Network Working Group                                            R. Bush
Internet-Draft                                 Internet Initiative Japan
Intended status: BCP                                    October 19, 2011                                       March 8, 2012
Expires: April 21, September 9, 2012

                   BGPsec Operational Considerations


   Deployment of the BGPsec architecture and protocols has many
   operational considerations.  This document attempts to collect and
   present them.  It is expected to evolve as BGPsec is formalized and
   initially deployed.

Requirements Language

   The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
   document are to be interpreted as described in RFC 2119 [RFC2119].

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

   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 April 21, September 9, 2012.

Copyright Notice

   Copyright (c) 2011 2012 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.  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  . . . . . . . . . . . . . . . . . . . . . . . . . 3
   2.  Suggested Reading . . . . . . . . . . . . . . . . . . . . . . . 3
   3.  RPKI Distribution and Maintenance . . . . . . . . . . . . . . . 3
   4.  AS/Router Certificates  . . . . . . . . . . . . . . . . . . . . 4 3
   5.  Within a Network  . . . . . . . . . . . . . . . . . . . . . . . 4
   6.  Considerations for Edge Sites . . . . . . . . . . . . . . . . . 5 4
   7.  Beaconing Considerations  . . . . . . . . . . . . . . . . . . . 5
   8.  Routing Policy  . . . . . . . . . . . . . . . . . . . . . . . . 6
   9. 5
   8.  Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
   10. 6
   9.  Security Considerations . . . . . . . . . . . . . . . . . . . . 8
   11. 6
   10. IANA Considerations . . . . . . . . . . . . . . . . . . . . . . 8
   12. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . 8
   13. 7
   11. References  . . . . . . . . . . . . . . . . . . . . . . . . . . 8
     13.1. 7
     11.1.  Normative References . . . . . . . . . . . . . . . . . . . 8
     13.2. 7
     11.2.  Informative References . . . . . . . . . . . . . . . . . . 8 7
   Author's Address  . . . . . . . . . . . . . . . . . . . . . . . . . 9 8

1.  Introduction

   BGPsec is a new protocol with many operational considerations.  It is
   expected to be deployed incrementally over a number of years.  As
   core BGPsec-capable routers may require large memory and crypto
   assist, and/or modern
   CPUs, it is thought that origin validation based on the RPKI will
   occur over the next two one to five three years and that BGPsec will start to
   deploy late in that window.

   BGPsec relies on widespread propagation of the Resource Public Key
   Infrastructure (RPKI) [I-D.ietf-sidr-arch]. [RFC6480].  How the RPKI is distributed and
   maintained globally and within an operator's infrastructure may be
   different for BGPsec than for origin validation.

   BGPsec need be spoken only by a an AS's eBGP speaking, AKA border,
   routers, and is designed so that it can be used to protect
   announcements which are originated by small edge routers, and this routers.  This has
   special operational considerations.

   Different prefixes have different timing and replay protection

2.  Suggested Reading

   It is assumed that the reader understands BGP, [RFC4271], BGPsec,
   [I-D.lepinski-bgpsec-overview], the RPKI, see [I-D.ietf-sidr-arch], [RFC6480], the RPKI
   Repository Structure, see [I-D.ietf-sidr-repos-struct], [RFC6481], and ROAs, see [I-D.ietf-sidr-roa-format]. [RFC6482].

3.  RPKI Distribution and Maintenance


   All non-ROA considerations in the section on RPKI is a distributed database containing certificates, CRLs,
   manifests, ROAs, Distribution and Ghostbuster Records as
   Maintenance of [I-D.ietf-sidr-pfx-validate] apply.

4.  AS/Router Certificates

   As described in
   [I-D.ietf-sidr-repos-struct].  Policies and considerations for RPKI
   object generation and maintenance are discussed elsewhere.

   A local valid cache containing all RPKI data may [I-D.ymbk-bgpsec-rtr-rekeying] routers MAY be gathered from the
   global distributed database using the rsync protocol capable
   of generating their own public/private key-pairs and a validation
   tool such as rcynic.

   Validated caches may also be created having their
   certificates signed and maintained from other
   validated caches.  Network operators SHOULD take maximum advantage of
   this feature to minimize load on published in the global distributed RPKI

   As RPKI-based origin validation relies on by the availability of RPKI
   data, operators SHOULD locate caches close to routers that require
   these data and services. CA system,
   and/or MAY be given public/private key-pairs by the operator.

   A router can peer with site/operator MAY use a single certificate/key in all their
   routers, one certificate/key per router, or more nearby

   For redundancy, any granularity in

   A large operator, concerned that a router SHOULD peer with more than compromise of one cache at router's key
   would make other routers vulnerable, MAY accept a more complex
   certificate/key distribution burden to reduce this exposure.

   On the
   same time.  Peering with two or more, at least one local and others
   remote, is recommended.

   If an operator trusts upstreams to carry their traffic, they SHOULD
   also trust the RPKI data those upstreams cache, and SHOULD peer with
   those caches.  Note that this places an obligation on those upstreams
   to maintain fresh and reliable caches.

   A transit provider or a network with peers SHOULD validate NLRI in
   announcements made by upstreams, downstreams, and peers.  To minimize
   impact on the global RPKI, they SHOULD fetch from and then revalidate
   data from caches provided by their upstreams.

   An environment where private address space is announced in eBGP the
   operator MAY have private RPKI objects which cover these private
   spaces.  This will require a trust anchor created and owned by that
   environment, see [I-D.ietf-sidr-ltamgmt].

4.  AS/Router Certificates

   A site/operator MAY use a single certificate/key in all their
   routers, one certificate/key per router, or any granularity in

   A large operator, concerned that a compromise of one router's key
   would make many routers vulnerable, MAY accept a more complex
   certificate/key distribution burden to reduce this exposure.

   On the other extreme, an edge site other extreme, an edge site with one or two routers MAY use a
   single certificate/key.

   Routers MAY be capable of generating their own keys and having their
   certificates signed and published in the RPKI by their NOC.  This
   would mean that a router's private key need never leave the router.

5.  Within a Network

   BGPsec is spoken by edge routers in a network, those which border
   other networks/ASs.

   In a fully BGPsec enabled AS, Route Reflectors MUST have BGPsec
   enabled if and only if there are eBGP speakers in their client cone. cone
   (the transitive closure of their customers' customers' customers'

   A BGPsec capable router MAY use the data it receives to influence
   local policy within its network, see Section 8. 7.  In deployment this
   policy should fit into the AS's existing policy, preferences, etc.
   This allows a network to incrementally deploy BGPsec capable border

   eBGP speakers which face more critical peers or up/downstreams would
   be candidates for the earliest deployment.  Both securing one's own
   announcements and validating received announcements should be
   considered in partial deployment.


   As they are not signed, an eBGP listener MUST SHOULD NOT strongly trust non-BGPsec
   unsigned markings such as communities received across a trust

6.  Considerations for Edge Sites

   An edge site which does not provide transit and trusts its
   upstream(s) SHOULD only originate a signed prefix announcement and
   need not validate received announcements.

   BGPsec protocol capability negotiation provides for a speaker signing
   the data it sends but being unable to accept signed data.  Thus a
   smallish edge router may hold only its own signing key(s) and sign
   it's announcement but not receive signed announcements and therefore
   not need to deal with the majority of the RPKI.  Thus such routers
   CPU, RAM, and crypto needs are trivial and additional hardware should
   not be needed.

   As the vast majority (84%) of ASs are stubs, and they announce the
   majority of prefixes, this allows for simpler and cheaper early
   incremental deployment.  It may also mean that edge sites concerned
   with routing security will be attracted to upstreams which support

7.  Beaconing Considerations

   The BGPsec protocol attempts to reduce exposure to replay attacks by
   allowing the route originator to sign an announcement with a validity
   period and re-announce well within that period.

   This re-announcement is termed 'beaconing'.  All timing values are,
   of course, jittered.

   It is only the originator of an NLRI which signs the announcement
   with a validity period.

   To reduce vulnerability to a lost beacon announcement, a router
   SHOULD beacon at a rate somewhat greater than half the signature
   validity period it uses.

   As beaconing places a load on the entire global routing system,
   careful thought MUST be given to any need to beacon frequently.  This
   would be based on a conservative estimation of the vulnerability to a
   replay attack.

   Beacon timing and signature validity periods SHOULD be as follows:

   The Exemplary Citizen:  Prefix originators who are not overly
      concerned about replay attacks might announce with a signature
      validity of multiple weeks and beacon one third of the validity

   Normal Prefix:  Most prefixes SHOULD announce with a signature
      validity of a week and beacon every three days.

   Critical Prefix:  Of course, we all think what we do is critical.
      But prefixes of top level DNS servers, and RPKI publication points
      are actually critical to large swaths of the Internet and are
      therefore tempting targets for replay attacks.  It is suggested
      that the beaconing of these prefixes SHOULD be two to four hours,
      with a signature validity they announce the
   majority of six to twelve hours.

      Note that prefixes, this allows for simpler and cheaper early
   incremental deployment.  It may incur route flap damping (RFD) also mean that edge sites concerned
   with current
      default but deprecated RFD parameters, see [I-D.ymbk-rfd-usable].

8. routing security will be attracted to upstreams which support

7.  Routing Policy

   Unlike origin validation based on the RPKI, BGPsec marks a received
   announcement as Valid or Invalid, there is no NotFound state.  How
   this is used in routing is up to the operator's local policy.  See

   As BGPsec will be rolled out over years and does not allow for
   intermediate non-signing edge routers, coverage will be spotty for a
   long time.  Hence a normal operator's policy SHOULD NOT be overly
   strict, perhaps preferring valid announcements and giving very low
   preference, but still using, invalid announcements.

   A BGPsec speaker validates signed paths at the eBGP edge.

   Local policy on the eBGP edge MAY convey the validation state of a
   BGP signed path through normal local policy mechanisms, e.g. setting
   a BGP community, or modifying a metric value such as local-preference
   or MED.  Some MAY choose to use the large Local-Pref hammer.  Others
   MAY choose to let AS-Path rule and set their internal metric, which
   comes after AS-Path in the BGP decision process.

   Because of possible RPKI version skew, an AS Path which does not
   validate at router R0 might validate at R1.  Therefore, signed paths
   that are invalid and yet propagated (because they are chosen as best
   path) SHOULD have their signatures kept intact and should MUST be signed if
   sent to external BGPsec speakers.

   This implies that updates which a speaker judges to be invalid MAY be
   propagated to iBGP peers.  Therefore, unless local policy ensures
   otherwise, a signed path learned via iBGP MAY be invalid.  If needed,
   the validation state SHOULD be signaled by normal local policy
   mechanisms such as communities or metrics.

   On the other hand, local policy on the eBGP edge might preclude iBGP
   or eBGP announcement of signed AS Paths which are invalid.

   If a

   A BGPsec speaker receives an unsigned path, it receiving a path SHOULD perform origin validation
   per [I-D.ietf-sidr-pfx-validate].

   If it is known that a BGPsec neighbor is not a transparent route
   server, and the router provides a knob to disallow a received pCount
   (prepend count, zero for transparent route servers) of zero, that
   knob SHOULD be applied.

9.  Routers should default to this knob
   disallowing pCount 0.

   To prevent exposure of the internals of BGP Confederations [RFC5065],
   a BGPsec speaker which is a Member-AS of a Confederation MUST NOT
   sign updates sent to another Member-AS of the same Confederation.

8.  Notes

   For protection from attacks replaying BGP data on the order of a day
   or longer old, re-keying routers with new keys (previously)
   provisioned in the RPKI is sufficient.  For one procedure, see

   Like the DNS, the global RPKI presents only a loosely consistent
   view, depending on timing, updating, fetching, etc.  Thus, one cache
   or router may have different data about a particular prefix than
   another cache or router.  There is no 'fix' for this, it is the
   nature of distributed data with distributed caches.

   Operators who manage certificates SHOULD have RPKI Ghostbuster
   Records (see [I-D.ietf-sidr-ghostbusters]), signed indirectly by End
   Entity certificates, for those certificates on which others' routing
   depends for certificate and/or ROA validation.

   As a router must evaluate certificates and ROAs which are time
   dependent, routers' clocks MUST be correct to a tolerance of
   approximately an hour.

   If a router has reason to believe its clock is seriouly seriously incorrect,
   e.g. it has a time earlier than 2011, it SHOULD NOT attempt to
   validate incoming updates.  It SHOULD defer validation until it
   believes it is within reasonable time tolerance.

   Servers should provide time service, such as NTP [RFC5905], to client


9.  Security Considerations

   BGPsec is all about security, routing security.

   The major security considerations for the BGPsec protocol are
   described in [I-D.ietf-sidr-bgpsec-protocol].


10.  IANA Considerations

   This document has no IANA Considerations.

12.  Acknowledgments

   The author wishes to thank the BGPsec design team.


11.  References


11.1.  Normative References

              Lepinski, M., "BGPSEC Protocol Specification",
              draft-ietf-sidr-bgpsec-protocol-01 (work in progress),
              October 2011.

              Bush, R., "The RPKI Ghostbusters Record",
              draft-ietf-sidr-ghostbusters-16 (work in progress),
              December 2011.


              Lepinski, M., Kent, S., M. and D. Kong, "A Profile for Route
              Origin Authorizations (ROAs)",
              draft-ietf-sidr-roa-format-12 S. Turner, "An Overview of BGPSEC",
              draft-lepinski-bgpsec-overview-00 (work in progress),
              March 2011.

   [RFC2119]  Bradner, S., "Key words for use in RFCs to Indicate
              Requirement Levels", BCP 14, RFC 2119, March 1997.

13.2.  Informative References


   [RFC6480]  Lepinski, M. and S. Kent, "An Infrastructure to Support
              Secure Internet Routing", draft-ietf-sidr-arch-13 (work in
              progress), May 2011.

              Reynolds, M. RFC 6480, February 2012.

   [RFC6481]  Huston, G., Loomans, R., and S. Kent, "Local Trust Anchor Management G. Michaelson, "A Profile for the
              Resource Public Key Infrastructure",
              draft-ietf-sidr-ltamgmt-02 (work in progress), June 2011. Certificate Repository Structure", RFC 6481,
              February 2012.

   [RFC6482]  Lepinski, M., Kent, S., and D. Kong, "A Profile for Route
              Origin Authorizations (ROAs)", RFC 6482, February 2012.

11.2.  Informative References

              Mohapatra, P., Scudder, J., Ward, D., Bush, R., and R.
              Austein, "BGP Prefix Origin Validation",
              draft-ietf-sidr-pfx-validate-03 (work in progress),
              October 2011.

              Huston, G., Loomans,

              Gagliano, R., Patel, K., and G. Michaelson, "A Profile for
              Resource Certificate Repository Structure",
              draft-ietf-sidr-repos-struct-09 (work in progress),
              July 2011.

              Lepinski, M. and S. Turner, "An Overview of BGPSEC",
              draft-lepinski-bgpsec-overview-00 B. Weis, "BGPSEC router key
              roll-over as an alternative to beaconing",
              draft-rogaglia-sidr-bgpsec-rollover-00 (work in progress),
              March 2011.

              Pelsser, C., Bush, R., 2012.

              Turner, S., Patel, K., Mohapatra, P., and O.
              Maennel, "Making Route Flap Damping Usable",
              draft-ymbk-rfd-usable-01 R. Bush, "Router Keying for
              BGPsec", draft-ymbk-bgpsec-rtr-rekeying-00 (work in
              progress), June 2011. March 2012.

   [RFC4271]  Rekhter, Y., Li, T., and S. Hares, "A Border Gateway
              Protocol 4 (BGP-4)", RFC 4271, January 2006.

   [RFC5065]  Traina, P., McPherson, D., and J. Scudder, "Autonomous
              System Confederations for BGP", RFC 5065, August 2007.

   [RFC5905]  Mills, D., Martin, J., Burbank, J., and W. Kasch, "Network
              Time Protocol Version 4: Protocol and Algorithms
              Specification", RFC 5905, June 2010.

Author's Address

   Randy Bush
   Internet Initiative Japan
   5147 Crystal Springs
   Bainbridge Island, Washington  98110

   Phone: +1 206 780 0431 x1