Route Leak Detection and Filtering using Roles in Update and Open messages

For the purposes of this document BGP route leaks are when a BGP route was learned from transit provider or peer and is announced to another provider or peer. See . These are usually the result of misconfigured or absent BGP route filtering or lack of coordination between two BGP speakers. describes a method of marking and detecting leaks which relies on operator maintained markings. Unfortunately, in most cases, a leaking router will likely also be misconfigured to mark incorrectly. The proposed mechanism provides an opportunity to detect route leaks made by third parties but provides no support to prevent route leak creation. The leak prevention still relies on communities which are optional and often missed due to mistakes or misunderstanding of the BGP configuration process. It has been suggested to use white list filtering, relying on knowing the prefixes in the customer cone as import filtering, in order to detect route leaks. Unfortunately, a large number of incidents is created medium size transit operators use a single prefix list as only the ACL for export filtering, without community tagging and paying attention to the source of a learned route. So, if they learn a customer's route from their provider or peer - they will announce it in all directions, including other providers or peers. This misconfiguration affects a limited number of prefixes; but such route leaks will obviously bypass customer cone import filtering made by upper level upstream providers. Also, route tagging which relies on operator maintained policy configuration is too easily and too often misconfigured. This document specifies a new BGP Capability Code, Sec 4, which two BGP speakers MAY use to ensure that they MUST agree on their relationship; i.e. customer and provider or peers. Either or both may optionally be configured to require that this option be exchanged for the BGP Open to succeed. Also this document specifies a way to mark routes according to BGP Roles and a way to create double-boundary filters for prevention and detection of route leaks via a two new BGP Path Attributes.

BGP Role is new mandatory configuration option which must be set per each address family. It reflects the real-world agreement between two BGP speakers about their business relationship. Allowed Role values are: Provider - sender is a transit provider to neighbor; Customer - sender is customer of neighbor; Peer - sender and neighbor are peers; Internal - sender is part of an internal AS of an organization which has multiple ASs, is a confederation, ... Complex - sender has non-standard agreement and wants to use manual policies. Since BGP Role reflects the relationship between two BGP speakers, it could also be used for more than route leak mitigation.

The TLV (type, length, value) of the BGP Role capability are: Type - <TBD1>; Length - 1 (octet); Value - integer corresponding to speaker' BGP Role. Value Role name 0 Undefined 1 Sender is Peer 2 Sender is Provider 3 Sender is Customer 4 Sender is Internal 5 Sender is Complex

described how BGP Role is a reflection of the relationship between two BGP speakers. But the mere presence of BGP Role doesn't automatically guarantee role agreement between two BGP peers. To enforce correctness, use the BGP Role check with a set of constrains on how speakers' BGP Roles MUST corresponded. Of course, each speaker MUST announce and accept the BGP Role capability in the BGP OPEN message exchange. If a speaker receives a BGP Role capability, it SHOULD check value of the received capability with its own BGP Role. The allowed pairings are (first a sender's Role, second the receiver's Role): Sender Role Receiver Role Peer Peer Provider Customer Customer Provider Internal Internal Complex Complex In all other cases speaker MUST send a Role Mismatch Notification (code 2, sub-code <TBD2>).

A new BGP configuration option "strict mode" is defined with values of true or false. If set to true, then the speaker MUST refuse to establish a BGP session with peers which do not announce BGP Role capability in their OPEN message. If a speaker rejects a connection, it MUST send a Connection Rejected Notification (Notification with error code 6, subcode 5). By default strict mode SHOULD be set to false for backward compatibility with BGP speakers, that do not yet support this mechanism.

Complex role should be set only if relations between BGP neighbors could not be described using simple Customer/Provider/Peer roles. For a example, if neighbor is literal peer, but for some prefixes it provides full transit, complex role SHOULD be set on both sides. In this case configuration of detection and filtering mechanisms ( and ) should be set on per-prefix basis upon local policy.

The Internal Only To Customer (iOTC) attribute is a new optional, non-transitive BGP Path attribute with the Type Code <TBD3>. This attribute has zero length as it used only as a flag. There are two rules for setting the iOTC attribute: The iOTC attribute MUST be added to all incoming routes if the receiver's Role is Customer or Peer; Routes with the iOTC attribute set MUST NOT be announced if the sender's Role is Customer or Peer; These two rules provide mechanism that prevent route leak creation by an AS. In case of Complex role usage the way of iOTC process is not automated and upon local policy.

The External Only To Customer (eOTC) attribute is a new optional, transitive BGP Path attribute with the Type Code <TBD4>. This attribute has four bytes length and contain an AS number of AS, that added attribute to the route. There are two rules for setting the eOTC attribute: If eOTC is not set and sender's Role is Provider or Peer the eOTC attribute MUST be added with value equal to its ASN. If eOTC is set, receiver's Role is Provider or Peer, and its value is not equal to neighbor ASN then such incoming route is route leak and MUST be given a lower local preference, or they MAY be dropped. These two rules provide mechanism for route leak detection that is made by some party in ASPath. In case of Complex role usage the way of eOTC process is not automated and upon local policy.

In BGPsec enabled routers eOTC attribute MUST be turned into one bit of Flags field of Secure_Path Segment and MUST NOT be added as separate attribute. When route is transmitted from BGPsec enabled router to BGPsec disabled device, in addition to AS_PATH reconstruction MUST be performed eOTC attribute reconstruction. If corresponded bit was set in one of Secure_Path Segments, eOTC attribute SHOULD be added with value that equals to ASN in which segment it appears for the first time.

As BGP Role reflects the relationship between neighbors, it can also have other uses. As an example, BGP Role might affect route priority, or be used to distinguish borders of a network if a network consists of multiple AS. Though such uses may be worthwhile, they are not the goal of this document. Note that such uses would require local policy control. This document doesn't provide any security measures to check correctness of attributes usage in case of Complex role, so Complex role should be set with great caution.

This document defines a new Capability Codes option [to be removed upon publication: http://www.iana.org/assignments/capability-codes/capability-codes.xhtml] , named "BGP Role", assigned value <TBD1> . The length of this capability is 1. The BGP Role capability includes a Value field, for which IANA is requested to create and maintain a new sub-registry called "BGP Role Value". Assignments consist of Value and corresponding Role name. Initially this registry is to be populated with the data in . Future assignments may be made by a standard action procedure . This document defines new subcode, "Role Mismatch", assigned value <TBD2> in the OPEN Message Error subcodes registry [to be removed upon publication: http://www.iana.org/assignments/bgp-parameters/bgp-parameters.xhtml#bgp-parameters-6] . This document defines a new optional, non-transitive BGP Path Attributes option, named "Internal Only To Customer", assigned value <TBD3> [To be removed upon publication: http://www.iana.org/assignments/bgp-parameters/bgp-parameters.xhtml#bgp-parameters-2] . The length of this attribute is 0. This document defines a new optional, transitive BGP Path Attributes option, named "External Only To Customer", assigned value <TBD4> [To be removed upon publication: http://www.iana.org/assignments/bgp-parameters/bgp-parameters.xhtml#bgp-parameters-2] . The length of this attribute is 4.

This document proposes a mechanism for prevention and detection of route leaks, that are the result of BGP policy misconfiguration. That includes preventing route leaks created inside an AS (company), and route leak detection, if a route was leaked by third party. Deliberate sending of a known conflicting BGP Role could be used to sabotage a BGP connection. This is easily detectable. Deliberate mis-marking of the eOTC flag could be used to could affect BGP decision process but could not sabotage a route's propagation. BGP Role is disclosed only to an immediate BGP speaker, so it will not itself reveal any sensitive information to third parties. On the other hand, eOTC is a transitive BGP AS_PATH attribute which reveals a bit about a BGP speaker's business relationship. It will give a strong hint that some link isn't customer to provider, but will not help to distinguish if it is provider to customer or peer to peer. If eOTC is BGPsec signed, it can not be removed for business confidentiality.