<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE rfc SYSTEM "rfc2629.dtd">
<?xml-stylesheet type='text/xsl' href='rfc2629.xslt' ?>
<!-- ?rfc toc="yes"? -->
<?rfc sortrefs="yes"?>
<?rfc symrefs="yes"?>
<?rfc compact="yes"?>
<rfc category="bcp" docName="draft-huston-idr-as4bytes-survey-00.txt" ipr="full3978" >
 <front>
  <title abbrev="4-Byte AS Report">BGP support for 4-Byte AS Numbers - Implementation Survey Report</title>

  <author fullname="Geoff Huston" initials="G." surname="Huston">
      <organization abbrev="APNIC">Asia Pacific Network Information Centre</organization>
      <address>
        <postal></postal>
        <email>gih@apnic.net</email>
        <uri>http://www.apnic.net</uri>
      </address>
  </author>

  <date year="2005" />

  <area>Individual Submission</area>

  <workgroup>Individual Submission</workgroup>
    <abstract>
      <t>

      This document provides a survey of BGP-4 4-Byte AS Number support
      implementations.

      </t>
    </abstract>

  </front>

  <middle>
    <section anchor="intro" title="Survey Summary">
      <t>

      This document provides a survey of BGP-4 4-Byte AS Number Support
      <xref target="ID.4ByteAS" /> implementations. After a brief summary,
      each response is listed. The editor, makes no claim as to the
      accuracy of the information provided.

      </t>
    </section>

    <section title="Summary Forms">
      <section title="Juniper Networks">
      <t>

      Organization: Juniper Networks

      </t><t>


      Person filling out this form:
      
      </t><list><t>
      
      Bruno Rijsman &lt;brijsman@juniper.net&gt;

      </t></list>
      <t>
      
      Implementation:
      
      </t><list><t>
      
      JUNOSe 4-1-0 and later
      
      </t></list>
      <t>
      
      Does the implementation include all parts of the specification:
      
      </t><list><t>
      
      Yes
      
      </t></list>
      <t>

      Are there parts of the specification that are unclear where the
      implementor had to exercise some judgement that may impact
      interoperability?

      </t>
      <list>
      <list style="symbols"><t>

      It isn't clear what to do if the information in the old as-path
      is inconsistent with the information in the new as-path.

      </t><t>
 
      There some places where AS numbers are used where it wasn't
      clear how to deal with 4-octet as-numbers (e.g. extended
      communities).

      </t><t>

      It isn't spelled out that this capability cannot be dynamically
      negotiated.

      </t></list></list>
      <t>
 
      Has there been any interoperability testing?

      </t><list>
      <t>

      Yes; no problems were discovered.
         <vspace blankLines="1" />

      </t><list style="numbers"><t>

      NEW / OLD ineroperability testing with:
       <list style="none">
        <t>
         Juniper ERX (older version which does not support draft)
         <vspace blankLines="0" />
         Juniper M/T/J
         <vspace blankLines="0" />
         Cisco 7500
         </t>
       </list>
         <vspace blankLines="1" />
      </t><t>

      NEW / NEW interoperability testing with:
       <list style="none">
        <t>
         Juniper M/T/J
         <vspace blankLines="0" />
         Redback SmartEdge
         </t>
       </list>
         <vspace blankLines="1" />
      </t><t>

      Most deployed Juniper ERX routers run code which supports
      4-octet AS-numbers (and the feature is enabled by default).
      This provides some confidence that the draft does not cause
      interoperability problems.  Note however that the NEW_AS_PATH
      attribute is not generated unless the AS-path contains at least
      one AS-number greater than 65535 which is -as far as we know-
      not yet the case in the Internet today.

      </t></list></list><t>

      Has there been testing of the interface between this
      implementation and the 2-byte BGP implementation on the NEW
      (4-byte) to OLD (2byte) update path?

      </t>
      <list><t>

      Yes

      </t></list>
      <t>

      Has there been testing of the OLD (2-byte) to NEW (4-byte) path?

      </t><list><t>

      Yes

      </t></list><t>

      Have there been any issues noted with the mechanism to
      reconstruct the 4-byte AS path from the NEW_AS-PATH attribute
      and the 2-byte AS Path on an OLD -NEW BGP update session?

      </t><list><t>

      It isn't clear what to do if the information in the old as-path
      is inconsistent with the information in the new as-path.

      </t></list><t>

      Any other comments regarding the implementation

      </t><list><t>

      Some older versions of Cisco IOS send an unsupported capability
      notification (instead of ignoring the capability) when they
      receive a capability advertisement which they don't recognize
      and which has non-empty data. The 4-octet as-number capability
      is such a capability.  Our implementation recognizes this
      notification and stops automatically stops advertising the
      4-octet as-numbers capability (and others) until the next hard
      clear on the BGP session.

      </t></list>

      </section>

      <section title="Redback">

      <t>

      Organization: Redback

      </t><t>

      Person filling out this form:
      
      </t><list><t>
      
      Albert Tian &lt;tian@redback.com&gt;

      </t></list><t>
      
      Does the implementation include all parts of the specification:
      
      </t><list><t>
      
      Yes

      </t></list><t>

      Are there parts of the specification that are unclear where the
      implementor had to exercise some judgement that may impact
      interoperability?

      </t><list><t>

      No.

      </t></list><t>

      Has there been any interoperability testing?

      </t><list><t>
      
      Yes

      </t></list><t>

      Has there been testing of the interface between this
      implementation and the 2-byte BGP implementation on the NEW
      (4-byte) to OLD (2byte) update path?

      </t><list><t>

      Yes (Cisco: 2-byte; Redback: 4 byte).

      </t></list><t>

      Has there been testing of the OLD (2-byte) to NEW (4-byte) path?

      </t><list><t>

      Yes. (Cisco: 2-byte;  Redback: 4-byte).

      </t></list><t>

      Have there been any issues noted with the mechanism to
      reconstruct the 4-byte AS path from the NEW_AS-PATH attribute
      and the 2-byte AS Path on an OLD -NEW BGP update session?

      </t><list><t>

      No
  
      </t></list><t>

      Have there been any issues noted with the mechanism to
      reconstruct the 4-byte AS path from the NEW_AS-PATH attribute
      and the 2-byte AS Path on an OLD -> NEW BGP update session?

      </t><list><t>

      No.

      </t></list><t>

      Any other comments regarding the implementation

      </t><list><t>

      No

      </t></list>

      </section>
      </section>

    <section title="IANA Considerations">
      <t>

      No IANA considerations are noted in this document

      </t>
    </section>


    <section title="Security Considerations">
      <t>

      Security considerations are documented in <xref target="ID.4ByteAS" />.

      </t>
    </section>
  </middle>

  <back>
    <references title="References">

      <reference anchor="ID.4ByteAS">
        <front>
          <title>BGP support for four-octet AS number space</title>

          <author fullname="Q. Vohra" initials="Q" surname="Vohra">
            <organization>Juniper Networks</organization>
          </author>
          <author fullname="E. Chen" initials="E" surname="Chen">
            <organization>Cisco Systems</organization>
          </author>

          <date month="July" year="2005" />
        </front>

        <seriesInfo name="Work in progress, Internet Draft:"
                    value="draft-ietf-idr-as4bytes-10.txt" />
      </reference>

    </references>
  </back>
</rfc>