wdiff rfc8361.original rfc8361.txt

TRILL Working Group
Internet Engineering Task Force (IETF) W. Hao
INTERNET-DRAFT
Request for Comments: 8361 Y. Li
Intended Status: Standards Track Huawei Technologies
Updates: 6325 Huawei Technologies
Category: Standards Track M. Durrani
ISSN: 2070-1721 Equinix
S. Gupta
IP Infusion
A. Qu
MediaTec
Expires: July 2018 January 25,
April 2018

Transparent Interconnection of Lots of Links (TRILL):
Centralized Replication for Active-Active BUM
Broadcast, Unknown Unicast, and Multicast (BUM) Traffic
draft-ietf-trill-centralized-replication-13.txt

Abstract

In TRILL Transparent Interconnection of Lots of Links (TRILL) active-active
access, an RPF (Reverse a Reverse Path Forwarding) Forwarding (RPF) check failure issue may occur
when using the pseudo-nickname mechanism specified in RFC 7781. This draft
document describes a solution to resolve this RPF check failure issue
through centralized replication. All ingress RBridges Routing Bridges
(RBridges) send BUM (Broadcast, Broadcast, Unknown unicast Unicast, and Multicast) Multicast (BUM)
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 the TRILL base protocol RFC 6325. (RFC 6325).
To avoid RPF check failure on a an 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. This
document updates RFC 6325.

Status of this 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 an Internet Standards Track document.

This document is a product of the Internet Engineering Task Force (IETF), its areas, and its working groups. Note that
other groups may also distribute working documents as Internet-
Drafts.

Internet-Drafts are draft documents valid for a maximum
(IETF). It represents the consensus of six
months the IETF community. It has
received public review and may be updated, replaced, or obsoleted has been approved for publication by other
documents at any time. It the
Internet Engineering Steering Group (IESG). Further information on
Internet Standards is inappropriate to use Internet-Drafts
as reference material or to cite them other than as "work available in
progress."

The list Section 2 of RFC 7841.

Information about the current Internet-Drafts can be accessed at
http://www.ietf.org/1id-abstracts.html

The list status of Internet-Draft Shadow Directories can this document, any errata,
and how to provide feedback on it may be accessed obtained at
http://www.ietf.org/shadow.html.
https://www.rfc-editor.org/info/rfc8361.

This document is subject to BCP 78 and the IETF Trust's Legal
Provisions Relating to IETF Documents
(http://trustee.ietf.org/license-info)
(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 . . . . . . . . . . . . . . . . . . . . . . . . 2
2. Conventions used Used in this document............................4 This Document . . . . . . . . . . . . . . 3
3. Centralized replication solution overview....................5 Replication Solution Overview . . . . . . . . . . 4
4. Frame duplication Duplication from remote RBridge........................6 Remote RBridge . . . . . . . . . . . . 6
5. Local forwarding behavior Forwarding Behavior on ingress RBridge.................7 Ingress RBridge . . . . . . . . 6
6. Loop prevention Prevention among RBridges in an edge group..............8 Edge Group . . . . . . . 8
7. Centralized replication forwarding process...................9 Replication Forwarding Process . . . . . . . . . 8
8. BUM traffic load balancing Traffic Load-Balancing among multiple centralized nodes..11 Multiple Centralized Nodes . 10
9. Co-existing Coexisting with the CMT solution........................... 12 Solution (RFC 7783) . . . . . . . . . 11
10. Network upgrade analysis................................... 13 Upgrade Analysis . . . . . . . . . . . . . . . . . . 12
11. TRILL protocol extensions.................................. 13 Protocol Extensions . . . . . . . . . . . . . . . . . . 12
11.1. "R" and "C" Flag in the Nickname Flags APPsub-TLV......13 APPsub-TLV . . . 13
12. Security Considerations.................................... Considerations . . . . . . . . . . . . . . . . . . . 14
13. IANA Considerations........................................ 15 Considerations . . . . . . . . . . . . . . . . . . . . . 14
14. References ................................................ 15 . . . . . . . . . . . . . . . . . . . . . . . . . 14
14.1. Normative References.................................. 15 References . . . . . . . . . . . . . . . . . . 14
14.2. Informative References................................ 16
15. References . . . . . . . . . . . . . . . . . 15
Acknowledgments ........................................... . . . . . . . . . . . . . . . . . . . . . . . . . 16
Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 16

1. Introduction

The IETF TRILL (Transparent Interconnection of Lots of Links)
[RFC6325] protocol [RFC6325] provides multipath data forwarding
that is loop free and per hop 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 TRILL-specific header for user data.

Customer Equipment (CE) devices can be multi-homed multihomed 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]) (LAALP) [RFC7379] such as Multi-
Chassis Link Aggregation (MC-LAG) or Distributed Resilient Network
Interconnect (DRNI) [802.1AX]. [IEEE802.1AX]. An active-active flow-based load load-
sharing mechanism can achieve better load balancing load-balancing and high
reliability. A CE device can be a layer Layer 3 (L3) end system by itself
or a bridge switch through which layer 3 L3 end systems access to the TRILL
campus.

In active-active access, the pseudo-nickname solution in [RFC7781]
can be used to avoid MAC Media Access Control (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 the TRILL campus. The use of the nickname solves the
address flip-flop issue by setting the MAC address learnt nickname learned by a remote
RBridge to be the pseudo-
nickname. 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 were
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 Then, the RPF (Reverse Reverse Path Forwarding) Forwarding
(RPF) check required by TRILL [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) 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. This document updates [RFC6325]; under [RFC6325] multi-destination per [RFC6325], multi-
destination traffic is ingressed to a multi-destination TRILL Data packet
but under data
packet. However, per this document, when using the centralized
replication feature, multi-destination traffic is initially ingressed
to a unicast TRILL Data data packet.

2. Conventions used Used in this document 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 abbreviations and terminology in [RFC6325] are used herein with
the following additions:

BUM -

BUM: Broadcast, Unknown unicast, and Multicast

CE - As in [RFC7783],

CE: Customer Equipment (as in [RFC7783]), as relates to a
device (end station or bridge). The device can be
either physical or virtual equipment.

Data Label - Label: VLAN or FGL Fine-Grained Labeled (FGL) [RFC7172]

DF -

DF: Designated Forwarder [RFC7781]

FGL - Fine Grained

FGL: Fine-Grained Label [RFC7172].

LAALP - [RFC7172]

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

MAC flip flop - 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 -

MC-LAG: Multi-Chassis Link Aggregation.

RPF - Aggregation

RPF: Reverse Path Forwarding. Forwarding

3. Centralized replication solution overview 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 high-capacity core RBridges with many high bandwidth
adjacencies and this high-bandwidth
adjacencies. This constraint makes its it 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 that 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. "R-nickname".

When the centralized RBridge receives a unicast TRILL encapsulated TRILL-encapsulated
packet with its R-nickname as the egress nickname, it decapsulates
the packet. Then 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 the TRILL base protocol. protocol [RFC6325].
It MUST use a distribution tree whose tree root is the centralized
RBridge itself. (An RBridge may by be 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 were a locally ingressed frame
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 a BUM packet should be the centralized node facing port centralized-node-facing port, as the
multicast traffic always comes from the centralized node in this
solution. However However, the RPF port as the result of distribution tree
calculation as specified in [RFC6325] will be the real ingress
RBridge facing port
RBridge-facing port, as it uses the edge group's virtual RBridge as
the ingress RBridge, so the RPF check will fail.

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, the RPF check will accept
traffic on the centralized node facing 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 from [RFC6325] 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
root, which is stored in the TRILL Header "egress nickname" field.)
When determining the RFC RPF 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 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 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 campuses through the TRILL LSP Link State PDU (LSP) extension
specified in Section 11.

4. Frame duplication Duplication from remote Remote RBridge

Frame duplication may occur when a remote host sends a multi-
destination frame to a local CE which that has an active-active connection
to the TRILL campus. To avoid the local CE receiving multiple copies
from a remote RBridge, the designated forwarder Designated Forwarder (DF) mechanism is
supported for egress direction 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 using DF is to elect one RBridge per VLAN from an edge group
to be responsible for egressing the BUM traffic. [RFC7781] describes
the DF election mechanism among member RBridges involving involved in an edge
group.

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

5. Local forwarding behavior Forwarding Behavior on ingress 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 encapsulation; 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 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 the 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 the 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.

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 "centralized replication local forwarding behavior A. 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 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 ports
associated with different pseudo-nicknames.

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

6. Loop prevention Prevention among RBridges in an edge group Edge Group

If a CE sends a broadcast, unknown unicast, or multicast (BUM) 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
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 case, the
frame will loop back to the CE and the traffic split-
horizon 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 the
same active-active group as the packet's incoming port to the TRILL
campus.

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 the ingress RBridge itself, through a TRILL distribution
tree. If the same pseudo-nickname is used for two active-active
access CEs as the 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 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 Replication Forwarding Process

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

| | |
-------- | --------
| | |
+------+ +------+ +------+
|(RB1) | |(RB2) | | (RB3)|
+------+ +------+ +------+
* | * | * | ^
* | * | * | ^
* ----------*-------------*-- ^
***************************** | ^
* | ^
LAALP1 * LAALP2 | ^
+------+ +------+ +------+
| CE1 | | CE2 | | CE3 |
+------+ +------+ +------+

Figure 1: TRILL Active-Active Access

Note: The asterisk line, hyphen & vertical bar line, and circumflex
line in this figure indicate the connection of the various CEs to the
various RBs.

Figure 1 TRILL Active-active access

Assuming the centralized replication solution is used in the example
network of above figure 1, Figure 1: RB5 is the distribution tree root and
centralized replication node, node; CE1 and CE2 are active-active accessed
to RB1, RB2, and RB3 through LAALP1 and LAALP2 respectively, LAALP2, respectively; and CE3
is
single homed single-homed to RB3. The RBridge's own nicknames of RB1 to RB5
are nick1 to nick5 nick5, respectively. RB1, RB2, and RB3 use the same pseudo-
nickname
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 data packet. Then Then, it uses a
distribution tree for which it is the root to forward the packet
as a multi-destination TRILL Data 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 would not be re-encapsulate. re-
encapsulated. If it is sent to an R-nickname that is not a tree
root, it will either be will not be forwarded at all or, if it is re-encapsulated 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 The incoming
traffic port is the up port facing the distribution tree root, 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 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 a 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 local LAALPs connecting CE1 and CE2, they also
don't forward the traffic to local CE1 and CE2.

8. BUM traffic load balancing Traffic Load-Balancing among multiple centralized nodes Multiple Centralized Nodes

To support unicast TRILL encapsulation BUM traffic load balancing, 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 campus, the VLAN-based (or FGL-based [RFC7172]) load load-
balancing algorithm used by an ingress active-active access RBridge
is as follows:

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

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

For examples, example, there are 3 three R-Nicknames (RN). (RNs). The RNs will be
ordered RN0 to RN2. Assuming there are 5 five VLANs from VLAN ID 1 ID1 to ID 5
ID5 spreading among edge RBridges, the traffic in VLAN 1 VLAN1 will go to
RN1,
VLAN 2 VLAN2 will go to RN2, and so on.

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

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

+------+ +------+
|(RB6) | |(RB7) |
+------+ +------+
------------------|-----------|----------------------
| | | | |
+------+ +------+ +------+ +------+ +------+
|(RB1) | |(RB2) | |(RB3) | |(RB4) | |(RB5) |
+------+ +------+ +------+ +------+ +------+
| | | | |
------------ -------------------------
| |
+------+ +------+
| CE1 | | CE2 |
+------+ +------+

Figure 2 2: CMT and centralized replication co-existing scenario Centralized Replication Coexisting Scenario
Both the centralized replication solution and the Coordinated
Multicast Trees (CMT) [RFC7783] solution from [RFC7783] rely on using
pseudo-nicknames pseudo-
nicknames to avoid MAC flip-flop on remote RBridges. These two
solutions can co-exist coexist 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, 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 the Nickname Flags APPsub-TLV APPsub-
TLV with C-flag C flag set. A nickname with the C-flag C flag set is called a "C-
nickname".
"C-nickname". A transit RBridge will perform the centralized
replication specific
replication-specific RPF check algorithm if it receives TRILL Data data
packets with a C-nickname as the ingress nickname.

An edge group using CMT [RFC7783] MUST NOT set the C-nickname C 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 an 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 would, thus, reach
few (possibly none) of its intended targets. To avoid confusion, a
pseudo-nickname MUST NOT be shared between a centralized replication
edge group and a CMT-based edge group.

10. Network upgrade analysis 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 the distribution tree.

Active-active connection edge RBridges will typically need software
and hardware upgrade upgrades 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 Protocol Extensions

Two new flags, "R" and "C", are specified in the Nickname Flags
APPsub-TLV [RFC7780]. A nickname with the "R" R flag set is called an
R-nickname
"R-nickname" and a nickname with the the "C" C flag set is called a C-
nickname.
"C-nickname". The R-nickname is a specialized nickname attached to a
centralized node to differentiate unicast TRILL encapsulated 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, nickname; that is, the R-nickname flag is
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 the 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 active-
active access. If the R 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 the C flag is zero,
that nickname will not be used for that purpose.

It is possible, due

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

12. Security Considerations

This draft document does not introduce any extra security risks. A rogue
RBridge that is a tree root can attract traffic by advertising an R-
nickname.
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 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.

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

13. IANA Considerations

IANA is requested to assign has assigned 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 "NickFlags
Bits" subregistry of the TRILL
Parameters page "Transparent Interconnection of Lots of
Links (TRILL) Parameters" registry as follows:

Bit Mnemonic Description Reference
--- -------- -------------------- -----------
3
2 R Replication Nickname [This document]
4 [RFC8361]
3 C Special RFC RPF Check [This document] [RFC8361]

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>.
<https://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>. <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>.
<https://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>.
<https://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>.
<https://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 (TRILL) Use of IS-IS", RFC 7176,
DOI 10.17487/RFC7176, May 2014,
<http://www.rfc-editor.org/info/rfc7176>.
<https://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>.
<https://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>.
<https://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>. <https://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.
<https://www.rfc-editor.org/info/rfc7783>.

[IEEE802.1AX]
IEEE, "IEEE Standard for Local and metropolitan area
networks -- Link Aggregation", IEEE 802.1AX,
DOI 10.1109/IEEESTD.2017.7888436, March 2017,
<http://ieeexplore.ieee.org/document/7888436/>.

Acknowledgments

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

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