From 4bfd864f10b68b71482b35c818559068ef8d5797 Mon Sep 17 00:00:00 2001 From: Thomas Voss Date: Wed, 27 Nov 2024 20:54:24 +0100 Subject: doc: Add RFC documents --- doc/rfc/rfc8361.txt | 955 ++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 955 insertions(+) create mode 100644 doc/rfc/rfc8361.txt (limited to 'doc/rfc/rfc8361.txt') diff --git a/doc/rfc/rfc8361.txt b/doc/rfc/rfc8361.txt new file mode 100644 index 0000000..ae7be1d --- /dev/null +++ b/doc/rfc/rfc8361.txt @@ -0,0 +1,955 @@ + + + + + + +Internet Engineering Task Force (IETF) W. Hao +Request for Comments: 8361 Y. Li +Updates: 6325 Huawei Technologies +Category: Standards Track M. Durrani +ISSN: 2070-1721 Equinix + S. Gupta + IP Infusion + A. Qu + MediaTec + April 2018 + + + Transparent Interconnection of Lots of Links (TRILL): + Centralized Replication for Active-Active + Broadcast, Unknown Unicast, and Multicast (BUM) Traffic + +Abstract + + In Transparent Interconnection of Lots of Links (TRILL) active-active + access, a Reverse Path Forwarding (RPF) check failure issue may occur + when using the pseudo-nickname mechanism specified in RFC 7781. This + document describes a solution to resolve this RPF check failure issue + through centralized replication. All ingress Routing Bridges + (RBridges) send Broadcast, Unknown Unicast, and 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 per the TRILL base protocol (RFC 6325). + To avoid RPF check failure on 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 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/rfc8361. + + + +Hao, et al. Standards Track [Page 1] + +RFC 8361 Centralized Replication for BUM Traffic April 2018 + + +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 + (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 ....................................................2 + 2. Conventions Used in This Document ...............................3 + 3. Centralized Replication Solution Overview .......................4 + 4. Frame Duplication from Remote RBridge ...........................6 + 5. Local Forwarding Behavior on Ingress RBridge ....................6 + 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 ....10 + 9. Coexisting with the CMT Solution (RFC 7783) ....................11 + 10. Network Upgrade Analysis ......................................12 + 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 ...........................................14 + 14. References ....................................................15 + 14.1. Normative References .....................................15 + 14.2. Informative References ...................................16 + Acknowledgments ...................................................17 + Authors' Addresses ................................................17 + +1. Introduction + + The IETF TRILL protocol [RFC6325] provides multipath data forwarding + that is loop free and per-hop based 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 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] such as Multi- + + + +Hao, et al. Standards Track [Page 2] + +RFC 8361 Centralized Replication for BUM Traffic April 2018 + + + Chassis Link Aggregation (MC-LAG) or Distributed Resilient Network + Interconnect (DRNI) [IEEE802.1AX]. An active-active flow-based load- + sharing mechanism can achieve better load-balancing and high + reliability. A CE device can be a Layer 3 (L3) end system by itself + or a bridge switch through which L3 end systems access the TRILL + campus. + + In active-active access, the pseudo-nickname solution in [RFC7781] + can be used to avoid 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 nickname learned by a remote + RBridge to be 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 + traffic, any RBridges (RBn) sitting between the ingress RBridge and + the distribution tree root will treat the traffic as if it were + ingressed from the 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 Reverse Path 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 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]; per [RFC6325], multi- + destination traffic is ingressed to a multi-destination TRILL data + packet. However, per this document, when using the centralized + replication feature, multi-destination traffic is initially ingressed + to a unicast TRILL data packet. + +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. + + + + +Hao, et al. Standards Track [Page 3] + +RFC 8361 Centralized Replication for BUM Traffic April 2018 + + + The abbreviations and terminology in [RFC6325] are used herein with + the following additions: + + BUM: Broadcast, Unknown unicast, and Multicast + + 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: VLAN or Fine-Grained Labeled (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 + +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. This constraint makes 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 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". + + + + +Hao, et al. Standards Track [Page 4] + +RFC 8361 Centralized Replication for BUM Traffic April 2018 + + + When the centralized RBridge receives a unicast TRILL-encapsulated + packet with its R-nickname as the 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 per the TRILL base protocol [RFC6325]. + It MUST use a distribution tree whose tree root is the centralized + RBridge itself. (An RBridge may 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 were 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 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, 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. + + 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 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 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, which is stored in the TRILL Header "egress nickname" field.) + When determining the 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 + is introduced for centralized replication. When the centralized node + + + +Hao, et al. Standards Track [Page 5] + +RFC 8361 Centralized Replication for BUM Traffic April 2018 + + + 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 campuses through the TRILL Link State PDU (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 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 (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 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 involved 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 the same 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, + 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 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; it will be replicated + and forwarded to all destination RBridges through central + replication, including the ingress RBridge itself, along a TRILL + + + + + +Hao, et al. Standards Track [Page 6] + +RFC 8361 Centralized Replication for BUM Traffic April 2018 + + + 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 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 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 ports + associated with different pseudo-nicknames. + + 3. Local replication to non-edge-group ports. + + + + + + +Hao, et al. Standards Track [Page 7] + +RFC 8361 Centralized Replication for BUM Traffic April 2018 + + +6. Loop Prevention among RBridges in an Edge Group + + If a CE sends a 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 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, the destination CE + connected to same RBridge with the sender CE will receive two copies + of the traffic. + + + + + + + + + + + + + + + +Hao, et al. Standards Track [Page 8] + +RFC 8361 Centralized Replication for BUM Traffic April 2018 + + +7. Centralized 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. + + Assuming the centralized replication solution is used in the example + network of above Figure 1: RB5 is the distribution tree root and + centralized replication node; CE1 and CE2 are active-active accessed + to RB1, RB2, and RB3 through LAALP1 and LAALP2, respectively; and 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. + + + + + + + + + + +Hao, et al. Standards Track [Page 9] + +RFC 8361 Centralized Replication for BUM Traffic April 2018 + + + 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 would not be + re-encapsulated. If it is sent to an R-nickname that is not a + tree root, it either will not be 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. The incoming + traffic 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 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 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 + + + +Hao, et al. Standards Track [Page 10] + +RFC 8361 Centralized Replication for BUM Traffic April 2018 + + + 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 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, 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 example, there are three R-Nicknames (RNs). The RNs will be + ordered RN0 to RN2. Assuming there are five VLANs from VLAN ID1 to + ID5 spreading among edge RBridges, the traffic in VLAN1 will go to + RN1, 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 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. + +9. Coexisting with the CMT Solution (RFC 7783) + + +------+ +------+ + |(RB6) | |(RB7) | + +------+ +------+ + ------------------|-----------|---------------------- + | | | | | + +------+ +------+ +------+ +------+ +------+ + |(RB1) | |(RB2) | |(RB3) | |(RB4) | |(RB5) | + +------+ +------+ +------+ +------+ +------+ + | | | | | + ------------ ------------------------- + | | + +------+ +------+ + | CE1 | | CE2 | + +------+ +------+ + + Figure 2: CMT and Centralized Replication Coexisting Scenario + + + + +Hao, et al. Standards Track [Page 11] + +RFC 8361 Centralized Replication for BUM Traffic April 2018 + + + Both the centralized replication solution and the Coordinated + Multicast Trees (CMT) solution from [RFC7783] rely on using pseudo- + nicknames to avoid MAC flip-flop on remote RBridges. These two + solutions can 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; 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 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 the ingress nickname. + + An edge group using CMT [RFC7783] MUST NOT set the 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 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, 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 + + 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 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. + + + + + + + + +Hao, et al. Standards Track [Page 12] + +RFC 8361 Centralized Replication for BUM Traffic April 2018 + + +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 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 + 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 access. If the R flag is zero, that nickname will not be + used for that purpose. + + + + + + +Hao, et al. Standards Track [Page 13] + +RFC 8361 Centralized Replication for BUM Traffic April 2018 + + + o C = If C flag is one, it indicates that the TRILL traffic with + this nickname as an ingress nickname 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. + + Due to errors or due to transient inconsistent LSPs when the link + state database is converging after a configuration change or the + like, it is possible 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 were set if any + Nickname Flags APPsub-TLV for that nickname has the R / C flag set. + +12. Security Considerations + + This document 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. + + For general TRILL security considerations, see [RFC6325]. For + security considerations related to pseudo-nickname active-active, see + [RFC7781]. + +13. IANA Considerations + + IANA has assigned two bits in the Nickname Flags APPsubTLV flags for + the R and C bits discussed in Section 11.1 and update the "NickFlags + Bits" subregistry of the "Transparent Interconnection of Lots of + Links (TRILL) Parameters" registry as follows: + + Bit Mnemonic Description Reference + --- -------- -------------------- ----------- + 2 R Replication Nickname [RFC8361] + 3 C Special RPF Check [RFC8361] + + + + + + + + + + + + +Hao, et al. Standards Track [Page 14] + +RFC 8361 Centralized Replication for BUM Traffic April 2018 + + +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, + . + + [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, . + + [RFC6165] Banerjee, A. and D. Ward, "Extensions to IS-IS for Layer-2 + Systems", RFC 6165, DOI 10.17487/RFC6165, April 2011, + . + + [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, + . + + [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, + . + + [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, + . + + [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, + . + + [RFC8174] Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC + 2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174, + May 2017, . + + + + + + + +Hao, et al. Standards Track [Page 15] + +RFC 8361 Centralized Replication for BUM Traffic April 2018 + + +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, + . + + [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, . + + [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, + . + + [IEEE802.1AX] + IEEE, "IEEE Standard for Local and metropolitan area + networks -- Link Aggregation", IEEE 802.1AX, + DOI 10.1109/IEEESTD.2017.7888436, March 2017, + . + + + + + + + + + + + + + + + + + + + + + + + + + + +Hao, et al. Standards Track [Page 16] + +RFC 8361 Centralized Replication for BUM Traffic April 2018 + + +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 + China + + Email: haoweiguo@huawei.com + + + Yizhou Li + Huawei Technologies + 101 Software Avenue, + Nanjing 210012 + 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, et al. Standards Track [Page 17] + -- cgit v1.2.3