should use the
EL from the terminated segment at the top of the stack when EL from the terminated segment at the top of the stack when
forwarding onto the next segment. forwarding onto the next segment.
This option was rejected due to the significant change in label swap This option was rejected due to the significant change in label swap
operations that would be required for existing hardware. operations that would be required for existing hardware.
10.5. ELs at readable label stack depths 10.5. ELs at readable label stack depths
skipping to change at page 20, line 39 skipping to change at page 20, line 39
protocol such as an IGP. protocol such as an IGP.
Applying this method to the example in Section 3 above, if LSR P1 Applying this method to the example in Section 3 above, if LSR P1
needs to have the EL within a depth of 4, then the source LSR S needs to have the EL within a depth of 4, then the source LSR S
encoded label stack would be {L_N-P3, ELI, EL, L_A-L1, L_N-D, ELI, encoded label stack would be {L_N-P3, ELI, EL, L_A-L1, L_N-D, ELI,
EL} where all the ELs would typically have the same value. EL} where all the ELs would typically have the same value.
In the case where the RLD has different values along the path and the In the case where the RLD has different values along the path and the
LSR that is inserting <ELI, EL> pairs has no limit on how many pairs LSR that is inserting <ELI, EL> pairs has no limit on how many pairs
it can insert, and it knows the appropriate positions in the stack it can insert, and it knows the appropriate positions in the stack
where they should be inserted, then this option is the same as the where they should be inserted, this option is the same as the
recommended solution in Section 7. recommended solution in Section 7.
Note that a refinement of this solution which balances the number of Note that a refinement of this solution which balances the number of
pushed labels against the desired entropy is the solution described pushed labels against the desired entropy is the solution described
in Section 7. in Section 7.
11. Acknowledgements 11. Acknowledgements
The authors would like to thank John Drake, Loa Andersson, Curtis The authors would like to thank John Drake, Loa Andersson, Curtis
Villamizar, Greg Mirsky, Markus Jork, Kamran Raza, Carlos Pignataro, Villamizar, Greg Mirsky, Markus Jork, Kamran Raza, Carlos Pignataro,
 End of changes. 107 change blocks. 
204 lines changed or deleted 203 lines changed or added

This html diff was produced by rfcdiff 1.45. The latest version is available from http://tools.ietf.org/tools/rfcdiff/