RFC 5613
OSPF Link-Local Signaling
OSPF Link-Local Signaling 🎀
Okay bestie, this is RFC 5613, an IETF Standards Track document from the Network Working Group, dated August 2009. It was written by A. Zinin (Alcatel-Lucent), A. Roy, L. Nguyen (Cisco Systems), B. Friedman (Google, Inc.), and D. Yeung (Cisco Systems). It obsoletes RFC 4813, the experimental version, so this is the glow-up that brings it onto the Standards Track 💅✨
Abstract 🌸
OSPF is a link-state intra-domain routing protocol, and OSPF routers swap info on a link using packets with a super well-defined fixed format. The problem is that the format isn't flexible enough for new features that need to send arbitrary data 🙈 This document describes a backward-compatible technique for link-local signaling, which just means exchanging arbitrary data on a link. It also replaces the experimental spec from RFC 4813 and brings it onto the Standards Track 🦋
Status of This Memo 📋
This document specifies an Internet standards track protocol for the Internet community, and it's asking for discussion and suggestions for improvements 💬 For the standardization state and status of this protocol, check the current edition of the "Internet Official Protocol Standards" (STD 1). Distribution of this memo is unlimited 🌈
Copyright Notice 🔒
Copyright (c) 2009 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 in effect on the date of publication of this document (http://trustee.ietf.org/license-info). Please review these documents carefully, because they describe your rights and restrictions with respect to this document 📜
Table of Contents 🗂️
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.1. Requirements Notation . . . . . . . . . . . . . . . . . . 2
2. Proposed Solution . . . . . . . . . . . . . . . . . . . . . . 3
2.1. L-Bit in Options Field . . . . . . . . . . . . . . . . . . 4
2.2. LLS Data Block . . . . . . . . . . . . . . . . . . . . . . 4
2.3. LLS TLVs . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.4. Extended Options and Flags TLV . . . . . . . . . . . . . . 5
2.5. Cryptographic Authentication TLV (OSPFv2 ONLY) . . . . . . 6
2.6. Private TLVs . . . . . . . . . . . . . . . . . . . . . . . 7
3. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 7
4. Compatibility Issues . . . . . . . . . . . . . . . . . . . . . 9
5. Security Considerations . . . . . . . . . . . . . . . . . . . 9
6. References . . . . . . . . . . . . . . . . . . . . . . . . . . 9
6.1. Normative References . . . . . . . . . . . . . . . . . . . 9
6.2. Informative References . . . . . . . . . . . . . . . . . . 10
Appendix A. Acknowledgements . . . . . . . . . . . . . . . . . . 11
Appendix B. Changes from RFC 4813 . . . . . . . . . . . . . . . . 111. Introduction ✨
This document describes an extension to OSPFv2 [OSPFV2] and OSPFv3 [OSPFV3] that lets routers on the same link trade extra info 🫶 OSPFv2 and OSPFv3 packet formats are fixed and can't be extended, so this doc proposes tacking an optional data block onto existing OSPFv2 and OSPFv3 packets. That block is made of Type/Length/Value (TLV) triplets. When the spec applies to both versions, we just say OSPF, and when the text is specific to one, we say OSPFv2 or OSPFv3 💅
One obvious idea would be a brand-new packet type. But that means extra packets floating around the network, which might not be wanted and could cause backward compatibility drama 💀 So this document shows how to exchange data using standard OSPF packet types instead. Iconic restraint 👑
1.1. Requirements Notation 📝
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 [KEY] 🔑
2. Proposed Solution 🌈
To do link-local signaling (LLS), OSPF routers stick a special data block on the end of OSPF packets. If cryptographic authentication is being used, the block goes right after the authentication data block instead 🔐 The LLS block's length doesn't count toward the OSPF packet length, but it does count toward the IPv4/IPv6 packet length. Figure 1 shows how it's attached 🦋
+---------------------+ -- -- +---------------------+
| IP Header | ^ ^ | IPv6 Header |
| Length = HL+X+Y+Z | | Header Length | | Length = HL+X+Y |
| | v v | |
+---------------------+ -- -- +---------------------+
| OSPF Header | ^ ^ | OSPFv3 Header |
| Length = X | | | | Length = X |
|.....................| | X | X |.....................|
| | | | | |
| OSPFv2 Data | | | | OSPFv3 Data |
| | v v | |
+---------------------+ -- -- +---------------------+
| | ^ ^ | |
| Authentication Data | | Y | Y | LLS Data |
| | v v | |
+---------------------+ -- -- +---------------------+
| | ^
| LLS Data | | Z
| | v
+---------------------+ --
Figure 1: LLS Data Block in OSPFv2 and OSPFv3The LLS block MAY be attached to OSPF Hello and Database Description (DD) packets 📦 The LLS block MUST NOT be attached to any other OSPF packet type when generating packets, and MUST be ignored when receiving them 🚫
The data in an LLS block attached to a Hello packet MAY be used for dynamic signaling, since Hello packets can be sent at any time ⏰ That said, delivery of LLS data in Hello packets isn't guaranteed 😬 Data sent with DD packets is guaranteed to be delivered as part of the adjacency forming process ✨
This document doesn't say how routers should interpret the data sent through the LLS mechanism 🤷♀️ Routers that don't understand LLS can still receive these packets, so changes made because of LLS block TLVs don't affect basic routing when talking to non-LLS routers. Very polite of them 💖
2.1. L-Bit in Options Field 🎀
A new L-bit (L stands for LLS) is added to the OSPF Options field (see Figures 2a and 2b) 💜 Routers set the L-bit in Hello and DD packets to show that the packet has an LLS data block. In other words, the LLS data block is only looked at if the L-bit is set 👀
+---+---+---+---+---+---+---+---+
| * | O | DC| L |N/P| MC| E | * |
+---+---+---+---+---+---+---+-+-+
Figure 2a: OSPFv2 Options Field
0 1 2
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+--+-+-+--+--+--+--+--+--+
| | | | | | | | | | | | | | |L|AF|*|*|DC| R| N|MC| E|V6|
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+--+-+-+--+--+--+--+--+--+
Figure 2b: OSPFv3 Options FieldThe L-bit MUST NOT be set except in Hello and DD packets that contain an LLS block 🚨 No sneaking it in anywhere else, bestie 🙅♀️
2.2. LLS Data Block 🍓
The data block used for link-local signaling is formatted as described below. Figure 3 has the visual 🖼️
0 1 2 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Checksum | LLS Data Length |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| |
| LLS TLVs |
. .
. .
. .
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Figure 3: Format of LLS Data BlockThe Checksum field has the standard IP checksum for the entire contents of the LLS block. Before computing it, the checksum field is set to 0 🧮 If the checksum is incorrect, the OSPF packet MUST be processed, but the LLS block MUST be discarded 🗑️ Ouch, but the packet still gets handled, so the routing isn't ruined 😭
The 16-bit LLS Data Length field holds the length (in 32-bit words) of the LLS block, header and payload included 📏
Now, this one's important 👑 If the OSPF packet is cryptographically authenticated, the LLS data block MUST also be cryptographically authenticated. In that case, the regular LLS checksum isn't calculated, and it's just set to 0 instead 🔐
The rest of the block is a set of Type/Length/Value (TLV) triplets, described in Section 2.3. All TLVs MUST be 32-bit aligned (with padding if necessary) 🧩
2.3. LLS TLVs 🦋
The contents of an LLS data block are built out of TLVs. Figure 4 shows the TLV format 💅
The Type field holds the TLV ID, which is unique for each type of TLV. The Length field holds the length of the Value field (in bytes). The Value field is variable and holds arbitrary data 🌸
0 1 2 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Type | Length |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| |
. .
. Value .
. .
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Figure 4: Format of LLS TLVsNote that TLVs are always padded to a 32-bit boundary. Padding bytes aren't counted in the TLV Length field, but they are counted in the LLS Data Length field in the LLS block header 🧠 Unrecognized TLV types are ignored, so no drama there 🫶
Part 1 wraps here, babe 🎀 Part 2 picks up with Section 3 (IANA Considerations) and everything after it, so stay tuned 💖
2.4. Extended Options and Flags TLV ✨
okay so this subsection is all about a TLV called the Extended Options and Flags (EOF) TLV 🎀 The format is shown in Figure 5.
Bits in the Value field have no semantics from the point of view of the LLS mechanism 🦋 Bits MAY be allocated to announce OSPF link-local capabilities. Bits MAY also be allocated to perform boolean link-local signaling. so much room to play, bestie 💜
The Value field in the EOF-TLV is 4 bytes long.
The Type field in the EOF-TLV is 1. The EOF-TLV MUST only appear once in the LLS data block. 🔥
0 1 2 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| 1 | 4 |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Extended Options and Flags |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+Figure 5: Format of the EOF-TLV 🍓
[OOB] and [RESTART] are currently using bits in the Extended Options field of the EOF-TLV 💅 The Extended Options and Flags bits themselves are defined in Section 3.
2.5. Cryptographic Authentication TLV (OSPFv2 ONLY) 🔐
okay this one is the security queen 👑 This document defines a special TLV used for cryptographic authentication (CA-TLV) of the LLS data block. This TLV MUST only be included in the LLS block when cryptographic authentication is enabled on the corresponding interface. The message digest of the LLS block MUST be calculated using the same key and authentication algorithm as used for the OSPFv2 packet. The cryptographic sequence number is included in the TLV and MUST be the same as the one in the OSPFv2 authentication data for the LLS block to be considered authentic. 🫶
The TLV is built like this:
0 1 2 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| 2 | AuthLen |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Sequence Number |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| |
. .
. AuthData .
. .
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+Figure 6: Format of Cryptographic Authentication TLV
The Type field for the CA-TLV has the value 2.
The Length field in the header holds the length of the data portion of the TLV. That includes the 4 bytes for Sequence Number plus the length of the message digest block for the whole LLS block, in bytes. 📏
The Sequence Number field carries the cryptographic sequence number that is used to prevent simple replay attacks 🛡️ For the LLS block to be considered authentic, the Sequence Number in the CA-TLV MUST match the Sequence Number in the OSPFv2 packet header Authentication field (which MUST be present). If there's a Sequence Number mismatch or an Authentication failure, the whole LLS block MUST be ignored. 🙈
The CA-TLV MUST NOT appear more than once in the LLS block. Also, when present, this TLV MUST be the last TLV in the LLS block. If it shows up more than once, only the first occurrence is processed and any others MUST be ignored. no duplicates allowed, periodt 💀
The AuthData field contains the message digest calculated for the LLS data block up to the CA-TLV AuthData field. In other words, it excludes the CA-TLV AuthData itself. 🧚
The CA-TLV is not applicable to OSPFv3, and it MUST NOT be added to any OSPFv3 packet. If it is found on reception, this TLV MUST be ignored. 🚫
2.6. Private TLVs 🌈
LLS type values in the range of 32768-65536 are reserved for private use. The first four octets of the Value field MUST be the private enterprise code [ENTNUM]. This is what lets multiple vendor private extensions coexist in one network, which is honestly iconic 💖
3. IANA Considerations 📋
This document uses the registry that was originally created in [RFC4813]. IANA updated the following registry to point to this document instead:
- "Open Shortest Path First (OSPF) Link-Local Signalling (LLS) - Type/Length/Value Identifiers (TLV)"
IANA allocated the L-bit in the "OSPFv2 Options Registry" and "OSPFv3 Options Registry" as per Section 2.1. ✨
LLS TLV types are maintained by IANA. Extensions to OSPF that require a new LLS TLV type MUST be reviewed by a Designated Expert from the routing area. 🧐
The criteria for allocating LLS TLVs are:
- LLS should not be used for information that would be better suited to be advertised in a link-local link state advertisement (LSA).
- LLS should be confined to signaling between direct neighbors.
- Discretion should be used in the volume of information signaled using LLS due to the obvious MTU and performance implications.
Following the policies outlined in [IANA], LLS type values in the range of 0-32767 are allocated through an IETF Review, and LLS type values in the range of 32768-65535 are reserved for private use. 📜
This document assigns the following LLS TLV types in OSPFv2/OSPFv3:
TLV Type Name Reference
0 Reserved
1 Extended Options and Flags [RFC5613]
2 Cryptographic Authentication+ [RFC5613]
3-32767 Reserved for assignment by the IANA
32768-65535 Private Use
+ Cryptographic Authentication TLV is only defined for OSPFv2IANA renamed the sub-registry from "LLS Type 1 Extended Options" to "LLS Type 1 Extended Options and Flags". 🏷️
This document also assigns the following bits in the EOF-TLV outlined in Section 2.5:
Bit Name Reference
0x00000001 LSDB Resynchronization (LR) [RFC4811]
0x00000002 Restart Signal (RS-bit) [RFC4812]Future allocation of Extended Options and Flags bits MUST be reviewed by a Designated Expert from the routing area. 🦋
4. Compatibility Issues 🤝
The changes to OSPF packet formats are compatible with standard OSPF, since OSPF routers that don't support LLS will ignore the LLS data block after the OSPF packet or cryptographic message digest. As of this writing, there are implementations deployed with [RFC4813]-compliant software. Routers not implementing [RFC4813] ignore the LLS data at the end of the OSPF packet. 🌸
Careful consideration should be given to carrying additional LLS data, as it may affect OSPF adjacency bring-up time due to additional propagation delay and/or processing time. ⏳
5. Security Considerations 🔒
Security considerations inherited from OSPFv2 are described in [OSPFV2]. This technique gives the same level of security as basic OSPFv2 by letting LLS data be authenticated with the same cryptographic authentication that OSPFv2 uses (see Section 2.5 for more details). 🔐
Security considerations inherited from OSPFv3 are described in [OSPFV3] and [OSPFV3AUTH]. OSPFv3 uses IPsec for authentication and encryption. With IPsec, the AH (Authentication Header), ESP (Encapsulating Security Payload), or both are applied to the entire OSPFv3 payload, including the LLS block. 🛡️
This part is serious, bestie: the LLS block is only as safe as the authentication you wrap around it, so don't skip Section 2.5 on OSPFv2 or the IPsec setup on OSPFv3 💀
6. References 📚
6.1. Normative References
- [IANA] Narten, T. and H. Alvestrand, "Guidelines for Writing an IANA Considerations Section in RFCs", BCP 26, RFC 5226, May 2008.
- [KEY] Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", BCP 14, RFC 2119, March 1997.
- [OSPFV2] Moy, J., "OSPF Version 2", STD 54, RFC 2328, April 1998.
- [OSPFV3] Coltun, R., Ferguson, D., Moy, J., and A. Lindem, "OSPF for IPv6", RFC 5340, July 2008.
- [OSPFV3AUTH] Gupta, M. and N. Melam, "Authentication/Confidentiality for OSPFv3", RFC 4552, June 2006.
6.2. Informative References
- [ENTNUM] IANA, "PRIVATE ENTERPRISE NUMBERS", http://www.iana.org.
- [OOB] Nguyen, L., Roy, A., and A. Zinin, "OSPF Out-of-Band Link State Database (LSDB) Resynchronization", RFC 4811, March 2007.
- [RESTART] Nguyen, L., Roy, A., and A. Zinin, "OSPF Restart Signaling", RFC 4812, March 2007.
- [RFC4813] Friedman, B., Nguyen, L., Roy, A., Yeung, D., and A. Zinin, "OSPF Link-Local Signaling", RFC 4813, March 2007.
Appendix A. Acknowledgements 💖
The authors would like to acknowledge Russ White, Acee Lindem, and Manral Vishwas for their review of this document. 🫶
Appendix B. Changes from RFC 4813 🔄
This section describes the substantive changes from [RFC4813].
- Added OSPFv3 support
- Private TLVs MUST use private enterprise code
- Clarified requirement levels at several places
- Changed from Experimental to Standards Track
Authors' Addresses 🎀
Alex Zinin
Alcatel-Lucent
Singapore
EMail: alex.zinin@alcatel-lucent.com
Abhay Roy
Cisco Systems
170 West Tasman Drive
San Jose, CA 95134
USA
EMail: akr@cisco.com
Liem Nguyen
Cisco Systems
170 West Tasman Drive
San Jose, CA 95134
USA
EMail: lhnguyen@cisco.com
Barry Friedman
Google, Inc.
1600 Amphitheatre Parkway
Mountain View, CA 94043
USA
EMail: barryf@google.com
Derek Yeung
Cisco Systems
170 West Tasman Drive
San Jose, CA 95134
USA
EMail: myeung@cisco.com📦 where this page comes from
- ietf: rfc5613 2009-08
- 💖 girlified by claude-haiku-5-5 on : a language model's rewrite of the original, under its terms