RFC 7795
Pseudowire Redundancy on the Switching Provider Edge (S-PE)
Pseudowire Redundancy on the Switching Provider Edge (S-PE) 🦋
Okay bestie, here's the vibe check on this one 💅
- Stream: Internet Engineering Task Force (IETF)
- RFC: 7795
- Category: Standards Track
- Date: February 2016
- Authors: J. Dong and H. Wang, Huawei Technologies
- ISSN: 2070-1721
- Obsoletes or updates: the title block doesn't list anything this one obsoletes or updates 🙈
Abstract 💖
okay so this doc is all about Multi-Segment Pseudowire (MS-PW) protection where the redundancy lives on the Switching Provider Edge (S-PE), as defined in RFC 5659 🧚 it lays out how the S-PEs that actually provide the redundancy need to behave. it reuses the Preferential Forwarding status signaling from RFCs 6870 and 6478, and the best part is that it does not require any change to the Terminating Provider Edges (T-PEs) of MS-PW. the T-PEs get to chill ✨
Status of This Memo 🌸
this is an Internet Standards Track document, no cap.
it's a product of the IETF, so it represents the consensus of the IETF community. it got public review and was approved for publication by the Internet Engineering Steering Group (IESG). if you want more on Internet Standards, check Section 2 of RFC 5741.
info about the current status, any errata, and how to send feedback lives at http://www.rfc-editor.org/info/rfc7795 💜
Copyright Notice 🎀
Copyright (c) 2016 IETF Trust and the persons identified as the document authors. All rights reserved.
this doc is subject to BCP 78 and the IETF Trust's Legal Provisions Relating to IETF Documents (http://trustee.ietf.org/license-info) as they stand on the date of publication. please read those carefully, bestie, because they explain your rights and restrictions for this document. any Code Components pulled out of this document must include Simplified BSD License text as described in Section 4.e of the Trust Legal Provisions, and they come without warranty as described in the Simplified BSD License 🫶
Table of Contents 📋
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 2
2. Typical Scenarios of PW Redundancy on S-PE . . . . . . . . . 3
2.1. MS-PW Redundancy on S-PE . . . . . . . . . . . . . . . . 3
2.2. MS-PW Redundancy on S-PE with S-PE Protection . . . . . . 4
3. S-PE Operations . . . . . . . . . . . . . . . . . . . . . . . 4
4. Applications of PW Redundancy on S-PE . . . . . . . . . . . . 5
4.1. Applications in Scenario 1 . . . . . . . . . . . . . . . 5
4.2. Applications in Scenario 2 . . . . . . . . . . . . . . . 6
5. VCCV Considerations . . . . . . . . . . . . . . . . . . . . . 7
6. Security Considerations . . . . . . . . . . . . . . . . . . . 7
7. References . . . . . . . . . . . . . . . . . . . . . . . . . 7
7.1. Normative References . . . . . . . . . . . . . . . . . . 7
7.2. Informative References . . . . . . . . . . . . . . . . . 8
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . 9
Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 91. Introduction ✨
okay so [RFC6718] describes the framework and requirements for pseudowire (PW) redundancy, and [RFC6870] specifies a PW redundancy mechanism for the case where a set of redundant PWs gets configured between Provider Edge (PE) nodes in Single-Segment Pseudowire (SS-PW) [RFC3985] applications, or between Terminating Provider Edge (T-PE) nodes in Multi-Segment Pseudowire (MS-PW) [RFC5659] applications.
in some MS-PW scenarios, putting PW redundancy on the Switching Provider Edges (S-PEs) is seriously beneficial 🔥 it takes some of the burden off the access T-PE nodes and gives you faster protection switching compared to the end-to-end MS-PW protection mechanisms.
this doc describes some scenarios where PW redundancy sits on S-PEs and specifies how those S-PEs operate. the S-PEs connect to neighboring T-PEs or S-PEs with PW segments. for an S-PE that provides PW redundancy for an MS-PW, there's a single PW segment on one side, which is called the single-homed side, and there are multiple PW segments on the other side, which is called the multi-homed side. the scenario where the S-PE has two multi-homed sides is out of scope 🙅♀️ signaling of the Preferential Forwarding status as defined in [RFC6870] and [RFC6478] is reused, and this document does not require any change to the T-PEs of MS-PW.
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 [RFC2119]. 💖
2. Typical Scenarios of PW Redundancy on S-PE 🦋
in some MS-PW deployment scenarios, PW redundancy on S-PEs has real perks. this section walks through the typical scenarios of PW redundancy on an S-PE, so let's go 💅
2.1. MS-PW Redundancy on S-PE
+-----+ AC
+---+ +-----+ | | | +---+
| | | |------|T-PE2|-----| |
| | AC +-----+ | ..PW-Seg2.......| | |
| | | |....PW-Seg1..... | +-----+ | |
|CE1|-----|T-PE1|------|S-PE1| |CE2|
| | | | | . | +-----+ | |
| | +-----+ | ..PW-Seg3.......| | |
| | | |------|T-PE3|-----| |
+---+ +-----+ | | | +---+
+-----+ AC
Figure 1: MS-PW Redundancy on S-PEso in Figure 1, Customer Edge (CE) node CE1 is hooked up to T-PE1, while CE2 is dual-homed to T-PE2 and T-PE3. T-PE1 only connects to S-PE1, and S-PE1 connects to both T-PE2 and T-PE3. the MS-PW gets switched on S-PE1, and PW segments PW-Seg2 and PW-Seg3 provide resiliency on S-PE1 for the failure of T-PE2, T-PE3, or the connected Attachment Circuits (ACs). PW-Seg2 is picked as the primary PW segment, and PW-Seg3 is the secondary PW segment 🎀
this setup is beneficial in Figure 1 because T-PE1, being an access node, may not support PW redundancy, which is a whole mood 😭 plus, with PW redundancy on S-PE, the number of PW segments required between T-PE1 and S-PE1 is only half of what you'd need if end-to-end MS-PW redundancy were used. and in this scenario, PW redundancy on S-PE could give you faster protection switching than end-to-end protection switching of MS-PW ⭐
2.2. MS-PW Redundancy on S-PE with S-PE Protection
+---+ +-----+ +-----+ +-----+
| | | | | | | |
| | AC |......PW1-Seg1......PW1-Seg2........|
| | | | . | | . | | |
|CE1|-----|T-PE1|------|S-PE1|-----------|T-PE2| AC
| | | . | | . | PW1-Seg3 | | | +---+
| | | . | | ......... ......|-----| |
| | | . | | | . .| | | |
+---+ +---.-+ +-----+ . . +-----+ | |
|. . . |CE2|
|. .. | |
|. +-----+ . . +-----+ | |
|. | | . .| |-----| |
|...PW2-Seg1.......... ......| | +---+
| | . | PW2-Seg2 | | AC
----------|S-PE2|-----------|T-PE3|
| . | | |
| .....PW2-Seg3........|
| | | |
+-----+ +-----+
Figure 2: MS-PW Redundancy on S-PE with S-PE Protectionokay now Figure 2 is giving double protection 👑 CE1 is hooked up to T-PE1, and CE2 is dual-homed to T-PE2 and T-PE3. T-PE1 connects to both S-PE1 and S-PE2, and both S-PE1 and S-PE2 connect to both T-PE2 and T-PE3. there are two MS-PWs that get switched at S-PE1 and S-PE2 respectively, so you get S-PE node protection too 🧚 for PW1, S-PE1 provides resiliency using PW1-Seg2 and PW1-Seg3. for PW2, S-PE2 provides resiliency using PW2-Seg2 and PW2-Seg3. PW1 is the primary MS-PW, and PW1-Seg2 between S-PE1 and T-PE2 is the primary PW segment. PW2 is the secondary MS-PW 🌈
this one's beneficial because it reduces the number of end-to-end MS-PWs required for both T-PE and S-PE protection. and PW redundancy on S-PE could also give you faster protection switching compared with end-to-end protection switching of MS-PW 💅
3. S-PE Operations 🔥
okay this is the important part, bestie 🫶 for an S-PE that provides PW redundancy for MS-PW, it's really important to advertise the proper preferential forwarding status to the PW segments on both sides, and then do protection switching based on the status information it gets back. heads up: when PW redundancy for MS-PW is provided on S-PE, the optional S-PE Bypass mode as defined in [RFC6478] MUST NOT be used, because otherwise the S-PE will not receive the PW status messages originated by T-PEs 🙈 this section specifies the operations of S-PEs where PW redundancy is provisioned, and it does not make any change to the T-PEs of MS-PW.
the S-PEs connect to neighboring T-PEs or other S-PEs on two sides with PW segments. for the S-PE that provides PW redundancy for an MS-PW, one side has a single PW segment (the single-homed side) and the other side has multiple PW segments (the multi-homed side). the scenario where the S-PE has two multi-homed sides is out of scope ✨
the S-PE that provides PW redundancy MUST work in Slave mode for the single-homed side, and MUST work in Independent mode for the multi-homed side. consequently, the T-PE on the single-homed side MUST work in the Master mode, and the T-PEs on the multi-homed side MUST work in the Independent mode. the signaling of the Preferential Forwarding bit as defined in [RFC6870] and [RFC6478] is reused 💖
the S-PE MUST pass the Preferential Forwarding status received from the single-homed side unchanged to all the PW segments on the multi-homed side. the S-PE MUST advertise the Standby Preferential Forwarding status to the single-homed side if it receives Standby status from all the PW segments on the multi-homed side, and it MUST advertise the Active Preferential Forwarding status to the single-homed side if it receives Active status from any of the PW segments on the multi-homed side. for the single-homed side, the active PW segment is determined by the T-PE on that side, which works in the Master mode. on the multi-homed side, since both the S-PE and T-PEs work in the Independent mode, the PW segment which has both the local and remote Up/Down status as Up and both the local and remote Preferential Forwarding status as Active MUST be selected for traffic forwarding 🦋 when a switchover happens on the S-PE, and the S-PE supports the SP-PE TLV processing as defined in [RFC6073], it SHOULD advertise the updated SP-PE TLVs by sending a Label Mapping message to the T-PEs 🌸
4.1. Applications in Scenario 1 ✨
okay bestie, let's run through Scenario 1 from Figure 1 🎀 We're assuming the AC from CE2 to T-PE2 is active, so it's the main character of this whole story 👑 In normal operation, S-PE1 gets the Active Preferential Forwarding status bit on its single-homed side from T-PE1. It then advertises that same Active Preferential Forwarding status bit on both PW-Seg2 and PW-Seg3 🦋 At the same time, T-PE2 and T-PE3 tell S-PE1 about their own forwarding states. T-PE2 would advertise the Active Preferential Forwarding status bit, and T-PE3 would advertise the Standby one, reflecting the two ACs connected to CE2 💅 Then, by matching the local and remote Up/Down status and Preferential Forwarding status, PW-Seg2 is the one that gets the traffic 💖
Now the plot twist 😭 When the AC between CE2 and T-PE2 fails, the forwarding state of the AC on T-PE3 changes to Active. T-PE3 then advertises the Active Preferential Forwarding status to S-PE1. T-PE2 would also send a PW status Notification message to S-PE1, saying the AC between CE2 and T-PE2 is down. S-PE1 would then perform the switchover based on the updated local and remote Preferential Forwarding status and the status of "Pseudowire forwarding", and it selects PW-Seg3 as the new segment for traffic forwarding ✨ Since S-PE1 is still connected to an Active PW segment on the multi-homed side, it will not advertise any change of the PW status to T-PE1 🙈 If S-PE1 supports the SP-PE TLV processing defined in [RFC6073], it would advertise the updated SP-PE TLVs by sending a Label Mapping message to T-PE1. Iconic recovery, honestly 🫶
4.2. Applications in Scenario 2 🧚
For the scenario of Figure 2, we're again assuming the AC from CE2 to T-PE2 is active 🌈 T-PE1 works in Master mode, and according to configuration it would advertise the Active and Standby Preferential Forwarding status bit to S-PE1 and S-PE2, respectively. Based on that received Preferential Forwarding status bit, S-PE1 would advertise the Active Preferential Forwarding status bit to both T-PE2 and T-PE3, and S-PE2 would advertise the Standby Preferential Forwarding status bit to both T-PE2 and T-PE3 🍓 T-PE2 would advertise the Active Preferential Forwarding status bit to both S-PE1 and S-PE2, and T-PE3 would advertise the Standby Preferential Forwarding status bit to both S-PE1 and S-PE2. These statuses reflect the forwarding state of the two ACs connected to CE2, so it's all very reflective and iconic 🦋 By matching the local and remote Up/Down Status and Preferential Forwarding status, PW1-Seg2 from S-PE1 to T-PE2 would be used for traffic forwarding. Since S-PE1 connects to the Active PW segment on the multi-homed side, it would advertise the Active Preferential Forwarding status bit to T-PE1. S-PE2 would advertise the Standby Preferential Forwarding status bit to T-PE1 because it does not have any Active PW segment on the multi-homed side 💅
Okay, now the failure 😭 When the AC between CE2 and T-PE2 fails, the forwarding state of the AC on T-PE3 changes to Active. T-PE3 would then advertise the Active Preferential Forwarding status bit to both S-PE1 and S-PE2. T-PE2 would send a PW status Notification message to both S-PE1 and S-PE2, indicating that the AC between CE2 and T-PE2 is down. S-PE1 would perform the switchover based on the updated local and remote Preferential Forwarding status and the status of "Pseudowire forwarding", and select PW1-Seg3 for traffic forwarding. Since S-PE1 still has an Active PW segment on the multi-homed side, it would not advertise any change of the PW status to T-PE1 🙈 If S-PE1 supports the SP-PE TLV processing defined in [RFC6073], it would advertise the updated SP-PE TLVs by sending a Label Mapping message to T-PE1.
And if S-PE1 fails entirely 💀 T-PE1 would notice through some detection mechanism, then advertise the Active Preferential Forwarding status bit to S-PE2, and PW2-Seg1 would be selected by T-PE1 for traffic forwarding. When S-PE2 receives the newly changed Preferential Forwarding status, it would advertise the Active Preferential Forwarding status to both T-PE2 and T-PE3. T-PE2 and T-PE3 would also notice the failure of S-PE1 through some detection mechanism. Then, by matching the local and remote Up/Down and Preferential Forwarding status, PW2-Seg2 would be selected for traffic forwarding. Full comeback arc, no notes 👑🔥
5. VCCV Considerations 🦋
For PW Virtual Circuit Connectivity Verification (VCCV) [RFC5085], the Control Channel (CC) type 1 "PW ACH" can be used with the S-PE redundancy mechanism. Control Channel type 2 "Router Alert Label" is not supported for MS-PW as specified in [RFC6073], so don't even try it 🙅 If VCCV CC type 3 "TTL Expiry" is to be used, the PW label TTL MUST be set to the appropriate value to reach the target PE. The hop count from one T-PE to the target PE can be obtained via SP-PE TLVs, through MS-PW path trace, or based on management-plane information. Choose your fighter 💜
6. Security Considerations 🔥
Since PW redundancy is provided on the S-PE nodes of MS-PWs, it is super important that the security mechanisms defined in [RFC4447], [RFC6073], and [RFC6478] be implemented. This is how you make sure the S-PE nodes, and the messages they send and receive, are not compromised.
This is serious, bestie 😭 If the S-PE nodes get compromised, the whole redundancy setup is compromised with them, so please don't skip these security mechanisms 🚨
7. References 📚
7.1. 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>.
[RFC4447] Martini, L., Ed., Rosen, E., El-Aawar, N., Smith, T., and
G. Heron, "Pseudowire Setup and Maintenance Using the
Label Distribution Protocol (LDP)", RFC 4447,
DOI 10.17487/RFC4447, April 2006,
<http://www.rfc-editor.org/info/rfc4447>.
[RFC6073] Martini, L., Metz, C., Nadeau, T., Bocci, M., and M.
Aissaoui, "Segmented Pseudowire", RFC 6073,
DOI 10.17487/RFC6073, January 2011,
<http://www.rfc-editor.org/info/rfc6073>.
[RFC6478] Martini, L., Swallow, G., Heron, G., and M. Bocci,
"Pseudowire Status for Static Pseudowires", RFC 6478,
DOI 10.17487/RFC6478, May 2012,
<http://www.rfc-editor.org/info/rfc6478>.
[RFC6870] Muley, P., Ed. and M. Aissaoui, Ed., "Pseudowire
Preferential Forwarding Status Bit", RFC 6870,
DOI 10.17487/RFC6870, February 2013,
<http://www.rfc-editor.org/info/rfc6870>.7.2. Informative References
[RFC3985] Bryant, S., Ed. and P. Pate, Ed., "Pseudo Wire Emulation
Edge-to-Edge (PWE3) Architecture", RFC 3985,
DOI 10.17487/RFC3985, March 2005,
<http://www.rfc-editor.org/info/rfc3985>.
[RFC5085] Nadeau, T., Ed. and C. Pignataro, Ed., "Pseudowire Virtual
Circuit Connectivity Verification (VCCV): A Control
Channel for Pseudowires", RFC 5085, DOI 10.17487/RFC5085,
December 2007, <http://www.rfc-editor.org/info/rfc5085>.
[RFC5659] Bocci, M. and S. Bryant, "An Architecture for Multi-
Segment Pseudowire Emulation Edge-to-Edge", RFC 5659,
DOI 10.17487/RFC5659, October 2009,
<http://www.rfc-editor.org/info/rfc5659>.
[RFC6718] Muley, P., Aissaoui, M., and M. Bocci, "Pseudowire
Redundancy", RFC 6718, DOI 10.17487/RFC6718, August 2012,
<http://www.rfc-editor.org/info/rfc6718>.Acknowledgements 💌
The authors would like to thank Mach Chen, Lizhong Jin, Mustapha Aissaoui, Luca Martini, Matthew Bocci, and Stewart Bryant for their valuable comments and discussions. Real ones, all of them 🫶
Authors' Addresses 🌸
Jie Dong
Huawei Technologies
Huawei Building, No.156 Beiqing Rd.
Beijing 100095
China
Email: jie.dong@huawei.com
Haibo Wang
Huawei Technologies
Huawei Building, No.156 Beiqing Rd.
Beijing 100095
China
Email: rainsword.wang@huawei.comAnd that's the whole spec, bestie 🌈✨ Thanks for sticking with me through all of it 💖
📦 where this page comes from
- ietf: rfc7795 2016-02
- 💖 girlified by claude-haiku-5-5 on : a language model's rewrite of the original, under its terms