Internet Engineering Task Force (IETF)
Request for Comments: 8584
Updates: 7432
Category: Standards Track
ISSN: 2070-1721
J. Rabadan, Ed.
Nokia
S. Mohanty, Ed.
A. Sajassi
Cisco
J. Drake
Juniper
K. Nagaraj
S. Sathappan
Nokia
April 2019

Framework for Ethernet VPN Designated Forwarder Election Extensibility

Abstract

An alternative to the default Designated Forwarder (DF) selection algorithm in Ethernet VPNs (EVPNs) is defined. The DF is the Provider Edge (PE) router responsible for sending Broadcast, Unknown Unicast, and Multicast (BUM) traffic to a multihomed Customer Edge (CE) device on a given VLAN on a particular Ethernet Segment (ES). In addition, the ability to influence the DF election result for a VLAN based on the state of the associated Attachment Circuit (AC) is specified. This document clarifies the DF election Finite State Machine in EVPN services. Therefore, it updates the EVPN specification (RFC 7432).

Status of This Memo

This is an Internet Standards Track document.

This document is a product of the Internet Engineering Task Force (IETF). It represents the consensus of the IETF community. It has received public review and has been approved for publication by the Internet Engineering Steering Group (IESG). Further information on Internet Standards is available in Section 2 of RFC 7841.

Information about the current status of this document, any errata, and how to provide feedback on it may be obtained at https://www.rfc-editor.org/info/rfc8584.

Copyright Notice

Copyright © 2019 IETF Trust and the persons identified as the document authors. All rights reserved.

This document is subject to BCP 78 and the IETF Trust's Legal Provisions Relating to IETF Documents (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 Simplified BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Simplified BSD License.

Table of Contents

   1. Introduction ....................................................3
      1.1. Conventions and Terminology ................................3
      1.2. Default Designated Forwarder (DF) Election in EVPN
           Services ...................................................5
      1.3. Problem Statement ..........................................8
           1.3.1. Unfair Load Balancing and Service Disruption ........8
           1.3.2. Traffic Black-Holing on Individual AC Failures .....10
      1.4. The Need for Extending the Default DF Election in
           EVPN Services .............................................12
   2. Designated Forwarder Election Protocol and BGP Extensions ......13
      2.1. The DF Election Finite State Machine (FSM) ................13
      2.2. The DF Election Extended Community ........................16
           2.2.1. Backward Compatibility .............................19
   3. The Highest Random Weight DF Election Algorithm ................19
      3.1. HRW and Consistent Hashing ................................20
      3.2. HRW Algorithm for EVPN DF Election ........................20
   4. The AC-Influenced DF Election Capability .......................22
      4.1. AC-Influenced DF Election Capability for
           VLAN-Aware Bundle Services ................................24
   5. Solution Benefits ..............................................25
   6. Security Considerations ........................................26
   7. IANA Considerations ............................................27
   8. References .....................................................28
      8.1. Normative References ......................................28
      8.2. Informative References ....................................29
   Acknowledgments ...................................................30
   Contributors ......................................................30
   Authors' Addresses ................................................31

1. Introduction

The Designated Forwarder (DF) in Ethernet VPNs (EVPNs) is the Provider Edge (PE) router responsible for sending Broadcast, Unknown Unicast, and Multicast (BUM) traffic to a multihomed Customer Edge (CE) device on a given VLAN on a particular Ethernet Segment (ES). The DF is elected from the set of multihomed PEs attached to a given ES, each of which advertises an ES route for the ES as identified by its Ethernet Segment Identifier (ESI). By default, the EVPN uses a DF election algorithm referred to as "service carving". The DF election algorithm is based on a modulus function (V mod N) that takes the number of PEs in the ES (N) and the VLAN value (V) as input. This document addresses inefficiencies in the default DF election algorithm by defining a new DF election algorithm and an ability to influence the DF election result for a VLAN, depending on the state of the associated Attachment Circuit (AC). In order to avoid any ambiguity with the identifier used in the DF election algorithm, this document uses the term "Ethernet Tag" instead of "VLAN". This document also creates a registry with IANA for future DF election algorithms and capabilities (see Section 7). It also presents a formal definition and clarification of the DF election Finite State Machine (FSM). Therefore, this document updates [RFC7432], and EVPN implementations MUST conform to the prescribed FSM.

The procedures described in this document apply to DF election in all EVPN solutions, including those described in [RFC7432] and [RFC8214]. Apart from the formal description of the FSM, this document does not intend to update other procedures described in [RFC7432]; it only aims to improve the behavior of the DF election on PEs that are upgraded to follow the procedures described in this document.

1.1. Conventions and Terminology

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.

  • AC: Attachment Circuit. An AC has an Ethernet Tag associated with it.
  • ACS: Attachment Circuit Status.
  • BUM: Broadcast, unknown unicast, and multicast.
  • DF: Designated Forwarder.
  • NDF: Non-Designated Forwarder.
  • BDF: Backup Designated Forwarder.
  • Ethernet A-D per ES route: Refers to Route Type 1 as defined in [RFC7432] or to Auto-discovery per Ethernet Segment route.
  • Ethernet A-D per EVI route: Refers to Route Type 1 as defined in [RFC7432] or to Auto-discovery per EVPN Instance route.
  • ES: Ethernet Segment.
  • ESI: Ethernet Segment Identifier.
  • EVI: EVPN Instance.
  • MAC-VRF: A Virtual Routing and Forwarding table for Media Access Control (MAC) addresses on a PE.
  • BD: Broadcast Domain. An EVI may be comprised of one BD (VLAN-based or VLAN Bundle services) or multiple BDs (VLAN-aware Bundle services).
  • Bridge table: An instantiation of a BD on a MAC-VRF.
  • HRW: Highest Random Weight.
  • VID: VLAN Identifier.
  • CE-VID: Customer Edge VLAN Identifier.
  • Ethernet Tag: Used to represent a BD that is configured on a given ES for the purpose of DF election. Note that any of the following may be used to represent a BD: VIDs (including Q-in-Q tags), configured IDs, VNIs (Virtual Extensible Local Area Network (VXLAN) Network Identifiers), normalized VIDs, I-SIDs (Service Instance Identifiers), etc., as long as the representation of the BDs is configured consistently across the multihomed PEs attached to that ES. The Ethernet Tag value MUST be different from zero.
  • Ethernet Tag ID: Refers to the identifier used in the EVPN routes defined in [RFC7432]. Its value may be the same as the Ethernet Tag value (see the definition for Ethernet Tag) when advertising routes for VLAN-aware Bundle services. Note that in the case of VLAN-based or VLAN Bundle services, the Ethernet Tag ID is zero.
  • DF election procedure: Also called "DF election". Refers to the process in its entirety, including the discovery of the PEs in the ES, the creation and maintenance of the PE candidate list, and the selection of a PE.
  • DF algorithm: A component of the DF election procedure. Strictly refers to the selection of a PE for a given <ES, Ethernet Tag>.
  • RR: Route Reflector. A network routing component for BGP [RFC4456]. It offers an alternative to the logical full-mesh requirement of the Internal Border Gateway Protocol (IBGP). The purpose of the RR is concentration. Multiple BGP routers can peer with a central point, the RR -- acting as a route reflector server -- rather than peer with every other router in a full mesh. This results in an O(N) peering as opposed to O(N^2).
  • TTL: Time To Live.

This document also assumes that the reader is familiar with the terminology provided in [RFC7432].

1.2. Default Designated Forwarder (DF) Election in EVPN Services

[RFC7432] defines the DF as the EVPN PE responsible for:

  • Flooding BUM traffic on a given Ethernet Tag on a particular ES to the CE. This is valid for Single-Active and All-Active EVPN multihoming.
  • Sending unicast traffic on a given Ethernet Tag on a particular ES to the CE. This is valid for Single-Active multihoming.

Figure 1 illustrates an example that we will use to explain the DF function.

                        +---------------+
                        |   IP/MPLS     |
                        |   Core        |
          +----+ ES1 +----+           +----+
          | CE1|-----|    |           |    |____ES2
          +----+     | PE1|           | PE2|    \
                     |    |           +----+     \+----+
                     +----+             |         | CE2|
                        |             +----+     /+----+
                        |             |    |____/   |
                        |             | PE3|    ES2 /
                        |             +----+       /
                        |               |         /
                        +-------------+----+     /
                                      | PE4|____/ES2
                                      |    |
                                      +----+

Figure 1: EVPN Multihoming

Figure 1 illustrates a case where there are two ESes: ES1 and ES2. PE1 is attached to CE1 via ES1, whereas PE2, PE3, and PE4 are attached to CE2 via ES2, i.e., PE2, PE3, and PE4 form a redundancy group. Since CE2 is multihomed to different PEs on the same ES, it is necessary for PE2, PE3, and PE4 to agree on a DF to satisfy the above-mentioned requirements.

The effect of forwarding loops in a Layer 2 network is particularly severe because of the broadcast nature of Ethernet traffic and the lack of a TTL. Therefore, it is very important that, in the case of a multihomed CE, only one of the PEs be used to send BUM traffic to it.

One of the prerequisites for this support is that participating PEs must agree amongst themselves as to who would act as the DF. This needs to be achieved through a distributed algorithm in which each participating PE independently and unambiguously selects one of the participating PEs as the DF, and the result should be consistent and unanimous.

The default algorithm for DF election defined by [RFC7432] at the granularity of (ESI, EVI) is referred to as "service carving". In this document, service carving and the default DF election algorithm are used interchangeably. With service carving, it is possible to elect multiple DFs per ES (one per EVI) in order to perform load balancing of traffic destined to a given ES. The objective is that the load-balancing procedures should carve up the BD space among the redundant PE nodes evenly, in such a way that every PE is the DF for a distinct set of EVIs.

The DF election algorithm (as described in [RFC7432], Section 8.5) is based on a modulus operation. The PEs to which the ES (for which DF election is to be carried out per EVI) is multihomed form an ordered (ordinal) list in ascending order by PE IP address value. For example, there are N PEs: PE0, PE1,... PE(N-1) ranked as per increasing IP addresses in the ordinal list; then, for each VLAN with Ethernet Tag V, configured on ES1, PEx is the DF for VLAN V on ES1 when x equals (V mod N). In the case of a VLAN Bundle, only the lowest VLAN is used. In the case when the planned density is high (meaning there are a significant number of VLANs and the Ethernet Tags are uniformly distributed), the thinking is that the DF election will be spread across the PEs hosting that ES and good load balancing can be achieved.

However, the described default DF election algorithm has some undesirable properties and, in some cases, can be somewhat disruptive and unfair. This document describes some of those issues and defines a mechanism for dealing with them. These mechanisms do involve changes to the default DF election algorithm, but they do not require any changes to the EVPN route exchange, and changes in the EVPN routes will be minimal.

In addition, there is a need to extend the DF election procedures so that new algorithms and capabilities are possible. A single algorithm (the default DF election algorithm) may not meet the requirements in all the use cases.

Note that while [RFC7432] elects a DF per <ES, EVI>, this document elects a DF per <ES, BD>. This means that unlike [RFC7432], where for a VLAN-aware Bundle service EVI there is only one DF for the EVI, this document specifies that there will be multiple DFs, one for each BD configured in that EVI.

1.3. Problem Statement

This section describes some potential issues with the default DF election algorithm.

1.3.1. Unfair Load Balancing and Service Disruption

There are three fundamental problems with the current default DF election algorithm.

  1. The algorithm will not perform well when the Ethernet Tag follows a non-uniform distribution -- for instance, when the Ethernet Tags are all even or all odd. In such a case, let us assume that the ES is multihomed to two PEs; one of the PEs will be elected as the DF for all of the VLANs. This is very suboptimal. It defeats the purpose of service carving, as the DFs are not really evenly spread across the PEs hosting the ES. In fact, in this particular case, one of the PEs does not get elected as the DF at all, so it does not participate in DF responsibilities at all. Consider another example where, referring to Figure 1, let's assume that (1) PE2, PE3, and PE4 are listed in ascending order by IP address and (2) each VLAN configured on ES2 is associated with an Ethernet Tag of the form (3x+1), where x is an integer. This will result in PE3 always being selected as the DF.
  1. The Ethernet Tag that identifies the BD can be as large as 2^24; however, it is not guaranteed that the tenant BD on the ES will conform to a uniform distribution. In fact, it is up to the customer what BDs they will configure on the ES. Quoting [Knuth]:

In general, we want to avoid values of M that divide r^k+a or r^k-a, where k and a are small numbers and r is the radix of the alphabetic character set (usually r=64, 256 or 100), since a remainder modulo such a value of M tends to be largely a simple superposition of key digits. Such considerations suggest that we choose M to be a prime number such that r^k!=a(modulo)M or r^k!=?a(modulo)M for small k & a.

In our case, N is the number of PEs (Section 8.5 of [RFC7432]). N corresponds to M above. Since N, N-1, or N+1 need not satisfy the primality properties of M, as per the modulo-based DF assignment [RFC7432], whenever a PE goes down or a new PE boots up (attached to the same ES), the modulo scheme will not necessarily map BDs to PEs uniformly.

  1. Disruption is another problem. Consider a case when the same ES is multihomed to a set of PEs. When the ES is DOWN in one of the PEs, say PE1, or PE1 itself reboots, or the BGP process goes down or the connectivity between PE1 and an RR goes down, the effective number of PEs in the system now becomes N-1, and DFs are computed for all the VLANs that are configured on that ES. In general, if the DF for a VLAN V happens not to be PE1, but some other PE, say PE2, it is likely that some other PE (different from PE1 and PE2) will become the new DF. This is not desirable. Similarly, when a new PE hosts the same ES, the mapping again changes because of the modulus operation. This results in needless churn. Again referring to Figure 1, say V1, V2, and V3 are VLANs configured on ES2 with associated Ethernet Tags of values 999, 1000, and 1001, respectively. So, PE1, PE2, and PE3 are the DFs for V1, V2, and V3, respectively. Now when PE3 goes down, PE2 will become the DF for V1 and PE1 will become the DF for V2.

One point to note is that the default DF election algorithm assumes that all the PEs who are multihomed to the same ES (and interested in the DF election by exchanging EVPN routes) use an Originating Router's IP address [RFC7432] of the same family. This does not need to be the case, as the EVPN address family can be carried over an IPv4 or IPv6 peering, and the PEs attached to the same ES may use an address of either family.

Mathematically, a conventional hash function maps a key k to a number i representing one of m hash buckets through a function h(k), i.e., i = h(k). In the EVPN case, h is simply a modulo-m hash function viz. h(V) = V mod N, where N is the number of PEs that are multihomed to the ES in question. It is well known that for good hash distribution using the modulus operation, the modulus N should be a prime number not too close to a power of 2 [CLRS2009]. When the effective number of PEs changes from N to N-1 (or vice versa), all the objects (VLAN V) will be remapped except those for which V mod N and V mod (N-1) refer to the same PE in the previous and subsequent ordinal rankings, respectively. From a forwarding perspective, this is a churn, as it results in reprogramming the PE ports as either blocking or non-blocking at the PEs where the DF state changes.

This document addresses this problem and furnishes a solution to this undesirable behavior.

1.3.2. Traffic Black-Holing on Individual AC Failures

The default DF election algorithm defined by [RFC7432] takes into account only two variables in the modulus function for a given ES: the existence of the PE's IP address in the candidate list and the locally provisioned Ethernet Tags.

If the DF for an <ESI, EVI> fails (due to physical link/node failures), an ES route withdrawal will make the NDF PEs re-elect the DF for that <ESI, EVI> and the service will be recovered.

However, the default DF election procedure does not provide protection against "logical" failures or human errors that may occur at the service level on the DF, while the list of active PEs for a given ES does not change. These failures may have an impact not only on the local PE where the issue happens but also on the rest of the PEs of the ES. Some examples of such logical failures are listed below:

(a) A given individual AC defined in an ES is accidentally shut down

or is not provisioned yet (hence, the ACS is DOWN), while the ES is operationally active (since the ES route is active).

(b) A given MAC-VRF with a defined ES is either shut down or not

provisioned yet, while the ES is operationally active (since the ES route is active). In this case, the ACS of all the ACs defined in that MAC-VRF is considered to be DOWN.

Neither (a) nor (b) will trigger the DF re-election on the remote multihomed PEs for a given ES, since the ACS is not taken into account in the DF election procedures. While the ACS is used as a DF election tiebreaker and trigger in Virtual Private LAN Service (VPLS) multihoming procedures [VPLS-MH], there is no procedure defined in the EVPN specification [RFC7432] to trigger the DF re-election based on the ACS change on the DF.

Figure 2 shows an example of logical AC failure.

                               +---+
                               |CE4|
                               +---+
                                 |
                            PE4  |
                           +-----+-----+
           +---------------|  +-----+  |---------------+
           |               |  | BD-1|  |               |
           |               +-----------+               |
           |                                           |
           |                   EVPN                    |
           |                                           |
           | PE1               PE2                PE3  |
           | (NDF)             (DF)               (NDF)|
       +-----------+       +-----------+       +-----------+
       |  | BD-1|  |       |  | BD-1|  |       |  | BD-1|  |
       |  +-----+  |-------|  +-----+  |-------|  +-----+  |
       +-----------+       +-----------+       +-----------+
              AC1\   ES12   /AC2  AC3\   ES23   /AC4
                  \        /          \        /
                   \      /            \      /
                    +----+              +----+
                    |CE12|              |CE23|
                    +----+              +----+

Figure 2: Default DF Election and Traffic Black-Holing

BD-1 is defined in PE1, PE2, PE3, and PE4. CE12 is a multihomed CE connected to ES12 in PE1 and PE2. Similarly, CE23 is multihomed to PE2 and PE3 using ES23. Both CE12 and CE23 are connected to BD-1 through VLAN-based service interfaces: CE12-VID 1 (VID 1 on CE12) is associated with AC1 and AC2 in BD-1, whereas CE23-VID 1 is associated with AC3 and AC4 in BD-1. Assume that, although not represented, there are other ACs defined on these ESes mapped to different BDs.

After executing the default DF election algorithm as described in [RFC7432], PE2 turns out to be the DF for ES12 and ES23 in BD-1. The following issues may arise:

(a) If AC2 is accidentally shut down or is not configured yet, CE12

traffic will be impacted. In the case of All-Active multihoming, the BUM traffic to CE12 will be "black-holed", whereas for Single-Active multihoming, all the traffic to/from CE12 will be discarded. This is because a logical failure in PE2's AC2 may not trigger an ES route withdrawal for ES12 (since there are still other ACs active on ES12); therefore, PE1 will not rerun the DF election procedures.

(b) If the bridge table for BD-1 is administratively shut down or is

not configured yet on PE2, CE12 and CE23 will both be impacted: BUM traffic to both CEs will be discarded in the case of All-Active multihoming, and all traffic will be discarded to/from the CEs in the case of Single-Active multihoming. This is because PE1 and PE3 will not rerun the DF election procedures and will keep assuming that PE2 is the DF.

   Quoting [RFC7432], "When an Ethernet tag is decommissioned on an
   Ethernet segment, then the PE MUST withdraw the Ethernet A-D per EVI
   route(s) announced for the <ESI, Ethernet tags> that are impacted by
   the decommissioning."  However, while this A-D per EVI route
   withdrawal is used at the remote PEs performing aliasing or backup
   procedures, it is not used to influence the DF election for the
   affected EVIs.

This document adds an optional modification of the DF election procedure so that the ACS may be taken into account as a variable in the DF election; therefore, EVPN can provide protection against logical failures.

1.4. The Need for Extending the Default DF Election in EVPN Services

Section 1.3 describes some of the issues that exist in the default DF election procedures. In order to address those issues, this document introduces a new DF election framework. This framework allows the PEs to agree on a common DF election algorithm, as well as the capabilities to enable during the DF election procedure. Generally, "DF election algorithm" refers to the algorithm by which a number of input parameters are used to determine the DF PE, while "DF election capability" refers to an additional feature that can be used prior to the invocation of the DF election algorithm, such as modifying the inputs (or list of candidate PEs).

Within this framework, this document defines a new DF election algorithm and a new capability that can influence the DF election result:

  • The new DF election algorithm is referred to as "Highest Random Weight" (HRW). The HRW procedures are described in Section 3.
  • The new DF election capability is referred to as "AC-Influenced DF election" (AC-DF). The AC-DF procedures are described in Section 4.
  • HRW and AC-DF mechanisms are independent of each other. Therefore, a PE may support either HRW or AC-DF independently or may support both of them together. A PE may also support the AC-DF capability along with the default DF election algorithm per [RFC7432].

In addition, this document defines a way to indicate the support of HRW and/or AC-DF along with the EVPN ES routes advertised for a given ES. Refer to Section 2.2 for more details.

2. Designated Forwarder Election Protocol and BGP Extensions

This section describes the BGP extensions required to support the new DF election procedures. In addition, since the EVPN specification [RFC7432] leaves several questions open as to the precise FSM behavior of the DF election, Section 2.1 precisely describes the intended behavior.

2.1. The DF Election Finite State Machine (FSM)

Per [RFC7432], the FSM shown in Figure 3 is executed per <ES, VLAN> in the case of VLAN-based service or <ES, [VLANs in VLAN Bundle]> in the case of a VLAN Bundle on each participating PE. Note that the FSM is conceptual. Any design or implementation MUST comply with behavior that is equivalent to the behavior outlined in this FSM.

                     VLAN_CHANGE                VLAN_CHANGE
                     RCVD_ES                    RCVD_ES
                     LOST_ES                    LOST_ES
                     +----+                     +-------+
                     |    |                     |       v
                     |  +-+----+   ES_UP       ++-------++
                     +->+ INIT +-------------->+ DF_WAIT |
                        ++-----+               +-------+-+
                         ^                             |
     +-----------+       |                             |DF_TIMER
     | ANY_STATE +-------+         VLAN_CHANGE         |
     +-----------+ ES_DOWN    +-----------------+      |
                              |    RCVD_ES      v      v
                     +--------++   LOST_ES     ++------+-+
                     | DF_DONE +<--------------+ DF_CALC +<-+
                     +---------+   CALCULATED  +-------+-+  |
                                                       |    |
                                                       +----+
                                                       VLAN_CHANGE
                                                       RCVD_ES
                                                       LOST_ES

Figure 3: DF Election Finite State Machine

Observe that each EVI is locally configured on each of the multihomed PEs attached to a given ES and that the FSM does not provide any protection against inconsistent configuration between these PEs. That is, for a given EVI, one or more of the PEs are inadvertently configured with a different set of VLANs for a VLAN-aware Bundle service or with different VLANs for a VLAN-based service.

The states and events shown in Figure 3 are defined as follows.

States:

  1. INIT: Initial state.
  1. DF_WAIT: State in which the participant waits for enough information to perform the DF election for the EVI/ESI/VLAN combination.
  1. DF_CALC: State in which the new DF is recomputed.
  1. DF_DONE: State in which the corresponding DF for the EVI/ESI/VLAN combination has been elected.
  1. ANY_STATE: Refers to any of the above states.

Events:

  1. ES_UP: The ES has been locally configured as "UP".
  1. ES_DOWN: The ES has been locally configured as "DOWN".
  1. VLAN_CHANGE: The VLANs configured in a bundle (that uses the ES) changed. This event is necessary for VLAN Bundles only.
   4.  DF_TIMER: DF timer [RFC7432] (referred to as "Wait timer" in this
   
       document) has expired.
  1. RCVD_ES: A new or changed ES route is received in an Update message with an MP_REACH_NLRI. Receiving an unchanged Update MUST NOT trigger this event.
  1. LOST_ES: An Update message with an MP_UNREACH_NLRI for a previously received ES route has been received. If such a message is seen for a route that has not been advertised previously, the event MUST NOT be triggered.
  1. CALCULATED: DF has been successfully calculated.

Corresponding actions when transitions are performed or states are entered/exited:

   1.   ANY_STATE on ES_DOWN:
        (i) Stop the DF Wait timer.
        (ii) Assume an NDF for the local PE.
   
   2.   INIT on ES_UP: Transition to DF_WAIT.
   
   3.   INIT on VLAN_CHANGE, RCVD_ES, or LOST_ES: Do nothing.
   
   4.   DF_WAIT on entering the state:
        (i) Start the DF Wait timer if not started already or expired.
        (ii) Assume an NDF for the local PE.
   
   5.   DF_WAIT on VLAN_CHANGE, RCVD_ES, or LOST_ES: Do nothing.
   
   6.   DF_WAIT on DF_TIMER: Transition to DF_CALC.
   
   7.   DF_CALC on entering or re-entering the state:
        (i) Rebuild the candidate list, perform a hash, and perform the
        election.
        (ii) Afterwards, the FSM generates a CALCULATED event against
        itself.
   
   8.   DF_CALC on VLAN_CHANGE, RCVD_ES, or LOST_ES: Do as prescribed in
        Transition 7.
   
   9.   DF_CALC on CALCULATED: Mark the election result for the VLAN or
        bundle, and transition to DF_DONE.
  1. DF_DONE on exiting the state: If a new DF election is triggered and the current DF is lost, then assume an NDF for the local PE for the VLAN or VLAN Bundle.
   11.  DF_DONE on VLAN_CHANGE, RCVD_ES, or LOST_ES: Transition to
        DF_CALC.

The above events and transitions are defined for the default DF election algorithm. As described in Section 4, the use of the AC-DF capability introduces additional events and transitions.

2.2. The DF Election Extended Community

For the DF election procedures to be consistent and unanimous, it is necessary that all the participating PEs agree on the DF election algorithm and capabilities to be used. For instance, it is not possible for some PEs to continue to use the default DF election algorithm while some PEs use HRW. For brownfield deployments and for interoperability with legacy PEs, it is important that all PEs have the ability to fall back on the default DF election. A PE can indicate its willingness to support HRW and/or AC-DF by signaling a DF Election Extended Community along with the ES route (Route Type 4).

The DF Election Extended Community is a new BGP transitive Extended Community attribute [RFC4360] that is defined to identify the DF election procedure to be used for the ES. Figure 4 shows the encoding of the DF Election Extended Community.

      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 = 0x06   | Sub-Type(0x06)| RSV |  DF Alg |    Bitmap     ~
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     ~     Bitmap    |            Reserved                           |
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

Figure 4: DF Election Extended Community

Where:

  • Type: 0x06, as registered with IANA (Section 7) for EVPN Extended Communities.
   o  Sub-Type: 0x06.  "DF Election Extended Community", as registered
   
      with IANA.
  • RSV/Reserved: Reserved bits for information that is specific to DF Alg.
  • DF Alg (5 bits): Encodes the DF election algorithm values (between 0 and 31) that the advertising PE desires to use for the ES. This document creates an IANA registry called "DF Alg" (Section 7), which contains the following values:
      -  Type 0: Default DF election algorithm, or modulus-based
         algorithm as defined in [RFC7432].
  • Type 1: HRW Algorithm (Section 3).
  • Types 2-30: Unassigned.
  • Type 31: Reserved for Experimental Use.
  • Bitmap (2 octets): Encodes "capabilities" to use with the DF election algorithm in the DF Alg field. This document creates an IANA registry (Section 7) for the Bitmap field, with values 0-15. This registry is called "DF Election Capabilities" and includes the bit values listed below.
                              1 1 1 1 1 1
          0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5
         +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
         | |A|                           |
         +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

Figure 5: Bitmap Field in the DF Election Extended Community

  • Bit 0 (corresponds to Bit 24 of the DF Election Extended Community): Unassigned.
      -  Bit 1: AC-DF Capability (AC-Influenced DF election; see
         Section 4).  When set to 1, it indicates the desire to use
         AC-DF with the rest of the PEs in the ES.
  • Bits 2-15: Unassigned.

The DF Election Extended Community is used as follows:

  • A PE SHOULD attach the DF Election Extended Community to any advertised ES route, and the Extended Community MUST be sent if the ES is locally configured with a DF election algorithm other than the default DF election algorithm or if a capability is required to be used. In the Extended Community, the PE indicates the desired "DF Alg" algorithm and "Bitmap" capabilities to be used for the ES.
  • Only one DF Election Extended Community can be sent along with an ES route. Note that the intent is not for the advertising PE to indicate all the supported DF election algorithms and capabilities but to signal the preferred one.
  • DF Alg values 0 and 1 can both be used with Bit 1 (AC-DF) set to 0 or 1.
  • In general, a specific DF Alg SHOULD determine the use of the reserved bits in the Extended Community, which may be used in a different way for a different DF Alg. In particular, for DF Alg values 0 and 1, the reserved bits are not set by the advertising PE and SHOULD be ignored by the receiving PE.
  • When a PE receives the ES routes from all the other PEs for the ES in question, it checks to see if all the advertisements have the Extended Community with the same DF Alg and Bitmap:
  • If they do, this particular PE MUST follow the procedures for the advertised DF Alg and capabilities. For instance, if all ES routes for a given ES indicate DF Alg HRW and AC-DF set to 1, then the PEs attached to the ES will perform the DF election as per the HRW algorithm and following the AC-DF procedures.
  • Otherwise, if even a single advertisement for Route Type 4 is received without the locally configured DF Alg and capability, the default DF election algorithm MUST be used as prescribed in [RFC7432]. This procedure handles the case where participating PEs in the ES disagree about the DF algorithm and capability to be applied.
  • The absence of the DF Election Extended Community or the presence of multiple DF Election Extended Communities (in the same route) MUST be interpreted by a receiving PE as an indication of the default DF election algorithm on the sending PE -- that is, DF Alg 0 and no DF election capabilities.
  • When all the PEs in an ES advertise DF Type 31, they will rely on the local policy to decide how to proceed with the DF election.
  • For any new capability defined in the future, the applicability/ compatibility of this new capability to/with the existing DF Alg values must be assessed on a case-by-case basis.
  • Likewise, for any new DF Alg defined in the future, its applicability/compatibility to/with the existing capabilities must be assessed on a case-by-case basis.

2.2.1. Backward Compatibility

Implementations that comply with [RFC7432] only (i.e., implementations that predate this specification) will not advertise the DF Election Extended Community. That means that all other participating PEs in the ES will not receive DF preferences and will revert to the default DF election algorithm without AC-DF.

Similarly, an implementation that complies with [RFC7432] only and that receives a DF Election Extended Community will ignore it and will continue to use the default DF election algorithm.

3. The Highest Random Weight DF Election Algorithm

The procedure discussed in this section is applicable to the DF election in EVPN services [RFC7432] and the EVPN Virtual Private Wire Service (VPWS) [RFC8214].

HRW as defined in [HRW1999] is originally proposed in the context of Internet caching and proxy server load balancing. Given an object name and a set of servers, HRW maps a request to a server using the object-name (object-id) and server-name (server-id) rather than the server states. HRW forms a hash out of the server-id and the object-id and forms an ordered list of the servers for the particular object-id. The server for which the hash value is highest serves as the primary server responsible for that particular object, and the server with the next-highest value in that hash serves as the backup server. HRW always maps a given object name to the same server within a given cluster; consequently, it can be used at client sites to achieve global consensus on object-to-server mappings. When that server goes down, the backup server becomes the responsible designate.

Choosing an appropriate hash function that is statistically oblivious to the key distribution and imparts a good uniform distribution of the hash output is an important aspect of the algorithm. Fortunately, many such hash functions exist. [HRW1999] provides pseudorandom functions based on the Unix utilities rand and srand and easily constructed XOR functions that satisfy the desired hashing properties. HRW already finds use in multicast and ECMP [RFC2991] [RFC2992].

3.1. HRW and Consistent Hashing

HRW is not the only algorithm that addresses the object-to-server mapping problem with goals of fair load distribution, redundancy, and fast access. There is another family of algorithms that also addresses this problem; these fall under the umbrella of the Consistent Hashing Algorithms [CHASH]. These will not be considered here.

3.2. HRW Algorithm for EVPN DF Election

This section describes the application of HRW to DF election. Let DF(V) denote the DF and BDF(V) denote the BDF for the Ethernet Tag V; Si is the IP address of PE i; Es is the ESI; and Weight is a function of V, Si, and Es.

Note that while the DF election algorithm provided in [RFC7432] uses a PE address and VLAN as inputs, this document uses an Ethernet Tag, PE address, and ESI as inputs. This is because if the same set of PEs are multihomed to the same set of ESes, then the DF election algorithm used in [RFC7432] would result in the same PE being elected DF for the same set of BDs on each ES; this could have adverse side effects on both load balancing and redundancy. Including an ESI in the DF election algorithm introduces additional entropy, which significantly reduces the probability of the same PE being elected DF for the same set of BDs on each ES. Therefore, when using the HRW algorithm for EVPN DF election, the ESI value in the Weight function below SHOULD be set to that of the corresponding ES.

In the case of a VLAN Bundle service, V denotes the lowest VLAN, similar to the "lowest VLAN in bundle" logic of [RFC7432].

  1. DF(V) = Si| Weight(V, Es, Si) >= Weight(V, Es, Sj), for all j. In the case of a tie, choose the PE whose IP address is numerically the least. Note that 0 <= i,j < number of PEs in the redundancy group.
   2.  BDF(V) = Sk| Weight(V, Es, Si) >= Weight(V, Es, Sk), and
   
       Weight(V, Es, Sk) >= Weight(V, Es, Sj).  In the case of a tie,
       choose the PE whose IP address is numerically the least.

Where:

  • DF(V) is defined to be the address Si (index i) for which Weight(V, Es, Si) is the highest; 0 <= i < N-1.
  • BDF(V) is defined as that PE with address Sk for which the computed Weight is the next highest after the Weight of the DF. j is the running index from 0 to N-1; i and k are selected values.

Since the Weight is a pseudorandom function with the domain as the three-tuple (V, Es, S), it is an efficient and deterministic algorithm that is independent of the Ethernet Tag V sample space distribution. Choosing a good hash function for the pseudorandom function is an important consideration for this algorithm to perform better than the default algorithm. As mentioned previously, such functions are described in [HRW1999]. We take as a candidate hash function the first one out of the two that are listed as preferred in [HRW1999]:

      Wrand(V, Es, Si) = (1103515245((1103515245.Si+12345) XOR
      D(V, Es))+12345)(mod 2^31)

Here, D(V, Es) is the 31-bit digest (CRC-32 and discarding the most significant bit (MSB), as noted in [HRW1999]) of the 14-octet stream (the 4-octet Ethernet Tag V followed by the 10-octet ESI). It is mandated that the 14-octet stream be formed by the concatenation of the Ethernet Tag and the ESI in network byte order. The CRC should proceed as if the stream is in network byte order (big-endian). Si is the address of the ith server. The server's IP address length does not matter, as only the low-order 31 bits are modulo significant.

A point to note is that the Weight function takes into consideration the combination of the Ethernet Tag, the ES, and the PE IP address, and the actual length of the server IP address (whether IPv4 or IPv6) is not really relevant. The default algorithm defined in [RFC7432] cannot employ both IPv4 and IPv6 PE addresses, since [RFC7432] does not specify how to decide on the ordering (the ordinal list) when both IPv4 and IPv6 PEs are present.

HRW solves the disadvantages pointed out in Section 1.3.1 of this document and ensures that:

  • With very high probability, the task of DF election for the VLANs configured on an ES is more or less equally distributed among the PEs, even in the case of two PEs (see the first fundamental problem listed in Section 1.3.1).
   o  If a PE that is not the DF or the BDF for that VLAN goes down or
   
      its connection to the ES goes down, it does not result in a DF or
      BDF reassignment.  This saves computation, especially in the case
      when the connection flaps.
  • More importantly, it avoids the third fundamental problem listed in Section 1.3.1 (needless disruption) that is inherent in the existing default DF election.
  • In addition to the DF, the algorithm also furnishes the BDF, which would be the DF if the current DF fails.

4. The AC-Influenced DF Election Capability

The procedure discussed in this section is applicable to the DF election in EVPN services [RFC7432] and EVPN VPWS [RFC8214].

The AC-DF capability is expected to be generally applicable to any future DF algorithm. It modifies the DF election procedures by removing from consideration any candidate PE in the ES that cannot forward traffic on the AC that belongs to the BD. This section is applicable to VLAN-based and VLAN Bundle service interfaces. Section 4.1 describes the procedures for VLAN-aware Bundle service interfaces.

In particular, when used with the default DF algorithm, the AC-DF capability modifies Step 3 in the DF election procedure described in [RFC7432], Section 8.5, as follows:

  1. When the timer expires, each PE builds an ordered candidate list of the IP addresses of all the PE nodes attached to the ES (including itself), in increasing numeric value. The candidate list is based on the Originating Router's IP addresses of the ES routes but excludes any PE from whom no Ethernet A-D per ES route has been received or from whom the route has been withdrawn. Afterwards, the DF election algorithm is applied on a per <ES, Ethernet Tag>; however, the IP address for a PE will not be considered to be a candidate for a given <ES, Ethernet Tag> until the corresponding Ethernet A-D per EVI route has been received from that PE. In other words, the ACS on the ES for a given PE must be UP so that the PE is considered to be a candidate for a given BD.

If the default DF algorithm is used, every PE in the resulting candidate list is then given an ordinal indicating its position in the ordered list, starting with 0 as the ordinal for the PE with the numerically lowest IP address. The ordinals are used to determine which PE node will be the DF for a given Ethernet Tag on the ES, using the following rule:

Assuming a redundancy group of N PE nodes, for VLAN-based service, the PE with ordinal i is the DF for an <ES, Ethernet Tag V> when (V mod N) = i. In the case of a VLAN (-aware) Bundle service, then the numerically lowest VLAN value in that bundle on that ES MUST be used in the modulo function as the Ethernet Tag.

It should be noted that using the Originating Router's IP Address field [RFC7432] in the ES route to get the PE IP address needed for the ordered list allows for a CE to be multihomed across different Autonomous Systems (ASes) if such a need ever arises.

The modified Step 3, above, differs from [RFC7432], Section 8.5, Step 3 in two ways:

  • Any DF Alg can be used -- not only the described modulus-based DF Alg (referred to as the default DF election or "DF Alg 0" in this document).
  • The candidate list is pruned based upon non-receipt of Ethernet A-D routes: a PE's IP address MUST be removed from the ES candidate list if its Ethernet A-D per ES route is withdrawn. A PE's IP address MUST NOT be considered to be a candidate DF for an <ES, Ethernet Tag> if its Ethernet A-D per EVI route for the <ES, Ethernet Tag> is withdrawn.

The following example illustrates the AC-DF behavior applied to the default DF election algorithm, assuming the network in Figure 2:

(a) When PE1 and PE2 discover ES12, they advertise an ES route for

ES12 with the associated ES-Import Extended Community and the DF Election Extended Community indicating AC-DF = 1; they start a DF Wait timer (independently). Likewise, PE2 and PE3 advertise an ES route for ES23 with AC-DF = 1 and start a DF Wait timer.

(b) PE1 and PE2 advertise an Ethernet A-D per ES route for ES12.

PE2 and PE3 advertise an Ethernet A-D per ES route for ES23.

(c) In addition, PE1, PE2, and PE3 advertise an Ethernet A-D per EVI

route for AC1, AC2, AC3, and AC4 as soon as the ACs are enabled. Note that the AC can be associated with a single customer VID (e.g., VLAN-based service interfaces) or a bundle of customer VIDs (e.g., VLAN Bundle service interfaces).

(d) When the timer expires, each PE builds an ordered candidate list

of the IP addresses of all the PE nodes attached to the ES (including itself) as explained in the modified Step 3 above. Any PE from which an Ethernet A-D per ES route has not been received is pruned from the list.

(e) When electing the DF for a given BD, a PE will not be considered

to be a candidate until an Ethernet A-D per EVI route has been received from that PE. In other words, the ACS on the ES for a given PE must be UP so that the PE is considered to be a candidate for a given BD. For example, PE1 will not consider PE2 as a candidate for DF election for <ES12, VLAN-1> until an Ethernet A-D per EVI route is received from PE2 for <ES12, VLAN-1>.

(f) Once the PEs with ACS = DOWN for a given BD have been removed

from the candidate list, the DF election can be applied for the remaining N candidates.

Note that this procedure only modifies the existing EVPN control plane by adding and processing the DF Election Extended Community and by pruning the candidate list of PEs that take part in the DF election.

In addition to the events defined in the FSM in Section 2.1, the following events SHALL modify the candidate PE list and trigger the DF re-election in a PE for a given <ES, Ethernet Tag>. In the FSM shown in Figure 3, the events below MUST trigger a transition from DF_DONE to DF_CALC:

  1. Local AC going DOWN/UP.
  1. Reception of a new Ethernet A-D per EVI route update/withdrawal for the <ES, Ethernet Tag>.
  1. Reception of a new Ethernet A-D per ES route update/withdrawal for the ES.

4.1. AC-Influenced DF Election Capability for VLAN-Aware Bundle

Services

The procedure described in Section 4 works for VLAN-based and VLAN Bundle service interfaces because, for those service types, a PE advertises only one Ethernet A-D per EVI route per <ES, VLAN> or <ES, VLAN Bundle>. In Section 4, an Ethernet Tag represents a given VLAN or VLAN Bundle for the purpose of DF election. The withdrawal of such a route means that the PE cannot forward traffic on that particular <ES, VLAN> or <ES, VLAN Bundle>; therefore, the PE can be removed from consideration for DF election.

According to [RFC7432], in VLAN-aware Bundle services, the PE advertises multiple Ethernet A-D per EVI routes per <ES, VLAN Bundle> (one route per Ethernet Tag), while the DF election is still performed per <ES, VLAN Bundle>. The withdrawal of an individual route only indicates the unavailability of a specific AC and not necessarily all the ACs in the <ES, VLAN Bundle>.

This document modifies the DF election for VLAN-aware Bundle services in the following ways:

  • After confirming that all the PEs in the ES advertise the AC-DF capability, a PE will perform a DF election per <ES, VLAN>, as opposed to per <ES, VLAN Bundle> as described in [RFC7432]. Now, the withdrawal of an Ethernet A-D per EVI route for a VLAN will indicate that the advertising PE's ACS is DOWN and the rest of the PEs in the ES can remove the PE from consideration for DF election in the <ES, VLAN>.
  • The PEs will now follow the procedures in Section 4.

For example, assuming three bridge tables in PE1 for the same MAC-VRF (each one associated with a different Ethernet Tag, e.g., VLAN-1, VLAN-2, and VLAN-3), PE1 will advertise three Ethernet A-D per EVI routes for ES12. Each of the three routes will indicate the status of each of the three ACs in ES12. PE1 will be considered to be a valid candidate PE for DF election in <ES12, VLAN-1>, <ES12, VLAN-2>, and <ES12, VLAN-3> as long as its three routes are active. For instance, if PE1 withdraws the Ethernet A-D per EVI routes for <ES12, VLAN-1>, the PEs in ES12 will not consider PE1 as a suitable DF candidate for <ES12, VLAN-1>. PE1 will still be considered for <ES12, VLAN-2> and <ES12, VLAN-3>, since its routes are active.

5. Solution Benefits

The solution described in this document provides the following benefits:

(a) It extends the DF election as defined in [RFC7432] to address

the unfair load balancing and potential black-holing issues with the default DF election algorithm. The solution is applicable to the DF election in EVPN services [RFC7432] and EVPN VPWS [RFC8214].

(b) It defines a way to signal the DF election algorithm and

capabilities intended by the advertising PE. This is done by defining the DF Election Extended Community, which allows the advertising PE to indicate its support for the capabilities defined in this document as well as any subsequently defined DF election algorithms or capabilities.

(c) It is backwards compatible with the procedures defined in

[RFC7432]. If one or more PEs in the ES do not support the new procedures, they will all follow DF election as defined in [RFC7432].

6. Security Considerations

This document addresses some identified issues in the DF election procedures described in [RFC7432] by defining a new DF election framework. In general, this framework allows the PEs that are part of the same ES to exchange additional information and agree on the DF election type and capabilities to be used.

By following the procedures in this document, the operator will minimize such undesirable situations as unfair load balancing, service disruption, and traffic black-holing. Because such situations could be purposely created by a malicious user with access to the configuration of one PE, this document also enhances the security of the network. Note that the network will not benefit from the new procedures if the DF election algorithm is not consistently configured on all the PEs in the ES (if there is no unanimity among all the PEs, the DF election algorithm falls back to the default DF election as provided in [RFC7432]). This behavior could be exploited by an attacker that manages to modify the configuration of one PE in the ES so that the DF election algorithm and capabilities in all the PEs in the ES fall back to the default DF election. If that is the case, the PEs will be exposed to the unfair load balancing, service disruption, and black-holing mentioned earlier.

In addition, the new framework is extensible and allows for new security enhancements in the future. Note that such enhancements are out of scope for this document. Finally, since this document extends the procedures in [RFC7432], the same security considerations as those described in [RFC7432] are valid for this document.

7. IANA Considerations

IANA has:

  • Allocated Sub-Type value 0x06 in the "EVPN Extended Community Sub-Types" registry defined in [RFC7153] as follows:
      Sub-Type Value    Name                             Reference
      --------------    ------------------------------   -------------
      0x06              DF Election Extended Community   This document
  • Set up a registry called "DF Alg" for the DF Alg field in the Extended Community. New registrations will be made through the "RFC Required" procedure defined in [RFC8126]. Value 31 is for experimental use and does not require any other RFC than this document. The following initial values in that registry exist:
      Alg         Name                               Reference
      ----        -----------------------------      -------------
      0           Default DF Election                This document
      1           HRW Algorithm                      This document
      2-30        Unassigned
      31          Reserved for Experimental Use      This document
  • Set up a registry called "DF Election Capabilities" for the 2-octet Bitmap field in the Extended Community. New registrations will be made through the "RFC Required" procedure defined in [RFC8126]. The following initial value in that registry exists:
      Bit         Name                             Reference
      ----        ----------------                 -------------
      0           Unassigned
      1           AC-DF Capability                 This document
      2-15        Unassigned

8. References

8.1. Normative References

   [RFC7432]  Sajassi, A., Ed., Aggarwal, R., Bitar, N., Isaac, A.,
              Uttaro, J., Drake, J., and W. Henderickx, "BGP MPLS-Based
              Ethernet VPN", RFC 7432, DOI 10.17487/RFC7432,
              February 2015, <https://www.rfc-editor.org/info/rfc7432>.
   
   [RFC8214]  Boutros, S., Sajassi, A., Salam, S., Drake, J., and J.
              Rabadan, "Virtual Private Wire Service Support in Ethernet
              VPN", RFC 8214, DOI 10.17487/RFC8214, August 2017,
              <https://www.rfc-editor.org/info/rfc8214>.
   
   [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>.
   
   [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>.
   
   [RFC4360]  Sangli, S., Tappan, D., and Y. Rekhter, "BGP Extended
              Communities Attribute", RFC 4360, DOI 10.17487/RFC4360,
              February 2006, <https://www.rfc-editor.org/info/rfc4360>.
   
   [RFC7153]  Rosen, E. and Y. Rekhter, "IANA Registries for BGP
              Extended Communities", RFC 7153, DOI 10.17487/RFC7153,
              March 2014, <https://www.rfc-editor.org/info/rfc7153>.
   
   [RFC8126]  Cotton, M., Leiba, B., and T. Narten, "Guidelines for
              Writing an IANA Considerations Section in RFCs", BCP 26,
              RFC 8126, DOI 10.17487/RFC8126, June 2017,
              <https://www.rfc-editor.org/info/rfc8126>.

8.2. Informative References

   [VPLS-MH]  Kothari, B., Kompella, K., Henderickx, W., Balus, F., and
              J. Uttaro, "BGP based Multi-homing in Virtual Private LAN
              Service", Work in Progress,
              draft-ietf-bess-vpls-multihoming-03, March 2019.
   
   [CHASH]    Karger, D., Lehman, E., Leighton, T., Panigrahy, R.,
              Levine, M., and D. Lewin, "Consistent Hashing and Random
              Trees: Distributed Caching Protocols for Relieving Hot
              Spots on the World Wide Web", ACM Symposium on Theory of
              Computing, ACM Press, New York, DOI 10.1145/258533.258660,
              May 1997.
   
   [CLRS2009] Cormen, T., Leiserson, C., Rivest, R., and C. Stein,
              "Introduction to Algorithms (3rd Edition)", MIT
              Press, ISBN 0-262-03384-8, 2009.
   
   [RFC2991]  Thaler, D. and C. Hopps, "Multipath Issues in Unicast and
              Multicast Next-Hop Selection", RFC 2991,
              DOI 10.17487/RFC2991, November 2000,
              <https://www.rfc-editor.org/info/rfc2991>.
   
   [RFC2992]  Hopps, C., "Analysis of an Equal-Cost Multi-Path
              Algorithm", RFC 2992, DOI 10.17487/RFC2992, November 2000,
              <https://www.rfc-editor.org/info/rfc2992>.
   
   [RFC4456]  Bates, T., Chen, E., and R. Chandra, "BGP Route
              Reflection: An Alternative to Full Mesh Internal BGP
              (IBGP)", RFC 4456, DOI 10.17487/RFC4456, April 2006,
              <https://www.rfc-editor.org/info/rfc4456>.
   
   [HRW1999]  Thaler, D. and C. Ravishankar, "Using Name-Based Mappings
              to Increase Hit Rates", IEEE/ACM Transactions on
              Networking, Volume 6, No. 1, February 1998,
              <https://www.microsoft.com/en-us/research/wp-content/
              uploads/2017/02/HRW98.pdf>.
   
   [Knuth]    Knuth, D., "The Art of Computer Programming: Volume 3:
              Sorting and Searching", 2nd Edition, Addison-Wesley,
              Page 516, 1998.

Acknowledgments

The authors want to thank Ranganathan Boovaraghavan, Sami Boutros, Luc Andre Burdet, Anoop Ghanwani, Mrinmoy Ghosh, Jakob Heitz, Leo Mermelstein, Mankamana Mishra, Tamas Mondal, Laxmi Padakanti, Samir Thoria, and Sriram Venkateswaran for their review and contributions. Special thanks to Stephane Litkowski for his thorough review and detailed contributions.

They would also like to thank their working group chairs, Matthew Bocci and Stephane Litkowski, and their AD, Martin Vigoureux, for their guidance and support.

Finally, they would like to thank the Directorate reviewers and the ADs for their thorough reviews and probing questions, the answers to which have substantially improved the quality of the document.

Contributors

The following people have contributed substantially to this document and should be considered coauthors:

   Antoni Przygienda
   Juniper Networks, Inc.
   1194 N. Mathilda Ave.
   Sunnyvale, CA  94089
   United States of America

Email:

          prz@juniper.net

Vinod Prabhu
Nokia

   Email: vinod.prabhu@nokia.com

Wim Henderickx
Nokia

   Email: wim.henderickx@nokia.com

Wen Lin
Juniper Networks, Inc.

Email:

          wlin@juniper.net

Patrice Brissette
Cisco Systems

Email:

          pbrisset@cisco.com

Keyur Patel
Arrcus, Inc.

Email:

          keyur@arrcus.com

Autumn Liu
Ciena

Email:

          hliu@ciena.com

Authors' Addresses

   Jorge Rabadan (editor)
   Nokia
   777 E. Middlefield Road
   Mountain View, CA  94043
   United States of America
   
   Email: jorge.rabadan@nokia.com
   
   Satya Mohanty (editor)
   Cisco Systems, Inc.
   225 West Tasman Drive
   San Jose, CA  95134
   United States of America

Email:

          satyamoh@cisco.com
   
   Ali Sajassi
   Cisco Systems, Inc.
   225 West Tasman Drive
   San Jose, CA  95134
   United States of America

Email:

          sajassi@cisco.com
   John Drake
   Juniper Networks, Inc.
   1194 N. Mathilda Ave.
   Sunnyvale, CA  94089
   United States of America

Email:

          jdrake@juniper.net
   
   Kiran Nagaraj
   Nokia
   701 E. Middlefield Road
   Mountain View, CA  94043
   United States of America
   
   Email: kiran.nagaraj@nokia.com
   
   Senthil Sathappan
   Nokia
   701 E. Middlefield Road
   Mountain View, CA  94043
   United States of America
   
   Email: senthil.sathappan@nokia.com