IPv6 Multicast Source Routing Traffic Engineering

Segment Routing () leverages the mechanism of source routing. An ingress node steers a packet through an ordered list of instructions, called "segments". Each one of these instructions represents a function to be implemented at a specific location in the network. A function is locally defined on the node where it is executed. Network Programming combines Segment Routing functions to achieve a networking objective that goes beyond mere packet routing. defines the SRv6 Network Programming concept and specifies the main Segment Routing behaviors and network programming functions. Bit Index Explicit Replication (BIER) is an architecture that provides optimal multicast forwarding without requiring a protocol for explicitly building multicast distribution trees or per-flow state maintained by intermediate routers. When a multicast data packet enters BIER forwarding domain, the ingress node encapsulates the packet with a bitstring, each bitposition of which presents the egress nodes. To forward the packet to a given set of egress nodes, the bits corresponding to those egress nodes are set in the bitstring. The intermediate nodes in the BIER domain replicate and forward the packet based on the bitstring.The mechanism of forwarding a packet based on bitstring of BIER are specified in . An IPv6 based multicast source routing (MSR6) solution is defined in , which leverages the benefits of source routing over IPv6 data plane to provide simplified multicast TE and BE service in an IPv6 network without unnecessary multicast tree status and complex control plane protocols. MSR6 needs to reuse the advantages of SRv6 and BIER to implement source routing. MSR6 has two basic modes of forwarding: one is based on Shortest Path First(SPF), which is called MSR6 BE mode; the other is based on traffic engineered, which is called MSR6 TE mode. This document defines 2 new types of segment: End.RL and End.RL.X , and the corresponding packet processing procedures over the IPv6 data plane for the MSR6 TE solutions.

MSR6: Multicast Source Routing over IPv6, defined in . MRH: Multicast Routing Header, a new type of Routing Header which is used for MSR6. Multicast domain: A set of network device which could provide P2MP multicast transport. In this document, the multicast domain is an MSR domain, where every nodes support the capability of MSR6. Root Node: Root node is the beginning point of a multicast tree and also the initiation node of a multicast tunnel. It encapsulates the packet with a multicast header. The type of the encapsulation depends on the routing protocol used in the multicast domain. For MSR6 TE, the encapsulation is MSR6 TE header, which is an IPv6 header with MRH. Replication Endpoint: the intermediate node of a multicast tree, which replicates packet and forwards the packet to the downstream nodes. For MSR6, the Replication Node is called Replication Endpoint which can be indicated by the MSR6 Segment and replicate packets according to the multicast source routing information encapsulation in the MSR6 header of the packet. Leaf Node: Leaf node is the end point of a multicast tree and also the decapsulation node of a multicast tunnel. It decapsulates the multicast header in the packet and forwards the packet based on overlay encapsulation. Parent Node: The parent node is the node that does the packet replication, corresponding to the concept of a child node. Child Node: The child node is the downstream node that will receive the packet which has been replicated ,corresponding to the concept of a parent node.

In order to implement MSR6 TE, the nodes and links along the path must be explicitly specified in the packet. In SRv6, the segment list is suitable for source route of unicast path, which is P2P and could be presented by the data structure of a chain with a pointer; In the Segment Routing Header(SRH), the segment list is the chain, and the segment left is the pointer. But for the multicast, the path of multicast service is a tree, and can not be presented by data structure in a one-dimensional array. So this document introduces new segment to present node/link in a multicast tree and introduce structure information in the segment to indicate the parent-child relationship during multicast replication. And all the nodes/links along the multicast tree can be encoded into the segment list. In a multicast tree, when the packet is replicated from one node to multiple downstream nodes, parent-child relationship is built up between these nodes. MRH is supposed to encode the nodes along the P2MP path and also the parent-child relationship between them.

As defined in [RFC8402], an Segment Identifier is an IPv6 address explicitly associated with the segment. When an SRv6 SID is in the Destination Address field of an IPv6 header of a packet, it is routed through transit nodes in an IPv6 network as an IPv6 address. MSR6 Segment Identifier complies with the definition of the Segment Identifier in . Following the specification in RFC8986, an MSR6 SID is consisting of LOC:FUNCT:ARG. Locator could be represented as B:N, B is the MSR6 segment block(IPv6 prefix allocated for MSR6 segment by the operator) and N is the identifier of the node instantiating the SID. The MSR6 locator is routable and leads to the node which instantiates the SID; Function is the identification of a local behavior bound to the MSR6 SID; Argument is the additional information requested by the function. Typical MSR6 functions and arguments are defined in the following sections.

End.RL SID is one of the basic MSR6 SID as an extension of End SID defined in RFC8986. The encoding of segment list which is composed of End.RL SIDs follow the rules: There MUST be an End.RL SID for each Replication Endpoint in the multicast tree The End.RL SIDs, which represent child nodes that have the same parent node, MUST be arranged consecutively in the segment list The End.RL SID for a parent node MUST indicate the position of the End.RL SIDs for the child nodes There are 2 arguments for each End.RL SID: "Replication number" is used for indicating how many times the parent node is supposed to execute replication "Pointer" is used for indicating the position of End.RL SID for the first child node in the segment list; When the packet is received by an Replication Endpoint and the DA of this packet is a local SID with the function of End.RL, the packet will be replicated based on the "replication number". The DA of the 1st replicated packet is replaced by the SL="pointer" and the value of Segment Left is set to "pointer"; the DA of the 2nd replicated packet is replaced by the SL="pointer -1" and the value of Segment Left is set to "pointer-1". The packet is sent out based on the updated DA. Repeat the operation untill all the replicated packets are sent out. The replication node does the following when the DA of the packet is a local End.RL SID.

max_LE) or (Segments Left > Last Entry+1)) { S10. Send an ICMP Parameter Problem to the Source Address with Code 0 (Erroneous header field encountered) and Pointer set to the Segments Left field, interrupt packet processing, and discard the packet. S11. } S12. Decrement IPv6 Hop Limit by 1 S13. Replicate the packet based on the "replication number" in the argument of the SID S14. Set the Segment Left of the 1st replicated packet to "Pointer" in the argument of the SID S15. Update IPv6 DA with Segment List[Pointer] respectively in each replicated packet S16. Submit the packet to the egress IPv6 FIB lookup for transmission to the new destination S17. Repeat S14-S17 with "Pointer+n"(n=the number of packets which have been updated) untill all the packets are transmitted to based on the updated DA R18. }]]>

End.RL.X SID is one of the basic MSR6 SID as an extension of End.X SID defined in RFC8986. The encoding of segment list which is composed of End.RL.X SIDs follow the rules: There MUST be an End.RL.X SID for each downstream link connected to a Replication Endpoint in the multicast tree The End.RL.X SIDs corresponding to downstream links of the same Replication Endpoint MUST be arranged consecutively in the segment list The End.RL.X SID for a parent node MUST indicate the position of the End.RL.X SIDs for the child nodes There are 2 arguments for each End.RL SID: "Replication number" is used for indicating how many times the existing node is supposed to execute replication "Pointer" is used for indicating the position of End.RL.X SID for the first downstream link of the child node in the segment list; When the packet is received by a Replication Endpoint and the DA of this packet is a local SID with the function of End.RL.X, the packet will be replicated based on the "replication number". The DA of the 1st replicated packet is replaced by the Segment List[ pointer ] and the value of Segment Left is set to "pointer"; The packet is sent out through the link indicated by the SID. the DA of the 2nd replicated packet is replaced by the Segment List [SL +1] and the value of Segment Left is set to "SL+1"; The DA of the 2nd packet is replaced by the Segment List [ pointer2 ] (pointer2 is the argument in the Segment List [SL+1]) and the value of Segment Left is set to "pointer2"; The packet is sent out through the link indicated by the SID; Repeat the operation untill all the replicated packets are sent out through the indicated links.

max_LE) or (Segments Left > Last Entry+1)) { S10. Send an ICMP Parameter Problem to the Source Address with Code 0 (Erroneous header field encountered) and Pointer set to the Segments Left field, interrupt packet processing, and discard the packet. S11. } S12. Decrement IPv6 Hop Limit by 1 S13. Replicate the packet based on the "replication number" in the argument of the SID S14. Set the Segment Left of the 1st replicated packet to "Pointer" in the argument of the SID S15. Update IPv6 DA with Segment List[Pointer] S16. Transmit the packet through the indicated link S17. Set the Segment Left of the 2nd replicated packet to "Segment Left+1" in the argument of the SID S18. Update IPv6 DA with Segment List[Segment Left+1] S19. Set the Segment Left of the 2nd replicated packet to "Pointer 2" in the argument of the updated SID S20. Update IPv6 DA with Segment List[Pointer 2] S21. Transmit the packet through the indicated link S22. Repeat S17-S21 with "Pointer+n"(n=the number of packets which have been updated) untill all the packets are transmitted to based on the updated DA R23. }]]>

A new type of Routing Header is defined as follows for MSR6 Traffic Engineering called Multicast Routing Header(MRH).

Next Header: Defined in , section 4.4. Hdr Ext Len: Defined in , section 4.4. Routing Type: Allocated by IANA. Segments Left: Defined in , section 4.4. MRH Sub-type: 8-bit identifier of the sub-type. The value of sub-type is supposed to be assigned by IANA. Optional Type Length Value objects: TLV is used to carry other type of information that is not suitable to indicated in the segment list. The formart of M-SID defined in this document is shown in the following figure:

It includes: Locator: could represented as B:N where B is the MRH segment block (IPv6 prefix allocated for MRH segment by the operator) and N is the identifier of the node instantiating the SID. When the locator part of the MRH segment is routable, it leads to the node, which instantiates the SID. Function: an identification of a local behavior bound to the MRH. Replication is default behavior for any MRH segment, which doesn't need function indication. Replication number is used for indicating how many times the existing node is supposed to excute replication Pointer is used for indicating the position of MSR6 SID for the child node; Segments for the group of child nodes which belong to the same parent node MUST be encoded together in the segment list. So Pointer and replication number could determine the upper bound and lower bound of the value range of segment left for the child nodes. The destination address of the packet is the IPv6 address of the existing Replication Endpoint. The next header of the IPv6 header points to a Routing Header and the type of the routing header is MRH Type.

Different from unicast, MSR is designed for an explicit multicast tree rather than a p2p path: the former contains more nodes than the latter generally. So the header overhead is one of the key challenge for MSR6, which limits the scalability and the potential deployment scenario. There are some SRH compression solutions under discussion in IETF. These solutions could also be used for MSR6 with modification.

The compressed MRH6 is designed as follows:

Each M-SID in the MRH is shown in the following figure:

It includes: Node ID: is the identifier of the node instantiating the SID. Common Profix and Node ID together form a routable IPv6 address Function: an identification of a local behavior bound to the MRH. Replication is default behavior for any MRH segment, which doesn't need function indication. Arguments: follow the definition in section 6, including replication number and pointer. The processing process is: The destination address of the packet is the IPv6 address of the existing Replication Endpoint. The next header of the IPv6 header points to a Routing Header and the type of the routing header is MRH Type 2. The Replication Endpoint is supposed to Find the next groups of segments in the MRH which present the child nodes of the existing Replication Endpoint with Pointer-1 and Pointer-2; Replicate the packet based on the number of child nodes, which is the replication number; Replace the SID part of IPv6 destination address with the corresponding child node's segment in order for each replicated packet; Update the segment left field of each replicated packet based on the location child node's segment; Forward the packet to the output port based on the FIB with the existing destination address as an entry;

In order to

MSR-R<--|->MSR End<-|-->MSR-L<--| ]]> MSR-R: MSR6 Root Node, including Node A; MSR End: MSR6 Replication Endpoint, including Node B and C; MSR-L: MSR6 Leaf Nodes, including Node D, E, F, G;

In node A, the packet is encapsulated the packet with an IPv6 header carrying an MRH. The segment list in the MRH is as follows. The destination address of the IPv6 header is the 1st SID in the segment list, which is the local SID of node A. Based on the End.RL behavior defined in section 5.1, the packet is replicated to 2 (Replication Number+1=1+1=2) packets. In the 1st packet, the Segment Left is set to 2 and DA is replaced by Segment List[2]; In the 2nd packet, the Segment Left is set to 3 and DA is replaced by Segment List[3]; The packets are routed to the node B and node C respectively.

In MSR6 Replication Endpoint B, the destination address of the IPv6 header is the local SID of node B. Based on the End.RL behavior defined in section 5.1, the packet is replicated to 2 (Replication Number+1=1+1=2) packets. In the 1st packet, the Segment Left is set to 4 and DA is replaced by Segment List[4]; In the 2nd packet, the Segment Left is set to 5 and DA is replaced by Segment List[5]; The packets are routed to the node D and node E respectively. In MSR Endpoint C, the process is similar. In MSRE D, the destination address of the IPv6 header is the local SID of node D. Based on the End.RL behavior defined in section 5.1, when replication number=0, node D stops processing the MRH and continues to process the next header in the packet.

In node A, the packet is encapsulated the packet with an IPv6 header carrying an MRH. The segment list in the MRH is as follows. The destination address of the IPv6 header is the local SID of node A. Based on the End.RL.X behavior defined in section 5.2, the packet is replicated to 2 (Replication Number+1=1+1=2) packets. In the 1st packet, the Segment Left is set to 3 and DA is replaced by Segment List[3] and send the packet based on the specified adjacency A1; In the 2nd packet, the Segment Left is set to 2 and DA is replaced by Segment List[2]; Based on the arguments in the updated DA, the Segment Left is set to 5 and DA is replaced by Segment List[5] and send the packet based on the specified adjacency A2;

In node B, the packet is encapsulated the packet with an IPv6 header carrying an MRH. The segment list in the MRH is as follows. The destination address of the IPv6 header is the local SID of node B. Based on the End.RL.X behavior defined in section 5.2, the packet is replicated to 2 (Replication Number+1=1+1=2) packets. In the 1st packet, the DA is replaced by the corresponding leaf stored in the node and send the packet based on the specified adjacency B1; In the 2nd packet, the DA is replaced by the corresponding leaf stored in the node and send the packet based on the specified adjacency B2; In MSR Endpoint C, the process is similar. In MSRE D, the destination address of the IPv6 header is the local SID of node D. Node D stops processing the MRH, and begin to process the next header in the packet.

This document makes no request of IANA. Note to RFC Editor: this section may be removed on publication as an RFC.