| Internet-Draft | Centerlized ECMP RT-5 | March 2022 |
| Wang | Expires 22 September 2022 | [Page] |
In Section 5.8 of [I-D.wang-bess-evpn-arp-nd-synch-without-irb], centerlized RT-5 advertisement are used for common prefixes behind different CEs, This draft describes the requirements for such scenarios. Then this draft reuse the procedures defined in Section 6.2.2 of [I-D.wz-bess-evpn-vpws-as-vrf-ac] to support this scenario.¶
This Internet-Draft is submitted in full conformance with the provisions of BCP 78 and BCP 79.¶
Internet-Drafts are working documents of the Internet Engineering Task Force (IETF). Note that other groups may also distribute working documents as Internet-Drafts. The list of current Internet-Drafts is at https://datatracker.ietf.org/drafts/current/.¶
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."¶
This Internet-Draft will expire on 22 September 2022.¶
Copyright (c) 2022 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 Revised BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Revised BSD License.¶
In Figure 1, Both R1 and R2 advertise their prefixes to DGW1 according to Section 5 of [I-D.wang-bess-evpn-arp-nd-synch-without-irb]. Subnet SN1 can only be reached through R1, so we say SN1 is R1's exlusive prefix. Then subnet SN2 is R2's exlusive prefix, because SN2 can only be reached throug R2. But subnet SN21 can be reached either through R1, or through R2, thus we say SN21 is a common prefix of R1 and R2.¶
When there are both exlusive prefixes and common prefixes behind some CEs, some special requirements should be considered, especially when some of these CEs will not aware which prefixes are the common prefixes. This draft describes the requirements and solutions related to these scenarios .¶
R1 and R2 both establish a single CE-BGP session with DGW1. These CE-BGP session can be called the centerlized CE-BGP session.¶
SN21 is a common prefix of R1 and R2, when R4 send traffic to a host inside subnet SN21, DGW1' should load-balance that traffic between PE1, PE2 and PE3.¶
+--------------------->----------------------+
| CE-BGP |
| PE3 |
| +----------+ |
| | +------+ | ------> | CE-BGP
R2 | | | | | RT-2R | SN21,SN2
+-------+ | | VPNx | | 20.2(MAC34) | NH=20.2
| | P3.1 | | | | MAC34 |
| ...................(20.9)| | |
| . | | +------+ | | DGW1'
SN2--| . | +----------+ +---|---------+
| . | ^ <---------- | V |
| . | | RT-2 RT-5 | +---------+ |
|(20.2) | | 20.2 SN21 | | VPNx | |
| . | | ESI34 GW-IP=20.2 | | |....R4
+---| . | | | |(z.z.z.z)| |
| | . | +----------+ ------> | | | |
| | . | | +------+ | RT-2R | +---------+ |
| | ...................(20.9)| | 20.2 | |
| | | P4.1 | | | | MAC34 +-------------+
| +-------+ | | | |
SN21 PE1 | | VPNx | | ------>
| R1 | | | | RT-2R
| +-------+ | | | | 10.2
| | | P1.1 | | | | MAC21
| | ...................(10.9)| | DGW1
| | . | | +------+ | +-------------+
| | . | +----------+ | |
+---| . | ^ <---------- | +---------+ |
| . | | RT-2 RT-5 | | | |
|(10.2) | | 10.2 SN21 | | VPNx | |
| . | | ESI21 GW-IP=10.2 | | |....R3
| . | | | |(z.z.z.z)| |
SN1--| . | +----------+ ------> | +---------+ |
| . | | +------+ | RT-2R | ^ |
| ...................(10.9)| | 10.2 | | |
| | P2.1 | | | | MAC21 +---|---------+
+-------+ | | VPNx | | |
| | | | | | CE-BGP
| | +------+ | | SN21,SN1
| +----------+ | NH=10.2
| PE2 |
| CE-BGP |
+--------------------->----------------------+
Note that we just use centerlized CE-BGP session to discover CE-prefixes, but we still expect a distributed Layer 3 forwarding framework.¶
Most of the terminology used in this documents comes from [RFC7432] and [RFC9136] except for the following:¶
Before advertise SN1/SN2/SN21 to DGWs, R1 and R2 don't have to know which prefix is their common prefix, and which prefix is their exclusive prefix.¶
The CE-BGP session between R1 and DGW1 is established between 10.2 and z.z.z.z. The IP address 10.2 is called the uplink interface address of R1 in this document. The IP address z.z.z.z is called the centerlized loopback address of VPNx in this document. The IP address 10.9 is called the downlink VRF-interface address of PE1/PE2 in this document.¶
R1 advertises a BGP route for a prefix (say "SN21") behind it to DGW1 via that CE-BGP session. The nexthop for SN21 is R1's uplink interface address (say 10.2).¶
R2 advertises a BGP route for a prefix (say "SN21") behind it to DGW1' via that CE-BGP session. The nexthop for SN21 is R2's uplink interface address (say 20.2).¶
Note that the data packets from R1(R2) to the centerlized loopback address may be routed following the default route on R1(R2). Thus DGW1 doesn't need to use the CE-BGP session to advertise prefixes of VPNx to R1(R2).¶
When PE1 and PE2 learns the ARP entry of 10.2, it advertises a RT-2R route to DGW1 (and DGW1').¶
When PE1 and PE3 learns the ARP entry of 20.2, it advertises a RT-2R route to DGW1 (and DGW1').¶
When DGW1 receives the SN21 from the CE-BGP session. The nexthop for SN21 is 10.2. So DGW1 advertises a RT-5G route to PE1/PE2/PE3 for SN21. The GW-IP value of the RT-5G route for SN21 is 10.2.¶
When DGW1' receives the SN21 from the CE-BGP session. The nexthop for SN21 is 20.2. So DGW1 advertises a RT-5G route to PE1/PE2/PE3 for SN21. The GW-IP value of the RT-5G route for SN21 is 20.2.¶
DGW1 and DGW1' may be the same device, in such case DGW1 should use the ADD-PATH of [RFC7911] to advertise two GW-IPs for the same prefixe SN21.¶
Note that when other PEs receive these RT-5 route for SN21, the ECMP behavior is already defined in Section 4.1 of [RFC9136] as the following:¶
* Based on the BD-10 Route Target in DGW1 and DGW2, the IP
Prefix route is also imported, and SN1/24 is added to the IP-
VRF with Overlay Index IP2 pointing at the local BD-10. In
this example, it is assumed that the RT-5 from NVE2 is
preferred over the RT-5 from NVE3. If both routes were
equally preferable and ECMP enabled, SN1/24 would also be
added to the routing table with Overlay Index IP3.
The RT-2R routes advertisement between PE1 and PE2 is used to sync subnet 10.0's ARP entries to each other in order to avoid ARP missing. The ESI Value of these two RT-2R routes is ESI21.¶
The RT-2R routes advertisement between PE1 and PE3 is used to sync subnet 20.0's ARP entries to each other in order to avoid ARP missing. The ESI Value of these two RT-2R routes is ESI34.¶
There is no IANA consideration needed.¶