RFC 7896

Update to the Include Route Object (IRO) Specification in the Path Computation Element Communication Protocol (PCEP)

๐Ÿ“… published June 2016, from ietf

Update to the Include Route Object (IRO) Specification in the Path Computation Element Communication Protocol (PCEP) โœจ

Okay bestie, here's the title block tea ๐Ÿซ–๐Ÿ’…. This is RFC 7896, written by D. Dhody of Huawei Technologies, dated June 2016. It's a Standards Track document (the ISSN is 2070-1721), and it updates RFC 5440 ๐ŸŽ€. It doesn't obsolete anything.

Here's the vibe: PCEP lets a PCC and a PCE talk, or two PCEs talk to each other ๐Ÿฆ‹. RFC 5440 gave us the Include Route Object (IRO) to list the network elements a computed path has to go through, but it never said whether the IRO is an ordered or unordered list. Plus, it said the L bit had no meaning inside an IRO, which is kinda messy ๐Ÿ™ˆ. This doc fixes both things so implementations can actually interoperate ๐Ÿซถ.

1. Introduction ๐ŸŒธ

PCEP is the protocol for PCC-to-PCE and PCE-to-PCE chats ๐Ÿ’œ. [RFC5440] defines the IRO for specifying the network elements to traverse in the computed path. The spec just doesn't say if the IRO is ordered or unordered, and it says the L bit means nothing in there ๐Ÿ˜ญ.

[RFC5441] shows how you use an IRO to indicate the sequence of domains for inter-domain path computation, which is giving main character energy ๐Ÿ‘‘.

The group realized they needed a standard representation so everyone's on the same page, so interoperability isn't a vibe check ๐Ÿ“.

This document updates the IRO specifications in Section 7.12 of [RFC5440] โœจ.

2. Update in the IRO Specification ๐Ÿงš

Section 7.12 of [RFC5440] says the IRO is an optional object used to specify a set of network elements to traverse in the computed path. It says the L bit in the subobject has no meaning inside an IRO ๐Ÿ’€. It doesn't say whether the IRO is ordered or unordered, so yeah, that's the gap we're filling ๐Ÿ”ฅ.

2.1. Update to RFC 5440 ๐Ÿ’–

The IRO specification gets updated by removing the last line of Section 7.12 of [RFC5440], which currently reads:

"The L bit of such sub-object has no meaning within an IRO."

Also, Section 7.12 of [RFC5440] gets two new statements added at the end of its first paragraph ๐ŸŒˆ:

  • The content of an IRO is an ordered list of subobjects representing a series of abstract nodes (refer to Section 4.3.2 of [RFC3209]).
  • The L bit of an IRO subobject is set based on the loose or strict hop property of the subobject. It is set if the subobject represents a loose hop. If the bit is not set, the subobject represents a strict hop. The interpretation of the L bit is as per Section 4.3.3.1 of [RFC3209] ๐Ÿฆ‹.

3. Operational Considerations ๐ŸŽ€

Because [RFC5440] was kinda unclear, you might run into implementations that always treat IRO subobjects as loose hops ๐Ÿ˜ฌ. When those interwork with an implementation following this document, these things might happen:

  • If a non-conforming (to this document) PCC sends an IRO to a conforming (to this document) PCE, then the PCE may unexpectedly fail to find a path. The PCC may think the IRO subobjects are loose hops, but the PCE reads them as strict hops ๐Ÿ™ƒ.
  • If a conforming PCC sends an IRO containing strict hops to a non-conforming PCE, then the PCE may erroneously return a path that doesn't comply with the requested strict hops, since the PCE reads them all as loose hops ๐Ÿšจ. The PCC may check the returned path and catch the issue, or it may end up using an incorrect path ๐Ÿ˜ฑ.

4. Security Considerations ๐Ÿ”

This update to the IRO specification does not introduce any new security considerations, apart from the ones already mentioned in [RFC5440]. Clarifying the supported IRO ordering or the Loose hop bit handling won't have any negative security impact โœจ.

Worth noting though: PCEP runs over TCP. An analysis of the security issues for routing protocols that use TCP (including PCEP) is provided in [RFC6952] ๐Ÿ›ก๏ธ.

5. References ๐Ÿ“š

5.1. Normative References ๐Ÿ’œ

[RFC3209] Awduche, D., Berger, L., Gan, D., Li, T., Srinivasan, V., and G. Swallow, "RSVP-TE: Extensions to RSVP for LSP Tunnels", RFC 3209, DOI 10.17487/RFC3209, December 2001, <http://www.rfc-editor.org/info/rfc3209>.

[RFC5440] Vasseur, JP., Ed. and JL. Le Roux, Ed., "Path Computation Element (PCE) Communication Protocol (PCEP)", RFC 5440, DOI 10.17487/RFC5440, March 2009, <http://www.rfc-editor.org/info/rfc5440>.

5.2. Informative References ๐ŸŒท

[RFC5441] Vasseur, JP., Ed., Zhang, R., Bitar, N., and JL. Le Roux, "A Backward-Recursive PCE-Based Computation (BRPC) Procedure to Compute Shortest Constrained Inter-Domain Traffic Engineering Label Switched Paths", RFC 5441, DOI 10.17487/RFC5441, April 2009, <http://www.rfc-editor.org/info/rfc5441>.

[RFC6952] Jethanandani, M., Patel, K., and L. Zheng, "Analysis of BGP, LDP, PCEP, and MSDP Issues According to the Keying and Authentication for Routing Protocols (KARP) Design Guide", RFC 6952, DOI 10.17487/RFC6952, May 2013, <http://www.rfc-editor.org/info/rfc6952>.

Acknowledgments ๐Ÿซถ

Special thanks to the PCE chairs for guidance on this work ๐Ÿ‘‘.

Thanks to Francesco Fondelli for suggestions clarifying the L bit usage ๐Ÿ’….

Thanks to Adrian Farrel for the review and comments ๐ŸŒธ.

Thanks to Jonathan Hardwick for shepherding the document and writing the text in Section 3 ๐Ÿงš.

Thanks to Deborah Brungard for comments and for being the responsible AD ๐Ÿฆ‹.

Thanks to Peter Yee for the Gen-ART review ๐Ÿ’–.

Thanks to Alvaro Retana for comments during the IESG review ๐Ÿ™Œ.

Author's Address ๐Ÿ“ฎ

Dhruv Dhody Huawei Technologies Divyashree Techno Park, Whitefield Bangalore, Karnataka 560066 India

Email: dhruv.ietf@gmail.com

๐Ÿ“ฆ where this page comes from