Skip to main content

Centralized Replication for Active-Active BUM Traffic
draft-ietf-trill-centralized-replication-12

The information below is for an old version of the document.
Document Type
This is an older version of an Internet-Draft that was ultimately published as RFC 8361.
Authors Hao Weiguo , Yizhou Li , Muhammad Durrani , Sujay Gupta , Andrew Qu
Last updated 2018-01-11 (Latest revision 2018-01-04)
Replaces draft-hao-trill-centralized-replication
RFC stream Internet Engineering Task Force (IETF)
Formats
Reviews
Additional resources Mailing list discussion
Stream WG state Submitted to IESG for Publication
Document shepherd Donald E. Eastlake 3rd
Shepherd write-up Show Last changed 2017-11-11
IESG IESG state Became RFC 8361 (Proposed Standard)
Consensus boilerplate Yes
Telechat date (None)
Responsible AD Alia Atlas
Send notices to (None)
IANA IANA review state IANA OK - Actions Needed
draft-ietf-trill-centralized-replication-12
TRILL Working Group                                              W. Hao
INTERNET-DRAFT                                                    Y. Li
Intended Status: Standards Track                    Huawei Technologies
                                                             M. Durrani
                                                                Equinix
                                                               S. Gupta
                                                            IP Infusion
                                                                  A. Qu
                                                               MediaTec
Expires: July 2018                                     January 05, 2018

           Centralized Replication for Active-Active BUM Traffic
              draft-ietf-trill-centralized-replication-12.txt

Abstract

   In TRILL active-active access, an RPF (Reverse Path Forwarding) check
   failure issue may occur when using the pseudo-nickname mechanism
   specified in RFC 7781. This draft describes a solution to resolve
   this RPF check failure issue through centralized replication. All
   ingress RBridges send BUM (Broadcast, Unknown unicast and Multicast)
   traffic to a centralized node with unicast TRILL encapsulation.
   When the centralized node receives the BUM traffic, it decapsulates
   the packets and forwards them to their destination RBridges using
   a distribution tree established as per TRILL base protocol RFC 6325.
   To avoid RPF check failure on a RBridge sitting between the ingress
   RBridge and the centralized replication node, some change in the
   RPF calculation algorithm is required. RPF checks on each RBridge
   MUST be calculated as if the centralized node was the ingress
   RBridge, instead of being calculated using the actual ingress
   RBridge.

Status of this Memo

   This Internet-Draft is submitted to IETF in full conformance with
   the provisions of BCP 78 and BCP 79.

   Internet-Drafts are working documents of the Internet Engineering
   Task Force (IETF), its areas, and its working groups.  Note that

Hao & Li, et al        Expires July 05, 2018                  [Page 1]
Internet-Draft Centralized replication for BUM traffic    January 2018

   other groups may also distribute working documents as Internet-
   Drafts.

   Internet-Drafts are draft documents valid for a maximum of six
   months   and may be updated, replaced, or obsoleted by other
   documents at any time.  It is inappropriate to use Internet-Drafts
   as reference material or to cite them other than as "work in
   progress."

   The list of current Internet-Drafts can be accessed at
   http://www.ietf.org/1id-abstracts.html

   The list of Internet-Draft Shadow Directories can be accessed at
   http://www.ietf.org/shadow.html.

Copyright Notice

   Copyright (c) 2018 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
   (http://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
   2. Conventions used in this document............................4
   3. Centralized replication solution overview....................5
   4. Frame duplication from remote RBridge........................6
   5. Local forwarding behavior on ingress RBridge.................7
   6. Loop prevention among RBridges in an edge group..............8
   7. Centralized replication forwarding process...................9
   8. BUM traffic load balancing among multiple centralized nodes..11
   9. Co-existing with the CMT solution........................... 12

Hao & Li,etc            Expires July 05, 2018                 [Page 2]
Internet-Draft Centralized replication for BUM traffic    January 2018

   10. Network upgrade analysis................................... 13
   11. TRILL protocol extensions.................................. 13
      11.1. "R" and "C" Flag in the Nickname Flags APPsub-TLV......13
   12. Security Considerations.................................... 14
   13. IANA Considerations........................................ 15
   14. References ................................................ 15
      14.1. Normative References.................................. 15
      14.2. Informative References................................ 16
   15. Acknowledgments ........................................... 16

1. Introduction

   The IETF TRILL (Transparent Interconnection of Lots of Links)
   [RFC6325] protocol provides loop free and per hop based multipath
   data forwarding with minimum configuration. TRILL uses IS-IS
   [RFC6165] [RFC7176] as its control plane routing protocol and
   defines a TRILL specific header for user data.

   Customer Equipment (CE) devices can be multi-homed to a set of edge
   RBridges forming an edge group where active-active service can be
   provided. In that case, all of the uplinks from a CE are handled via
   a Local Active-Active Link Protocol (LAALP [RFC7379]) such as Multi-
   Chassis Link Aggregation (MC-LAG) or Distributed Resilient Network
   Interconnect (DRNI) [802.1AX]. An active-active flow-based load
   sharing mechanism  can achieve better load balancing and high
   reliability. A CE device can be a layer 3 end system by itself or a
   bridge switch through which layer 3 end systems access to TRILL
   campus.

   In active-active access, the pseudo-nickname solution in [RFC7781]
   can be used to avoid MAC flip-flop on remote RBridges. The basic
   idea is to use a virtual RBridge (RBv) with a single pseudo-nickname
   to represent an edge group. Any member RBridge of that edge group
   uses this pseudo-nickname rather than its own nickname as the
   ingress nickname when it injects TRILL data frames to TRILL campus.
   The use of the nickname solves the address flip-flop issue by
   setting the MAC address learnt by a remote RBridge to the pseudo-
   nickname. However, it introduces another issue of incorrect packet
   dropping as follows: When a pseudo-nickname is used by an edge
   RBridge as the ingress nickname to forward BUM (Broadcast, Unknown
   unicast and Multicast) traffic, any RBridges (RBn) sitting between
   the ingress RBridge and the distribution tree root will treat the
   traffic as if it was ingressed from the virtual RBridge RBv. If the
   same distribution tree is used by different edge RBridges of the
   same RBv, the traffic may arrive at some RBn from different ports.
   Then the RPF (Reverse Path Forwarding) check required by TRILL

Hao & Li,etc            Expires July 05, 2018                 [Page 3]
Internet-Draft Centralized replication for BUM traffic    January 2018

   [RFC6325] fails, and the BUM traffic received on unexpected ports
   will be dropped by RBn.

   This document specifies a centralized replication solution for
   broadcast, unknown unicast and multicast (BUM) traffic forwarding to
   resolve the issue of incorrect packet drop caused by the RPF check
   failure in the virtual RBridge case. The basic idea is that all
   ingress RBridges send BUM traffic to a centralized node, which
   MUST be a distribution tree root, using unicast TRILL
   encapsulation. When the centralized node receives the packets, it
   decapsulates and forwards them to their destination RBridges using a
   distribution tree established as per the TRILL base protocol.

2. Conventions used in this document

   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.
   
   The acronyms and terminology in [RFC6325] are used herein with the
   following additions:

      BUM - Broadcast, Unknown unicast, and Multicast

      CE - As in [RFC7783], Customer Equipment device (end station or
   bridge). The device can be either physical or virtual equipment.

      Data Label - VLAN or FGL [RFC7172]

      DF - Designated Forwarder [RFC7781]

      FGL - Fine Grained Label [RFC7172].

      LAALP - Local Active-Active Link Protocol [RFC7379].

      MAC flip flop - A problem where the attachment point of a MAC
   address appears to a remote switch to keep changing. See Section 3.3
   of [RFC7379].

      MC-LAG - Multi-Chassis Link Aggregation.

      RPF - Reverse Path Forwarding.

Hao & Li,etc            Expires July 05, 2018                 [Page 4]
Internet-Draft Centralized replication for BUM traffic    January 2018

3. Centralized replication solution overview

   When an edge RBridge receives BUM traffic from a CE device, it uses
   unicast TRILL encapsulation instead of multicast encapsulation to
   send the packets to a centralized node. The centralized node MUST be
   a distribution tree root. Distribution tree roots are normally
   chosen to be high capacity core RBridges with many high bandwidth
   adjacencies and this constraint makes its practical, as described
   below, to support centralized replication with only software changes
   to transit RBridges.

   The TRILL header of the unicast TRILL encapsulation contains an
   "ingress RBridge nickname" field and an "egress RBridge nickname"
   field [RFC6325]. If the ingress RBridge receives the BUM packet from
   a port that is in an active-active edge group using [RFC7781], it
   sets the ingress RBridge nickname to be the pseudo-nickname rather
   than its own nickname to avoid MAC flip-flop (see Section 3.3 of
   [RFC7379]) on remote RBridges. The egress RBridge nickname is set
   to a special nickname of the centralized node which is used to
   differentiate the centralized replication purpose unicast TRILL
   encapsulation from a normal unicast TRILL encapsulation. This
   special nickname is called an R-nickname.

   When the centralized RBridge receives a unicast TRILL encapsulated
   packet with its R-nickname as egress nickname, it decapsulates the
   packet. Then the centralized RBridge replicates and forwards the
   BUM packet to the packet's destination RBridges using one of the
   distribution trees established as per TRILL base protocol. It MUST
   use a distribution tree whose tree root is the centralized RBridge
   itself. (An RBridge may by the root of more than one tree.) When the
   centralized RBridge forwards the BUM traffic, it simply sends it on
   the distribution tree as if it was a locally ingressed frame except
   that the ingress nickname remains the same as that in the packet it
   received to ensure that the MAC address learning by all egress
   RBridges is bound to the pseudo-nickname.

   When the replicated packet is forwarded by each RBridge along the
   distribution tree starting from the centralized node, an RPF check
   is performed as per [RFC6325]. For any RBridge sitting between the
   ingress RBridge and the centralized replication node, the incoming
   port of such BUM packet should be the centralized node facing port
   as the multicast traffic always comes from the centralized node in
   this solution. However the RPF port as the result of distribution
   tree calculation as specified in [RFC6325] will be the real ingress
   RBridge facing port as it uses the edge group's virtual RBridge as
   the ingress RBridge, so the RPF check will fail.

Hao & Li,etc            Expires July 05, 2018                 [Page 5]
Internet-Draft Centralized replication for BUM traffic    January 2018

   To solve this problem, some change in the RPF test is required. In
   this case, the RPF calculation on each RBridge should use the
   centralized node as the ingress RBridge for each tree for which that
   node is the root instead of the real ingress virtual RBridge to
   perform the calculation. As a result, RPF check will accept traffic
   on the centralized node facing port of the RBridge for multi-
   destination traffic. This prevents incorrect frame drops by the RPF
   check.

   The change in the actual RPF check on a received multi-destination
   TRILL data packet is easy. The [RFC6325] RPF check is a check to see
   if a triple of {ingress nickname, tree, receiving RBridge port} is
   allowed. (The tree is indicated by the nickname of its root which is
   stored in the TRILL Header "egress nickname" field.)  When
   determining the RFC check, if "ingress nickname" is using centralized
   replication (indicated by a C-nickname, see Section 9), then the check
   is based on distribution from the tree root. If "ingress nickname"
   is not using centralized replication, then the check is based on
   distribution from the RBridge having the ingress nickname.

   To differentiate the centralized replication unicast TRILL
   encapsulation from normal unicast TRILL encapsulation, the R-
   nickname is introduced for centralized replication. When the
   centralized node receives unicast TRILL encapsulation traffic with
   the egress nickname R-nickname, it decapsulates the packet and then
   forwards the packet to the destination RBridges through a
   distribution tree for which it is the root by re-encapsulation as
   aforementioned. In TRILL, RBridges can hold multiple nicknames so
   the centralized RBridge simply obtains another nickname to use as
   the R-nickname. The centralized RBridge or RBridges should announce
   their R-nickname to all TRILL campus through the TRILL LSP extension
   specified in Section 11.

4. Frame duplication from remote RBridge

   Frame duplication may occur when a remote host sends a multi-
   destination frame to a local CE which has an active-active
   connection to the TRILL campus. To avoid local CE receiving multiple
   copies from a remote RBridge, the designated forwarder (DF)
   mechanism is supported for egress direction multicast traffic.

   The DF election mechanism [RFC7781] allows only one port of one
   RBridge in an active-active group to forward multicast traffic from
   the TRILL campus to the local access side for each VLAN. The basic
   idea of DF is to elect one RBridge per VLAN from an edge group to be

Hao & Li,etc            Expires July 05, 2018                 [Page 6]
Internet-Draft Centralized replication for BUM traffic    January 2018

   responsible for egressing the BUM traffic. [RFC7781] describes the
   DF election mechanism among member RBridges involving in an edge
   group.

   If the DF election mechanism is used for frame duplication
   prevention, access ports on an RBridge are categorized as one of
   three types: non-group, group DF port and group non-DF port. The
   last two types can be called group ports. Each of the group ports
   is associated with a pseudo-nickname. If consistent nickname
   allocation to edge group RBridges is used, it is possible that same
   pseudo-nickname is associated with more than one port on a single
   RBridge. A typical scenario is that CE1 is connected to RB1 & RB2
   by LAALP1 while CE2 is connected to RB1 & RB2 by LAALP2.
   In order to conserve the number of pseudo-nicknames used, member
   ports for both LAALP1 and LAALP2 on RB1 & RB2 are all associated
   with the same pseudo-nickname.

5. Local forwarding behavior on ingress RBridge

   When an ingress RBridge (RB1) receives BUM traffic from a local
   active-active connected CE (CE1) device, the traffic will be
   injected into the TRILL campus with TRILL encapsulation, and it will
   be replicated and forwarded to all destination RBridges through
   central replication, including the ingress RBridge itself, along a
   TRILL distribution tree. To avoid the traffic looping back to the
   original sender CE, an ingress nickname of the CE group's pseudo-
   nickname is used for traffic filtering.

   However, if there are two CEs, say CE1 and CE2, connecting to the
   ingress RB1 and each associated with the same pseudo-nickname, RB1
   needs to locally replicate and forward to CE2, because another copy
   of the BUM traffic between CE1 and CE2 through TRILL campus will be
   blocked by the traffic filtering.

   If CE1 and CE2 are not associated with the same pseudo-nickname, the
   copy of the BUM traffic between CE1 and CE2 through TRILL campus
   won't be blocked by the traffic filtering. To avoid duplicated
   traffic on receiver CE, there cannot be local replicated BUM traffic
   between these two CEs on ingress RB1.

   In summary, to ensure correct BUM traffic forwarding behavior for
   each CE, the local replication behavior on the ingress RBridge is as
   follows:

      1. Replicate to the active-active group ports associated with the
   same pseudo-nickname as that associated with the incoming port.

Hao & Li,etc            Expires July 05, 2018                 [Page 7]
Internet-Draft Centralized replication for BUM traffic    January 2018

      2. Do not replicate to active-active group ports associated with
   other pseudo-nicknames.

      3. Do not replicate to non-edge-group ports.

   The above local forwarding behavior on the ingress RBridge of RB1
   can be called centralized replication local forwarding behavior A.

   If ingress RBridge RB1 itself is the centralized replication node,
   BUM traffic injected by RB1 into the TRILL campus won't loop back to
   RB1. In this case, the local forwarding behavior is called
   centralized replication local forwarding behavior B. Behavior B on
   RB1 is as follows:

      1. Local replication to the ports associated with the same
   pseudo-nickname as that associated with the incoming port.

      2. Local replication to the group DF port associated with
   different pseudo-nicknames. Do not replicate to group non-DF port
   associated with different pseudo-nicknames.

      3. Local replication to non-edge-group ports.

6. Loop prevention among RBridges in an edge group

   If a CE sends a broadcast, unknown unicast, or multicast (BUM)
   packet through a DF port to an ingress RBridge, that RBridge will
   forward that packet to all or a subset of the other RBridges that
   only have non-DF ports for that active-active group. Because BUM
   traffic forwarding to non-DF ports isn't allowed, in this case the
   frame won't loop back to the CE.

   If a CE sends a BUM packet through a non-DF port to an ingress
   RBridge, say RB1, then RB1 will forward that packet to other
   RBridges that have a DF port for that active-active group. In this
   case the frame will loop back to the CE and the traffic split-
   horizon filtering mechanism is used to avoid looping back among
   RBridges in the edge group.

   This split-horizon mechanism relies on the ingress nickname field in
   the TRILL header to check if a packet's egress port belongs to a
   same active-active group as the packet's incoming port to the TRILL
   campus.

Hao & Li,etc            Expires July 05, 2018                 [Page 8]
Internet-Draft Centralized replication for BUM traffic    January 2018

   When the ingress RBridge receives BUM traffic from an active-active
   connected CE device, the traffic will be sent through the TRILL
   campus with TRILL encapsulation to a centralized RBridge. There it
   will be replicated and forwarded to its destination RBridges, which
   include ingress RBridge itself, through a TRILL distribution tree.
   If the same pseudo-nickname is used for two active-active access CEs
   as ingress nickname, an egress RBridge can use that nickname to
   filter traffic forwarding to all local CEs. In this case, the
   traffic between these two CEs goes through the local RBridge and
   another copy of the traffic from the TRILL campus is filtered. If
   different ingress nicknames are used for two connecting CE devices,
   the access ports connecting to these two CEs should be isolated from
   each other. The BUM traffic between these two CEs should go through
   the TRILL campus, otherwise the destination CE connected to same
   RBridge with the sender CE will receive two copies of the traffic.

7. Centralized replication forwarding process

                             +-----------+
                             |   (RB5)   |
                             +-----------+
                                   |
                             +-----------+
                             |   (RB4)   |
                             +-----------+

                              |     |    |
                      --------      |     --------
                     |              |             |
                   +------+      +------+      +------+
                   |(RB1) |      |(RB2) |      | (RB3)|
                   +------+      +------+      +------+
                     *   |         *  |          * |  ^
                     *   |         *  |          * |   ^
                     *   ----------*-------------*--    ^
                     ***************************** |     ^
              LAALP1 *                      LAALP2 |      ^
                 +------+                    +------+    +------+
                 |  CE1 |                    | CE2  |    | CE3  |
                 +------+                    +------+    +------+
             Figure 1  TRILL Active-active access

   Assuming the centralized replication solution is used in the example
   network of above figure 1, RB5 is the distribution tree root and

Hao & Li,etc            Expires July 05, 2018                 [Page 9]
Internet-Draft Centralized replication for BUM traffic    January 2018

   centralized replication node, CE1 and CE2 are active-active accessed
   to RB1, RB2, and RB3 through LAALP1 and LAALP2 respectively, CE3 is
   single homed to RB3. The RBridge's own nicknames of RB1 to RB5 are
   nick1 to nick5 respectively. RB1, RB2, and RB3 use the same pseudo-
   nickname for LAALP1 and LAALP2; that pseudo-nickname is P-nick. The
   R-nickname on the centralized replication node of RB5 is S-nick.

   The BUM traffic forwarding process from CE1 to CE2 and CE3 is as
   follows:

       1. CE1 sends BUM traffic to RB3.

       2. RB3 replicates and sends the BUM traffic to CE2 locally. RB2
   also sends the traffic to RB5 using unicast TRILL encapsulation. In
   the TRILL Header, the ingress nickname is set as P-nick and the
   egress nickname is set as S-nick.

       3. RB5 decapsulates the unicast TRILL Data packet. Then it uses a
   distribution tree for which it is the root to forward the packet as
   a multi-destination TRILL Data packet. The egress nickname in the
   multi-destination TRILL Header is the nick5 and the ingress nickname
   is still P-nick. If RB3 had sent the unicast to some nickname that
   was not an R-nickname, the packet will not be re-encapsulate. If it
   is sent to an R-nickname that is not a tree root, it will either be
   not forwarded at all or, if it is re-encapsulated and forwarded,
   will be subject to incorrect pruning and will not be delivered to
   all of its intended recipients.

      4. RB4 receives multicast TRILL traffic from RB5. Traffic
   incoming port is the up port facing the distribution tree root,
   RB4's RPF check will be correct based on the changed RPF port
   calculation algorithm in this document. After the RPF check is
   performed, it forwards the traffic to all other egress RBridges (RB1,
   RB2, and RB3).

      5. RB3 receives multicast TRILL traffic from RB4. It decapsulates
   the multi-destination TRILL Data packet. Because the ingress
   nickname of P-nick is equivalent to the nickname of local LAALPs
   connecting to CE1 and CE2, RB3 doesn't forward the traffic to CE1
   and CE2 to avoid duplicated frame. RB3 only forwards the packet to
   CE3.

      6. RB1 and RB2 receive multicast TRILL traffic from RB4. The
   forwarding process is similar to the process on RB3, i.e., because
   the ingress nickname of P-nick is equivalent to the nickname of the

Hao & Li,etc            Expires July 05, 2018                [Page 10]
Internet-Draft Centralized replication for BUM traffic    January 2018

   local LAALPs connecting CE1 and CE2, they also don't forward the
   traffic to local CE1 and CE2.

8. BUM traffic load balancing among multiple centralized nodes

   To support unicast TRILL encapsulation BUM traffic load balancing,
   multiple centralized replication nodes can be deployed and the
   traffic can be spread over these nodes based on data label (VLAN or
   FGL). Furthermore, if it was desirable for a centralized node to be
   sent more of this BUM traffic, it could hold two or more R-nicknames.
   The share of BUM traffic it would receive would be proportional to
   the number of R-nicknames it held.

   Assuming there are k different R-nicknames held by centralized nodes
   in a TRILL campus. The VLAN-based (or FGL-based [RFC7172]) load
   balancing algorithm used by ingress active-active access RBridge is
   as follows:

         1. All R-nicknames are ordered and numbered from 0 to k-1 in
   ascending order treating the nicknames as unsigned 16-bit integers.

         2. For data label ID m, choose the R-nickname whose number
   equals (m mod k) as egress nickname for BUM traffic unicast TRILL
   encapsulation.

   For examples, there are 3 R-Nicknames (RN). The RNs will be ordered
   RN0 to RN2. Assuming there are 5 VLANs from VLAN ID 1 to ID 5
   spreading among edge RBridges, the traffic in VLAN 1 will go to RN1,
   VLAN 2 will go to RN2, and so on.

   When an ingress RBridge participating in active-active connection
   receives BUM traffic from local CE, the RBridge decides which R-
   nickname to send the traffic to based on the VLAN-based load
   spreading algorithm, thus data-label-based load balancing for the
   BUM traffic can be achieved using multiple centralized
   nodes/multiple R-nicknames.

Hao & Li,etc            Expires July 05, 2018                [Page 11]
Internet-Draft Centralized replication for BUM traffic    January 2018

9. Co-existing with the CMT (RFC 7783) solution

                    +------+    +------+
                    |(RB6) |    |(RB7) |
                    +------+    +------+
      ------------------|-----------|----------------------
      |            |              |          |            |
   +------+    +------+       +------+    +------+     +------+
   |(RB1) |    |(RB2) |       |(RB3) |    |(RB4) |     |(RB5) |
   +------+    +------+       +------+    +------+     +------+
       |          |               |          |            |
       ------------               -------------------------
             |                               |
         +------+                         +------+
         |  CE1 |                         |  CE2 |
         +------+                         +------+
       Figure 2 CMT and centralized replication co-existing scenario

   Both the centralized replication solution and the Coordinated
   Multicast Trees (CMT) [RFC7783] solution rely on using
   pseudo-nicknames to avoid MAC flip-flop on remote RBridges. 
   These two solutions can co-exist in a single TRILL campus.
   Each solution can be selected by each active-active edge
   group of RBridges independently.

   As illustrated in Figure 2, RB1 and RB2 use CMT for CE1's active-
   active access, RB3, RB4, and RB5 use the centralized replication for
   CE2's active-active access.

   For the centralized replication solution, edge group RBridges MUST
   announce the local pseudo-nickname using Nickname Flags APPsub-TLV
   with C-flag set. A nickname with the C-flag set is called a "C-
   nickname". A transit RBridge will perform the centralized
   replication specific RPF check algorithm if it receives TRILL Data
   packets with a C-nickname as ingress nickname.

   An edge group using CMT [RFC7783] MUST NOT set the C-nickname flag
   on the pseudo-nickname it is using. This is already mandatory
   behavior because any RBridge originating a Nickname Flags APPsub-TLV
   is required by [RFC7780] to set any flag bit it does not know about
   to zero. If a edge RBridge using CMT [RFC7783] nevertheless set the
   C-bit for an edge group pseudo-nickname, it is very likely that BUM
   traffic encapsulated with that nickname as ingress would be
   incorrectly pruned early in its distribution and would thus reach
   few (possibly none) of its intended targets. To avoid confusion, a

Hao & Li,etc            Expires July 05, 2018                [Page 12]
Internet-Draft Centralized replication for BUM traffic    January 2018

   pseudo-nickname MUST NOT be shared between a centralized replication
   edge group and a CMT-based edge group.

10. Network upgrade analysis

   Centralized nodes will typically need software and hardware upgrades
   to support centralized replication, which stitches together the
   TRILL unicast traffic decapsulation process and the process of
   normal TRILL multicast traffic forwarding along distribution tree.

   Active-active connection edge RBridges will typically need software
   and hardware upgrade to support unicast TRILL encapsulation for BUM
   traffic; the process is similar to other head-end replication
   processes.

   Transit nodes typically need only a software upgrade to support the
   changed RPF port calculation algorithm.

11. TRILL protocol extensions

   Two new flags, "R" and "C", are specified in the Nickname Flags
   APPsub-TLV [RFC7780]. A nickname with the "R" flag set is called an
   R-nickname and a nickname with the the "C" flag set is called a C-
   nickname. The R-nickname is a specialized nickname attached to a
   centralized node to differentiate unicast TRILL encapsulated BUM
   traffic from normal unicast TRILL traffic. The C-nickname flag is
   set on the pseudo-nickname for each edge group that uses the
   centralized replication. A C-nickname is a specialized pseudo-
   nickname for which transit RBridges perform a different RPF check
   algorithm for TRILL data packets with the C-nickname in the ingress
   nickname field.

   When active-active edge RBridges use centralized replication to
   forward BUM traffic, the R-nickname is used as the egress nickname
   and the C-nickname is used as ingress nickname in the TRILL header
   for the unicast TRILL encapsulation of BUM traffic.

11.1. "R" and "C" Flag in the Nickname Flags APPsub-TLV

   If this APPsub-TLV is being advertised by an RBridge that does not
   have the nickname appearing in the Nickname Flags APPsub-TLV, the R
   and C flag bits in the APPsub-TLV MUST be treated as if they were
   zero. If an RBridge that is not a distribution tree root advertises
   an R-nickname, that nickname MUST NOT be treated as an R-nickname
   but rather as an ordinary nickname, that is, the R-nickname flag is

Hao & Li,etc            Expires July 05, 2018                [Page 13]
Internet-Draft Centralized replication for BUM traffic    January 2018

   ignored for all purposes if the nickname is held by an RBridge that
   is not a tree root.

              0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15
             +--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+
             |   Nickname                                    |
             +--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+
             |IN|SE|R | C|    RESV                           |
             +--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+
                             NICKFLAG RECORD

            o R = If R flag is one, it indicates that the advertising
   TRILL switch holding Nickname is a centralized replication node, and
   Nickname is used as egress nickname for edge group RBridges to
   inject BUM traffic into the TRILL campus when the edge group
   RBridges use this centralized replication solution for active-active
   access. If flag is zero, that nickname will not be used for that
   purpose.

            o C = If C flag is one, it indicates that the TRILL traffic
   with this nickname as an ingress nickname that requires the special
   RPF check algorithm specified in Section 3. If flag is zero, that
   nickname will not be used for that purpose.

   It is possible, due to errors or due to transient inconsistent LSPs
   when the link state database is converging after a configuration
   change or the like for there to be inconsistent Nickname Flags
   APPsub-TLVs for the same nickname. In this case it is RECOMMENDED
   that the nickname be treated as if the R / C flag was set if any
   Nickname Flags APPsub-TLV for that nickname has the R / C flag set.

12. Security Considerations

   This draft does not introduce any extra security risks. A rogue
   RBridge that is a tree root can attract traffic by advertising an R-
   nickname. However, this does not represent a substantial increase in
   risk as RBridges could cause problems in a number of other ways by
   advertising low cost adjacencies or making themselves the highest
   priority tree root or the like. In general, the protection against
   an untrusted device acting as an RBridge and wrecking havoc is to
   use IS-IS authentication [RFC5310] and configure and administer the
   TRILL campus so that only trusted RBridges have the authentication
   key.

Hao & Li,etc            Expires July 05, 2018                [Page 14]
Internet-Draft Centralized replication for BUM traffic    January 2018

   For general TRILL Security Considerations, see [RFC6325]. For
   Security Considerations related to pseudo-nickname active-active,
   see [RFC7781].

13. IANA Considerations

   IANA is requested to assign two bits in the Nickname Flags APPsubTLV
   flags for the R and C bits discussed in Section 11.1 [Bits 3 and 4
   suggested] and update the "NickFlags" Bits registry on the TRILL
   Parameters page as follows:

              Bit  Mnemonic   Description           Reference
             ---  --------  --------------------  -----------
               3    R        Replication Nickname  [This document]
               4    C        Special RFC Check     [This document]

14. References

14.1. Normative References

      [RFC2119]  Bradner, S., "Key words for use in RFCs to Indicate
   Requirement Levels", BCP 14, RFC 2119, DOI 10.17487/RFC2119, March
   1997, <http://www.rfc-editor.org/info/rfc2119>.

      [RFC5310]  Bhatia, M., Manral, V., Li, T., Atkinson, R., White,
   R., and M. Fanto, "IS-IS Generic Cryptographic Authentication", RFC
   5310, DOI 10.17487/RFC5310, February 2009, <https://www.rfc-
   editor.org/info/rfc5310>.

      [RFC6165]  Banerjee, A. and D. Ward, "Extensions to IS-IS for
   Layer-2 Systems", RFC 6165, DOI 10.17487/RFC6165, April 2011,
   <http://www.rfc-editor.org/info/rfc6165>.

      [RFC6325] Perlman, R., Eastlake 3rd, D., Dutt, D., Gai, S., and A.
   Ghanwani, "Routing Bridges (RBridges): Base Protocol Specification",
   RFC 6325, DOI 10.17487/RFC6325, July 2011, <http://www.rfc-
   editor.org/info/rfc6325>.

      [RFC7172] Eastlake 3rd, D., Zhang, M., Agarwal, P., Perlman, R.,
   and D. Dutt, "Transparent Interconnection of Lots of Links (TRILL):
   Fine-Grained Labeling", RFC 7172, DOI 10.17487/RFC7172, May 2014,
   <http://www.rfc-editor.org/info/rfc7172>.

      [RFC7176] Eastlake 3rd, D., Senevirathne, T., Ghanwani, A., Dutt,
   D., and A. Banerjee, "Transparent Interconnection of Lots of Links

Hao & Li,etc            Expires July 05, 2018                [Page 15]
Internet-Draft Centralized replication for BUM traffic    January 2018

   (TRILL) Use of IS-IS", RFC 7176, DOI 10.17487/RFC7176, May 2014,
   <http://www.rfc-editor.org/info/rfc7176>.

      [RFC7780] Eastlake 3rd, D., Zhang, M., Perlman, R., Banerjee, A.,
   Ghanwani, A., and S. Gupta, "Transparent Interconnection of Lots of
   Links (TRILL): Clarifications, Corrections, and Updates", RFC 7780,
   DOI 10.17487/RFC7780, February 2016, <http://www.rfc-
   editor.org/info/rfc7780>.
   
      [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>.
   

14.2. Informative References

      [RFC7781] Zhai, H., Senevirathne, T., Perlman, R., Zhang, M., and
   Y. Li, "Transparent Interconnection of Lots of Links (TRILL):
   Pseudo-Nickname for Active-Active Access", RFC 7781, DOI
   10.17487/RFC7781, February 2016, <http://www.rfc-
   editor.org/info/rfc7781>.

      [RFC7379] Li, Y., Hao, W., Perlman, R., Hudson, J., and H. Zhai,
   "Problem Statement and Goals for Active-Active Connection at the
   Transparent Interconnection of Lots of Links (TRILL) Edge", RFC 7379,
   DOI 10.17487/RFC7379, October 2014, <http://www.rfc-
   editor.org/info/rfc7379>.

      [RFC7783] Senevirathne, T., Pathangi, J., and J. Hudson,
   "Coordinated Multicast Trees (CMT) for Transparent Interconnection
   of Lots of Links (TRILL)", RFC 7783, DOI 10.17487/RFC7783, February
   2016, <http://www.rfc-editor.org/info/rfc7783>.

15. Acknowledgments

   The authors wish to acknowledge the important contributions of
   Donald Eastlake, Hongjun Zhai, Xiaomin Wu, Liang Xia, Francis Dupont.

Hao & Li,etc            Expires July 05, 2018                [Page 16]
Internet-Draft Centralized replication for BUM traffic    January 2018

Authors' Addresses

   Weiguo Hao
   Huawei Technologies
   101 Software Avenue,
   Nanjing 210012
   PR China
   Email: haoweiguo@huawei.com

   Yizhou Li
   Huawei Technologies
   101 Software Avenue,
   Nanjing 210012
   PR China
   Email: liyizhou@huawei.com

   Muhammad Durrani
   Equinix
   Email: mdurrani@equinix.com

   Sujay Gupta
   IP Infusion
   RMZ Centennial
   Mahadevapura Post
   Bangalore - 560048
   India
   Email: sujay.gupta@ipinfusion.com

   Andrew Qu
   MediaTec
   Email: laodulaodu@gmail.com

Hao & Li,etc            Expires July 05, 2018                [Page 17]