SPRING S. Agrawal
Internet-Draft Z. Ali
Intended status: Standards Track C. Filsfils
Expires: August 16, 2020 Cisco Systems
D. Voyer
Bell Canada
G. Dawra
Z. Li
Huawei Technologies
February 13, 2020

SRv6 and MPLS interworking


This document describes SRv6 and MPLS/SR-MPLS interworking and co-existence procedures.

Requirements Language

The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in RFC 2119.

Status of This Memo

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 August 16, 2020.

Copyright Notice

Copyright (c) 2020 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

Many of the deployments require SRv6 insertion in the brownfield networks. The incremental deployment of SRv6 into existing networks require SRv6 to interwork and co-exist with SR-MPLS/ MPLS.

There are various SRv6 and SR-MPLS/ MPLS interworking scenarios possible. They can be classified into the following four categories.

These scenarios cover various cascading of SRv6/ MPLS network, e.g., SR-MPLS/MPLS <-> SRv6 <-> SR-MPLS/MPLS <-> SRv6 <-> SR-MPLS/MPLS, etc. The draft addresses all these possible interworking scenarios.

In addition, the draft also addresses migration and coexistence of the SRv6 and SR-MPLS/ MPLS. Co-existence means a network that supports both SRv6 and MPLS in a given domain. This may be a transient state when brownfield SR-MPLS/ MPLS network upgrades to SRv6 (migration) or permanent state when some devices are not capable of SRv6 but supports native IPv6 and SR-MPLS/ MPLS.

1.1. Terminology

SID: A Segment Identifier which represents a specific segment in segment routing domain. The SID type used in this document is IPv6 address (also referenced as SRv6 Segment or SRv6 SID).

Node k has a classic IPv6 loopback address Ak::/128.

A SID at node k with locator block B and function F is represented by B:k:F::

A SID list is represented as <S1, S2, S3> where S1 is the first SID to visit, S2 is the second SID to visit and S3 is the last SID to visit along the SR path.

(SA,DA) (S3, S2, S1; SL) represents an IPv6 packet with:

-IPv6 header with source address SA, destination addresses DA and SRH as next-header

-SRH with SID list <S1, S2, S3> with SegmentsLeft = SL

Note the difference between the <> and () symbols: <S1, S2, S3> represents a SID list where S1 is the first SID and S3 is the last SID to traverse. (S3, S2, S1; SL) represents the same SID list but encoded in the SRH format where the rightmost SID in the SRH is the first SID and the leftmost SID in the SRH is the last SID. When referring to an SR policy in a high-level use-case, it is simpler to use the <S1, S2, S3> notation. When referring to an illustration of the detailed packet behavior, the (S3, S2, S1; SL) notation is more convenient.

domain: Without loss of the generality, domain is assumed to be instantiated by a single IGP instance or a network within IGP if there is clear separation of data plane.

2. Interworking Procedures

This documents refers to interworking as a stitching of SRv6 domain and SR-MPLS/ MPLS domain. Special stitching procedures are performed on routers which acts as border between such domains. Border routers need to support both SRv6 and SR-MPLS/ MPLS. Interworking is applicable when SRv6 domains are deployed and need to interop with existing SR-MPLS/ MPLS backbones or access networks.

This draft proposes two ways to stitch heterogeneous domains: a controller based solution and a BGP signaling based approach. The PCE based solution is applicable to both best effort as well as deployments where tight SLA guarantees are required (e.g., ODN like deployments scenarios). The BGP signaling covers the best effort case. Specifically, the draft proposes the following two ways to stitch heterogeneous domains end to end:

In summary the draft covers the following SRv6/ MPLS interworking scenarios.
- Carrying SRv6 over SR-MPLS (controller stitches domains).
- Carrying SRv6 over SR-MPLS (BGP stitches domains).
- Carrying SR-MPLS over SRv6 (controller stitches domains).
- Carrying SR-MPLS over SRv6 (BGP stitches domains).
- SRv6 to SR-MPLS translation (controller stitches domains).
- SRv6 to SR-MPLS translation (BGP stitches domains).
- SR-MPLS to SRv6 translation (controller stitches domains).
- SR-MPLS to SRv6 translation (BGP stitches domains).
- Cascaded domains (controller stitches domains).
- Cascaded domains (BGP stitches domains).

While the number of interworking scenarios is large, the few building blocks outlines in this draft address all of them. For the same reasons, without the loss of generality, the building blocks are illustrated using the SRv6 over SR-MPLS example of Figure 1 but the procedure equally applies to the other deployment scenarios.

                  2                       5                       8
               *     *                 /     \                 *     *  
            *           *           /           \           *           *
         *                 *     /                 \     *                 *
      1         SRv6          4         MPLS          7         SRv6          10
         *      IGP1       *     \      IGP2       /     *      IGP3       *
            *             *         \           /           *           *
               *     *                 \     /                 *     *
                  3                       6                       9


Example Network Scenario(6oM)

Figure 1

3. Building blocks for domain stitching

3.1. Stitching heterogenous domains using a Controller

This procedure provides a best-effort path as well as a path that satisfies the Service Level Agreement (SLA), across multiple domains. A PCE may act as an SDN controller. In that case, based on the SLA, the PCE computes and programs end to end path. The PCE is also aware of interworking requirement at border nodes, as each domain feeds topological information to the controller through BGP LS feeds. Intermediate domain of different data plane type is represented by Binding SID (BSID) of head end type in SID list. The intermediate domain BSID is programmed at domain entry border node with SID list through domain and exit node SID as last segment. In summary, an intermediate heterogenous domain is replaced by a BSID of the data plane nature of headend. The procedure work for all of deployment model mentioned above.

3.1.1. Illustration

The procedure is illustrated using the example of Figure 1.

When a service prefix (e.g., vpn or evpn) is received on head end with SLA (color extended community), the head- end (Node 1) node requests a PCE for a path to egress node that can satisfy the SLA. This is because Node 1 does not know how to compute the traffic engineered path through the multi-domain network to node 10. Node 1 requests SR PCE to compute a path to node 10 providing optimization objective, constraints(eg: low latency). The PCE computes low latency path via node 2, 5 and 8. The PCE identifies the end-to-end path is not consistent data plane and kicks in interworking procedures at the border node(4). It programs a policy at 4 that "Stitches" domains using SRv6 End.BM BSID.The PCE installs SRv6 End.BM BSID at node 4 with segments are node 5, 7. SR PCE responds back to node 1 with SRv6 segments via node 2, End.BM at node 4, node 8 and node 10.

The data plan operations for the above-mentioned interworking example are described in the following:

3.2. Stitching heterogenous domains usin BGP inter domain routing

For providing services across domains, edge node locators need to be reachable.

-Locators are advertised by edge nodes in the BGP ipv6 unicast address family (AFI=2,Safi=1) to border nodes. These locators are also advertised in its local IGP domain.

-On border nodes these prefixes are like any IPv6 global prefixes. These will be advertised in BGP IPv6 LU[AFI=2/SAFI=4] session using 3107 procedures in label core. It could be summary prefix for all locators in that domain.

-Remote domain border router advertising locator over SRv6 domain need to attach SRv6 SID in prefix SID attribute. SRv6 SID in this case will be End function of advertising border node.

-Ingress node learns remote locators over BGP ipv6 address family[AFI=2, SAFI=1]. These locators have prefix SID attribute containing SRv6 SID. This SRv6 SID is End function of advertising border node and helps to tunnel traffic to border node in remote domain.

-If locators are leaked into remote IGP and no tunneling of traffic will be needed in remote domain. Hence attaching SRv6 SID on remote border nodes can be avoided.

These procedures work for any of deployment model mentioned above. Below are some important aspects for Mo6, 6toM, Mto6
-Loopback address are advertised in BGP label unicast session to border node when advertised from MPLS domain. These are also advertised in local IGP.
-Border nodes advertises prefix over SRv6 domain in BGP IPv4/IPv6 session. It attaches prefix SID attribute with SRv6 SID. This SRv6 SID maps to label received in prefix update. -Remote border node allocates local label to advertise prefix in MPLS domain to ingress node. This local label maps to received SRv6 SID in prefix sid attribute of prefix.

3.3. 6toM and Mto6 considerations

For 6toM and Mto6 BGP inter domain or ODN multi domain stitching will work if SRv6 edge nodes are capable of handling vpn/service label. In 6toM scenario, ingress node should be able to encap vpn label and then perform T.Encap function with SRv6 SID associated with prefix nexthop. In Mto6 case, traffic will be received with SRv6 SID and vpn label below it on egress PE. So egress SRv6 capable node should be able to process vpn labels after decapsulating SRv6 SID and when next header is 137 in IPv6 header.

Service information encoded by SRv6 PE will be in SRv6 Service SID and MPLS PE will be vpn label/service label or just IP payload for internet. If SRv6 PE do not support vpn label, then we need some special handling to translate SRv6 service SID to vpn label and vice versa at border nodes. This will be detailed in future versions

4. FRR handling

Failure within domain are taken care by existing FRR(TILFA, rLFA, LFA etc) mechanisms.

Failure of border nodes are to be addressed in a future version of the document.

5. Migration and co-existence

These procedures would be detailed in a future revision

6. BGP based services interworking and migration

SRv6-based VPN (SRv6-VPN)/EVPN service information is encoded in SRv6 SIDs specifically END.DT*/END.DX*/END.DT2. MPLS-based VPN service information is encoded in labels. This requires special consideration during Migration and Interworking. Will be discussed more detail in future versions

7. IANA Considerations


8. Security Considerations

9. Acknowledgements

10. 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.

Authors' Addresses

Swadesh Agrawal Cisco Systems EMail: swaagraw@cisco.com
Zafar ALI Cisco Systems EMail: zali@cisco.com
Clarence Filsfils Cisco Systems EMail: cfilsfil@cisco.com
Daniel Voyer Bell Canada Canada EMail: daniel.voyer@bell.ca
Gaurav dawra LinkedIn USA EMail: gdawra.ietf@gmail.com
Zhenbin Li Huawei Technologies China EMail: lizhenbin@huawei.com