rfc8821xml2.original.xml   rfc8821.xml 
<?xml version="1.0" encoding="US-ASCII"?> <?xml version='1.0' encoding='utf-8'?>
<!-- edited with XMLSPY v5 rel. 3 U (http://www.xmlspy.com) <!DOCTYPE rfc SYSTEM "rfc2629-xhtml.ent">
by Daniel M Kohn (private) --> <rfc xmlns:xi="http://www.w3.org/2001/XInclude" docName="draft-ietf-teas-pce-nat
<!DOCTYPE rfc SYSTEM "rfc2629.dtd" [ ive-ip-17" number="8821" ipr="trust200902" obsoletes="" updates="" submissionTyp
<!ENTITY rfc2119 PUBLIC "" "http://xml.resource.org/public/rfc/bibxml/reference. e="IETF" category="info" consensus="true" xml:lang="en" tocInclude="true" tocDep
RFC.2119.xml"> th="3" symRefs="true" sortRefs="true" version="3">
]> <!-- xml2rfc v2v3 conversion 3.5.0 -->
<?rfc toc="yes"?>
<?rfc tocompact="yes"?>
<?rfc tocdepth="3"?>
<?rfc tocindent="yes"?>
<?rfc symrefs="yes"?>
<?rfc sortrefs="yes"?>
<?rfc comments="yes"?>
<?rfc inline="yes"?>
<?rfc compact="yes"?>
<?rfc subcompact="no"?>
<rfc category="info" docName="draft-ietf-teas-pce-native-ip-17"
ipr="trust200902">
<front> <front>
<title abbrev="pce in native ip">Path Computation Element (PCE) based <title abbrev="PCE in Native IP">PCE-Based
Traffic Engineering (TE) in Native IP Networks</title> Traffic Engineering (TE) in Native IP Networks</title>
<seriesInfo name="RFC" value="8821"/>
<author fullname="Aijun Wang" initials="A" surname="Wang"> <author fullname="Aijun Wang" initials="A" surname="Wang">
<organization>China Telecom</organization> <organization>China Telecom</organization>
<address> <address>
<postal> <postal>
<street>Beiqijia Town, Changping District</street> <street>Beiqijia Town</street>
<extaddr>Changping District</extaddr>
<city>Beijing</city> <city>Beijing</city>
<region/>
<code>102209</code> <code>102209</code>
<country>China</country> <country>China</country>
</postal> </postal>
<email>wangaj3@chinatelecom.cn</email> <email>wangaj3@chinatelecom.cn</email>
</address> </address>
</author> </author>
<author fullname="Boris Khasanov" initials="B" surname="Khasanov"> <author fullname="Boris Khasanov" initials="B" surname="Khasanov">
<organization>Yandex LLC</organization> <organization>Yandex LLC</organization>
<address> <address>
<postal> <postal>
<street/>
<street>Ulitsa Lva Tolstogo 16</street> <street>Ulitsa Lva Tolstogo 16</street>
<city>Moscow</city> <city>Moscow</city>
<country>Russian Federation</country>
<region/>
<code/>
<country>Russia</country>
</postal> </postal>
<email>bhassanov@yahoo.com</email> <email>bhassanov@yahoo.com</email>
</address> </address>
</author> </author>
<author fullname="Quintin Zhao" initials="Q" surname="Zhao"> <author fullname="Quintin Zhao" initials="Q" surname="Zhao">
<organization>Etheric Networks</organization> <organization>Etheric Networks</organization>
<address> <address>
<postal> <postal>
<street>1009 S CLAREMONT ST</street> <street>1009 S Claremont St</street>
<city>San Mateo</city>
<street/>
<city>SAN MATEO</city>
<region>CA</region> <region>CA</region>
<code>94402</code> <code>94402</code>
<country>United States of America</country>
<country>USA</country>
</postal> </postal>
<email>qzhao@ethericnetworks.com</email> <email>qzhao@ethericnetworks.com</email>
</address> </address>
</author> </author>
<author fullname="Huaimo Chen" initials="H" surname="Chen"> <author fullname="Huaimo Chen" initials="H" surname="Chen">
<organization>Futurewei</organization> <organization>Futurewei</organization>
<address> <address>
<postal> <postal>
<street/>
<city>Boston</city> <city>Boston</city>
<region>MA</region> <region>MA</region>
<code/>
<country>USA</country> <country>USA</country>
</postal> </postal>
<phone/>
<facsimile/>
<email>huaimo.chen@futurewei.com</email> <email>huaimo.chen@futurewei.com</email>
<uri/>
</address> </address>
</author> </author>
<date month="April" year="2021"/>
<date day="2" month="February" year="2021"/>
<area>RTG Area</area> <area>RTG Area</area>
<workgroup>TEAS Working Group</workgroup> <workgroup>TEAS Working Group</workgroup>
<keyword>RFC</keyword>
<abstract> <abstract>
<t>This document defines an architecture for providing traffic <t>This document defines an architecture for providing traffic
engineering in a native IP network using multiple BGP sessions and a engineering in a native IP network using multiple BGP sessions and a
Path Computation Element (PCE)-based central control mechanism. It Path Computation Element (PCE)-based central control mechanism. It
defines the Central Control Dynamic Routing (CCDR) procedures and defines the Centralized Control Dynamic Routing (CCDR) procedures and
identifies needed extensions for the Path Computation Element identifies needed extensions for the Path Computation Element
Communication Protocol (PCEP).</t> Communication Protocol (PCEP).</t>
</abstract> </abstract>
</front> </front>
<middle> <middle>
<section anchor="intro" title="Introduction"> <section anchor="intro">
<t><xref target="RFC8283"/>, based on an extension of the Path <name>Introduction</name>
Computation Element (PCE) architecture described in <xref <t><xref target="RFC8283"/>, based on an extension of the
target="RFC4655"/> , introduced a broader use applicability for a PCE as PCE architecture described in <xref target="RFC4655"/>, introduced a broad
a central controller. PCEP Protocol (PCEP) continues to be used as the er use applicability for a PCE as
protocol between PCE and Path Computation Client (PCC). Building on that a central controller. PCEP continues to be used as the
work, this document describes a solution using a PCE for centralized protocol between the PCE and the Path Computation Client (PCC). Building o
control in a native IP network to provide End-to-End (E2E) performance n that
work, this document describes a solution of using a PCE for centralized
control in a native IP network to provide end-to-end (E2E) performance
assurance and QoS for traffic. The solution combines the use of assurance and QoS for traffic. The solution combines the use of
distributed routing protocols and a centralized controller, referred to distributed routing protocols and a centralized controller, referred to
as Centralized Control Dynamic Routing (CCDR).</t> as Centralized Control Dynamic Routing (CCDR).</t>
<t><xref target="RFC8735"/> describes the scenarios and simulation <t><xref target="RFC8735"/> describes the scenarios and simulation
results for traffic engineering in a native IP network based on use of a results for traffic engineering in a native IP network based on use of a
CCDR architecture. Per <xref target="RFC8735"/>, the architecture for CCDR architecture. Per <xref target="RFC8735"/>, the architecture for
traffic engineering in a native IP network should meet the following traffic engineering in a native IP network should meet the following
criteria:</t> criteria:</t>
<ul>
<t><list style="symbols"> <li>Same solution for native IPv4 and IPv6 traffic.</li>
<t>Same solution for native IPv4 and IPv6 traffic.</t> <li>Support for intra-domain and inter-domain scenarios.</li>
<li>Achieve E2E traffic assurance, with determined QoS
<t>Support for intra-domain and inter-domain scenarios.</t>
<t>Achieve End to End traffic assurance, with determined QoS
behavior, for traffic requiring a service assurance (prioritized behavior, for traffic requiring a service assurance (prioritized
traffic).</t> traffic).</li>
<li>No changes in a router's forwarding behavior.</li>
<t>No changes in a router's forwarding behavior.</t> <li>Based on centralized control through a distributed network
control plane.</li>
<t>Based on centralized control through a distributed network <li>Support different network requirements such as high traffic
control plane.</t> volume and prefix scaling.</li>
<li>Ability to adjust the optimal path dynamically upon the changes
<t>Support different network requirements such as high traffic of network status. No need for reserving resources for physical links
volume and prefix scaling.</t> in advance.</li>
</ul>
<t>Ability to adjust the optimal path dynamically upon the changes
of network status. No need for physical links resources reservations
to be done in advance.</t>
</list></t>
<t>Building on the above documents, this document defines an <t>Building on the above documents, this document defines an
architecture meeting these requirements by using a multiple BGP session architecture meeting these requirements by using a strategy of multiple BG
strategy and a PCE as the centralized controller. The architecture P sessions
depends on the central control (PCE) element to compute the optimal and a PCE as the centralized controller. The architecture
path, and utilizes the dynamic routing behavior of IGP/BGP protocols for depends on the central control element (PCE) to compute the optimal
path and utilizes the dynamic routing behavior of IGP and BGP for
forwarding the traffic.</t> forwarding the traffic.</t>
</section> </section>
<section>
<section title="Terminology"> <name>Terminology</name>
<t>This document uses the following terms defined in <xref <t>This document uses the following terms defined in <xref target="RFC5440
target="RFC5440"/>:</t> "/>:</t>
<dl>
<t><list style="symbols"> <dt>PCE:</dt>
<t>PCE: Path Computation Element</t> <dd>Path Computation Element</dd>
<dt>PCEP:</dt>
<t>PCEP: PCE Protocol</t> <dd>PCE Protocol</dd>
<dt>PCC:</dt>
<t>PCC: Path Computation Client</t> <dd>Path Computation Client</dd>
</list></t> </dl>
<t>Other terms are used in this document:</t> <t>Other terms are used in this document:</t>
<dl>
<t><list style="symbols"> <dt>CCDR:</dt>
<t>CCDR: Central Control Dynamic Routing</t> <dd>Centralized Control Dynamic Routing</dd>
<dt>E2E:</dt>
<t>E2E: End to End</t> <dd>End to End</dd>
<dt>ECMP:</dt>
<t>ECMP: Equal-Cost Multipath</t> <dd>Equal-Cost Multipath</dd>
<dt>RR:</dt>
<t>RR: Route Reflector</t> <dd>Route Reflector</dd>
<dt>SDN:</dt>
<t>SDN: Software Defined Network</t> <dd>Software-Defined Network</dd>
</list></t> </dl>
</section> </section>
<section>
<section title="CCDR Architecture in Simple Topology"> <name>CCDR Architecture in a Simple Topology</name>
<t>Figure 1 illustrates the CCDR architecture for traffic engineering in <t><xref target="fig-ccdr-arch-simple"/> illustrates the CCDR architecture
a simple topology. The topology is composed of four devices which are for traffic engineering in
SW1, SW2, R1, R2. There are multiple physical links between R1 and R2. a simple topology. The topology is composed of four devices, which are
Traffic between prefix PF11(on SW1) and prefix PF21(on SW2) is normal SW1, SW2, R1, and R2. There are multiple physical links between R1 and R2.
traffic, traffic between prefix PF12(on SW1) and prefix PF22(on SW2) is Traffic between prefix PF11 (on SW1) and prefix PF21 (on SW2) is normal
traffic; traffic between prefix PF12 (on SW1) and prefix PF22 (on SW2) is
priority traffic that should be treated accordingly.</t> priority traffic that should be treated accordingly.</t>
<figure anchor="fig-ccdr-arch-simple">
<figure> <name>CCDR Architecture in a Simple Topology</name>
<artwork align="center"><![CDATA[ +-----+ <artwork name="" type="" alt="">
+-----+
+----------+ PCE +--------+ +----------+ PCE +--------+
| +-----+ | | +-----+ |
| | | |
| BGP Session 1(lo11/lo21)| | BGP Session 1(lo11/lo21)|
+-------------------------+ +-------------------------+
| | | |
| BGP Session 2(lo12/lo22)| | BGP Session 2(lo12/lo22)|
+-------------------------+ +-------------------------+
PF12 | | PF22 PF12 | | PF22
PF11 | | PF21 PF11 | | PF21
+---+ +-----+-----+ +-----+-----+ +---+ +---+ +-----+-----+ +-----+-----+ +---+
|SW1+---------+(lo11/lo12)+-------------+(lo21/lo22)+--------------+SW2| |SW1+---------+(lo11/lo12)+-------------+(lo21/lo22)+-----------+SW2|
+---+ | R1 +-------------+ R2 | +---+ +---+ | R1 +-------------+ R2 | +---+
+-----------+ +-----------+ +-----------+ +-----------+
---+ | R1 +-------------+ R2 | +---+
Figure 1: CCDR architecture in simple topology </artwork>
]]></artwork>
</figure> </figure>
<t>In the intra-domain scenario, IGP and BGP combined with a PCE are
<t/> deployed between R1 and R2. In the inter-domain scenario, only native
BGP is deployed. The traffic between each address pair may
<t>In the Intra-AS scenario, IGP and BGP combined with a PCE are
deployed between R1 and R2. In the inter-AS scenario, only the native
BGP protocol is deployed. The traffic between each address pair may
change in real time and the corresponding source/destination addresses change in real time and the corresponding source/destination addresses
of the traffic may also change dynamically.</t> of the traffic may also change dynamically.</t>
<t>The key ideas of the CCDR architecture for this simple topology are <t>The key ideas of the CCDR architecture for this simple topology are
the following:</t> the following:</t>
<ul>
<t><list style="symbols"> <li>Build two BGP sessions between R1 and R2 via the different
<t>Build two BGP sessions between R1 and R2, via the different
loopback addresses on these routers (lo11 and lo12 are the loopback loopback addresses on these routers (lo11 and lo12 are the loopback
address of R1, lo21 and lo22 are the loopback address of R2).</t> addresses of R1, and lo21 and lo22 are the loopback addresses of R2).<
/li>
<t>Using the PCE, set the explicit peer route on R1 and R2 for BGP <li>Using the PCE, set the explicit peer route on R1 and R2 for BGP
next hop to different physical link addresses between R1 and R2. The next hop to different physical link addresses between R1 and R2. The
explicit peer route can be set in the format of a static route, explicit peer route can be set in the format of a static route,
which is different from the route learned from the IGP protocol.</t> which is different from the route learned from IGP.</li>
<li>Send different prefixes via the established BGP sessions. For
<t>Send different prefixes via the established BGP sessions. For
example, send PF11/PF21 via the BGP session 1 and PF12/PF22 via the example, send PF11/PF21 via the BGP session 1 and PF12/PF22 via the
BGP session 2.</t> BGP session 2.</li>
</list></t> </ul>
<t>After the above actions, the bidirectional traffic between the PF11
<t>After the above actions, the bi-directional traffic between the PF11 and PF21, and the bidirectional traffic between PF12 and PF22, will go
and PF21, and the bi-directional traffic between PF12 and PF22 will go
through different physical links between R1 and R2.</t> through different physical links between R1 and R2.</t>
<t>If there is more traffic between PF12 and PF22 that needs assured <t>If there is more traffic between PF12 and PF22 that needs assured
transport, one can add more physical links between R1 and R2 to reach transport, one can add more physical links between R1 and R2 to reach
the next hop for BGP session 2. In this case, the prefixes that are the next hop for BGP session 2. In this case, the prefixes that are
advertised by the BGP peers need not be changed.</t> advertised by the BGP peers need not be changed.</t>
<t>If, for example, there is bidirectional priority traffic from
<t>If, for example, there is bi-directional priority traffic from another address pair (for example, prefix PF13/PF23), and the total
another address pair (for example prefix PF13/PF23), and the total
volume of priority traffic does not exceed the capacity of the volume of priority traffic does not exceed the capacity of the
previously provisioned physical links, one need only advertise the newly previously provisioned physical links, one need only advertise the newly
added source/destination prefixes via the BGP session 2. The added source/destination prefixes via the BGP session 2. The
bi-directional traffic between PF13/PF23 will go through the same bidirectional traffic between PF13/PF23 will go through the same
assigned dedicated physical links as the traffic between PF12/PF22.</t> assigned, dedicated physical links as the traffic between PF12/PF22.</t>
<t>Such a decoupling philosophy of the IGP/BGP traffic link and the <t>Such a decoupling philosophy of the IGP/BGP traffic link and the
physical link achieves a flexible control capability for the network physical link achieves a flexible control capability for the network
traffic, satisfying the needed QoS assurance to meet the application's traffic, satisfying the needed QoS assurance to meet the application's
requirement. The router needs only support native IP and multiple BGP requirement. The router needs only to support native IP and multiple BGP
sessions setup via different loopback addresses.</t> sessions set up via different loopback addresses.</t>
<t/>
</section> </section>
<section>
<section title="CCDR Architecture in Large Scale Topology"> <name>CCDR Architecture in a Large-Scale Topology</name>
<t>When the priority traffic spans a large-scale network, such as that <t>When the priority traffic spans a large-scale network, such as that
illustrated in Figure 2, the multiple BGP sessions cannot be established illustrated in <xref target="fig-ccdr-arch-large"/>, the multiple BGP sess
hop by hop within one AS. For such a scenario, we propose using a Route ions cannot be established
hop by hop within one autonomous system. For such a scenario, we propose u
sing a Route
Reflector (RR) <xref target="RFC4456"/> to achieve a similar effect. Reflector (RR) <xref target="RFC4456"/> to achieve a similar effect.
Every edge router will establish two BGP sessions with the RR via Every edge router will establish two BGP sessions with the RR via
different loopback addresses respectively. The other steps for traffic different loopback addresses respectively. The other steps for traffic
differentiation are the same as that described in the CCDR architecture differentiation are the same as that described in the CCDR architecture
for the simple topology.</t> for the simple topology.</t>
<t>As shown in <xref target="fig-ccdr-arch-large"/>, if we select R3 as th
<t>As shown in Figure 2, if we select R3 as the RR, every edge router(R1 e RR, every edge router (R1
and R7 in this example) will build two BGP session with the RR. If the and R7 in this example) will build two BGP sessions with the RR. If the
PCE selects the dedicated path as R1-R2-R4-R7, then the operator should PCE selects the dedicated path as R1-R2-R4-R7, then the operator should
set the explicit peer routes via PCEP protocol on these routers set the explicit peer routes via PCEP on these routers
respectively, pointing to the BGP next hop (loopback addresses of R1 and respectively, pointing to the BGP next hop (loopback addresses of R1 and
R7, which are used to send the prefix of the priority traffic) to the R7, which are used to send the prefix of the priority traffic) to the
selected forwarding address.</t> selected forwarding address.</t>
<figure anchor="fig-ccdr-arch-large">
<figure align="right"> <name>CCDR Architecture in a Large-Scale Network</name>
<artwork><![CDATA[ +-----+ <artwork name="" type="" alt="">
+-----+
+----------------+ PCE +------------------+ +----------------+ PCE +------------------+
| +--+--+ | | +--+--+ |
| | | | | |
| | | | | |
| +--+---+ | | +--+---+ |
+----------------+R3(RR)+-----------------+ +----------------+R3(RR)+-----------------+
PF12 | +--+---+ | PF22 PF12 | +--+---+ | PF22
PF11 | | PF21 PF11 | | PF21
+---+ ++-+ +--+ +--+ +-++ +---+ +---+ ++-+ +--+ +--+ +-++ +---+
|SW1+-------+R1+----------+R5+----------+R6+---------+R7+--------+SW2| |SW1+-------+R1+----------+R5+----------+R6+---------+R7+-------+SW2|
+---+ ++-+ +--+ +--+ +-++ +---+ +---+ ++-+ +--+ +--+ +-++ +---+
| | | |
---+ ++-+ +--+ +--+ +-++ +---+
| | | |
| +--+ +--+ | | +--+ +--+ |
+------------+R2+----------+R4+-----------+ +------------+R2+----------+R4+-----------+
+--+ +--+ +--+ +--+
Figure 2: CCDR architecture in large-scale network </artwork>
]]></artwork>
</figure> </figure>
</section> </section>
<section>
<section title="CCDR Multiple BGP Sessions Strategy"> <name>CCDR Multiple BGP Sessions Strategy</name>
<t>Generally, different applications may require different QoS criteria, <t>Generally, different applications may require different QoS criteria,
which may include:</t> which may include:</t>
<ul>
<t><list style="symbols"> <li>Traffic that requires low latency and is not sensitive to packet
<t>Traffic that requires low latency and is not sensitive to packet loss.</li>
loss.</t> <li>Traffic that requires low packet loss and can endure higher
latency.</li>
<t>Traffic that requires low packet loss and can endure higher <li>Traffic that requires low jitter.</li>
latency.</t> </ul>
<t>These different traffic requirements are summarized in <xref target="ta
<t>Traffic that requires low jitter.</t> b-traffic-req"/>.</t>
</list>These different traffic requirements can be summarized in the <table anchor="tab-traffic-req">
following table:</t> <name>Traffic Requirement Criteria</name>
<thead>
<t><figure align="right"> <tr>
<artwork><![CDATA[ +----------------+-------------+-------------- <th>Prefix Set No.</th>
-+-----------------+ <th>Latency</th>
| Prefix Set No. | Latency | Packet Loss | Jitter | <th>Packet Loss</th>
+----------------+-------------+---------------+-----------------+ <th>Jitter</th>
| 1 | Low | Normal | Don't care | </tr>
+----------------+-------------+---------------+-----------------+ </thead>
| 2 | Normal | Low | Don't care | <tbody>
+----------------+-------------+---------------+-----------------+ <tr>
| 3 | Normal | Normal | Low | <td align="center">1</td>
+----------------+-------------+---------------+-----------------+ <td>Low</td>
Table 1. Traffic Requirement Criteria <td>Normal</td>
]]></artwork> <td>Don't care</td>
</figure></t> </tr>
<tr>
<td align="center">2</td>
<td>Normal</td>
<td>Low</td>
<td>Don't care</td>
</tr>
<tr>
<td align="center">3</td>
<td>Normal</td>
<td>Normal</td>
<td>Low</td>
</tr>
</tbody>
</table>
<t>For Prefix Set No.1, we can select the shortest distance path to <t>For Prefix Set No.1, we can select the shortest distance path to
carry the traffic; for Prefix Set No.2, we can select the path that has carry the traffic; for Prefix Set No.2, we can select the path that has
end to end under-loaded links; for Prefix Set No.3, we can let traffic E2E under-loaded links; for Prefix Set No.3, we can let traffic
pass over a determined single path, as no Equal Cost Multipath (ECMP) pass over a determined single path, as no ECMP
distribution on the parallel links is desired.</t> distribution on the parallel links is desired.</t>
<t>It is almost impossible to provide an E2E path
<t>It is almost impossible to provide an End-to-End (E2E) path
efficiently with latency, jitter, and packet loss constraints to meet efficiently with latency, jitter, and packet loss constraints to meet
the above requirements in a large-scale IP-based network only using a the above requirements in a large-scale, IP-based network only using a
distributed routing protocol, but these requirements can be met with the distributed routing protocol, but these requirements can be met with the
assistance of PCE, as that described in <xref target="RFC4655"/> and assistance of PCE, as described in <xref target="RFC4655"/> and
<xref target="RFC8283"/>. The PCE will have the overall network view, <xref target="RFC8283"/>. The PCE will have the overall network view,
ability to collect the real-time network topology, and the network ability to collect the real-time network topology, and the network
performance information about the underlying network. The PCE can select performance information about the underlying network. The PCE can select
the appropriate path to meet the various network performance the appropriate path to meet the various network performance
requirements for different traffic.</t> requirements for different traffic.</t>
<t>The architecture to implement the CCDR multiple BGP sessions strategy
<t>The architecture to implement the CCDR Multiple BGP sessions strategy
is as follows:</t> is as follows:</t>
<t>The PCE will be responsible for the optimal path computation for the <t>The PCE will be responsible for the optimal path computation for the
different priority classes of traffic:</t> different priority classes of traffic:</t>
<ul>
<t><list style="symbols"> <li>PCE collects topology information via BGP-LS <xref target="RFC7752">
<t>PCE collects topology information via BGP-LS<xref </xref> and link utilization information via the
target="RFC7752"> </xref> and link utilization information via the
existing Network Monitoring System (NMS) from the underlying existing Network Monitoring System (NMS) from the underlying
network.</t> network.</li>
<li>PCE calculates the appropriate path based upon the application's
<t>PCE calculates the appropriate path based upon the application's requirements and sends the key parameters to edge/RR routers (R1, R7,
requirements, and sends the key parameters to edge/RR routers(R1, R7 and R3 in <xref target="fig-ccdr-arch-multi"/>) to establish multiple
and R3 in Figure 3) to establish multiple BGP sessions. The loopback BGP sessions. The loopback
addresses used for the BGP sessions should be planned in advance and addresses used for the BGP sessions should be planned in advance and
distributed in the domain.</t> distributed in the domain.</li>
<li>PCE sends the route information to the routers (R1, R2, R4, and R7 i
<t>PCE sends the route information to the routers (R1,R2,R4,R7 in n
Figure 3) on the forwarding path via PCEP, to build the path to the <xref target="fig-ccdr-arch-multi"/>) on the forwarding path via PCEP
BGP next-hop of the advertised prefixes. The path to these BGP to build the path to the
next-hop will also be learned via the IGP protocol, but the route BGP next hop of the advertised prefixes. The path to these BGP
from the PCEP has the higher preference. Such design can assure the next hops will also be learned via IGP, but the route
IGP path to the BGP next-hop can be used to protect the path from the PCEP has the higher preference. Such a design can assure the
assigned by PCE.</t> IGP path to the BGP next hop can be used to protect the path
assigned by PCE.</li>
<t>PCE sends the prefixes information to the PCC(edge routers that <li>PCE sends the prefix information to the PCC (edge routers that
have established BGP sessions) for advertising different prefixes have established BGP sessions) for advertising different prefixes
via the specified BGP session.</t> via the specified BGP session.</li>
<li>The priority traffic may share some links or nodes if the path the
<t>The priority traffic may share some links or nodes, if path the
shared links or nodes can meet the requirement of application. When shared links or nodes can meet the requirement of application. When
the priority traffic prefixes were changed but the total volume of the priority traffic prefixes are changed, but the total volume of
priority traffic does not exceed the physical capacity of the priority traffic does not exceed the physical capacity of the
previous E2E path, the PCE needs only change the prefixed advertised previous E2E path, the PCE needs only change the prefixes advertised
via the edge routers (R1,R7 in Figure 3).</t> via the edge routers (R1 and R7 in <xref target="fig-ccdr-arch-multi"/
>).</li>
<t>If the volume of priority traffic exceeds the capacity of the <li>If the volume of priority traffic exceeds the capacity of the
previous calculated path, the PCE can recalculate and add the previous calculated path, the PCE can recalculate and add the
appropriate paths to accommodate the exceeding traffic. After that, appropriate paths to accommodate the exceeding traffic. After that,
the PCE needs to update the on-path routers to build the forwarding the PCE needs to update the on-path routers to build the forwarding
path hop by hop.</t> path hop by hop.</li>
</list><figure align="right"> </ul>
<artwork><![CDATA[ +------------+ <figure anchor="fig-ccdr-arch-multi">
<name>CCDR Architecture for Multi-BGP Sessions Deployment</name>
<artwork name="" type="" alt="">
+------------+
| Application| | Application|
+------+-----+ +------+-----+
| |
+--------+---------+ +--------+---------+
+----------+SDN Controller/PCE+-----------+ +----------+SDN Controller/PCE+-----------+
| +--------^---------+ | | +--------^---------+ |
| | | | | |
| | | | | |
PCEP | BGP-LS|PCEP | PCEP PCEP | BGP-LS|PCEP | PCEP
| | | | | |
| +--v---+ | | +--v---+ |
+----------------+R3(RR)+-----------------+ +----------------+R3(RR)+-----------------+
PF12 | +------+ | PF22 PF12 | +------+ | PF22
PF11 | | PF21 PF11 | | PF21
+---+ +v-+ +--+ +--+ +-v+ +---+ +---+ +v-+ +--+ +--+ +-v+ +---+
|SW1+-------+R1+----------+R5+----------+R6+---------+R7+--------+SW2| |SW1+-------+R1+----------+R5+----------+R6+---------+R7+-------+SW2|
+---+ ++-+ +--+ +--+ +-++ +---+ +---+ ++-+ +--+ +--+ +-++ +---+
| | | |
---+ ++-+ +--+ +--+ +-++ +---+
| | | |
| +--+ +--+ | | +--+ +--+ |
+------------+R2+----------+R4+-----------+ +------------+R2+----------+R4+-----------+
+--+ +--+ +--+ +--+
</artwork>
Figure 3: CCDR architecture for Multi-BGP sessions deployment]]></artwork> </figure>
</figure></t>
</section> </section>
<section>
<section title="PCEP Extension for Critical Parameters Delivery"> <name>PCEP Extension for Critical Parameters Delivery</name>
<t>The PCEP protocol needs to be extended to transfer the following <t>PCEP needs to be extended to transfer the following
critical parameters:</t> critical parameters:</t>
<ul>
<t><list style="symbols"> <li>Peer information that is used to build the BGP session.</li>
<t>Peer information that is used to build the BGP session</t> <li>Explicit route information for BGP next hop of advertised
prefixes.</li>
<t>Explicit route information for BGP next hop of advertised <li>Advertised prefixes and their associated BGP session.</li>
prefixes</t> </ul>
<t>Once the router receives such information, it should establish
<t>Advertised prefixes and their associated BGP session.</t>
</list>Once the router receives such information, it should establish
the BGP session with the peer appointed in the PCEP message, build the the BGP session with the peer appointed in the PCEP message, build the
end-to-end dedicated path hop-by-hop, and advertise the prefixes that E2E dedicated path hop by hop, and advertise the prefixes that
are contained in the corresponding PCEP message.</t> are contained in the corresponding PCEP message.</t>
<t>The dedicated path is preferred by making sure that the explicit <t>The dedicated path is preferred by making sure that the explicit
route created by PCE has the higher priority (lower route preference) route created by PCE has the higher priority (lower route preference)
than the route information created by other dynamic protocols.</t> than the route information created by other dynamic protocols.</t>
<t>All of the above dynamically created states (BGP sessions, explicit rou
<t>All above dynamically created states (BGP sessions, Explicit route tes,
and Prefix advertised prefix) will be cleared on the expiration of the and advertised prefixes) will be cleared on the expiration of the
state timeout interval which is based on the existing Stateful PCE <xref state timeout interval, which is based on the existing stateful PCE <xref
target="RFC8231"/> and PCECC <xref target="RFC8283"/> mechanism.</t> target="RFC8231"/> and PCE as a Central Controller (PCECC) <xref target="RFC8283
"/> mechanism.</t>
<t>Regarding the BGP session, it is not different from that configured <t>Regarding the BGP session, it is not different from that configured
manually or via NETCONF/YANG. Different BGP sessions are used mainly for manually or via Network Configuration Protocol (NETCONF) and YANG. Differe nt BGP sessions are used mainly for
the clarification of the network prefixes, which can be differentiated the clarification of the network prefixes, which can be differentiated
via the different BGP nexthop. Based on this strategy, if we manipulate via the different BGP next hop. Based on this strategy, if we manipulate
the path to the BGP nexthop, then the path to the prefixes that were the path to the BGP next hop, then the path to the prefixes that were
advertised with the BGP sessions will be changed accordingly. Details of advertised with the BGP sessions will be changed accordingly. Details of
communications between PCEP and BGP subsystems in the router's control communications between PCEP and BGP subsystems in the router's control
plane are out of scope of this draft.</t> plane are out of scope of this document.</t>
</section> </section>
<section>
<section title="Deployment Consideration"> <name>Deployment Considerations</name>
<section title="Scalability"> <section>
<name>Scalability</name>
<t>In the CCDR architecture, only the edge routers that connect with <t>In the CCDR architecture, only the edge routers that connect with
the PCE are responsible for the prefixes advertisement via the the PCE are responsible for the prefix advertisement via the
multiple BGP sessions deployment. The route information for these multiple BGP sessions deployment. The route information for these
prefixes within the on-path routers is distributed via the BGP prefixes within the on-path routers is distributed via BGP.
protocol.</t> </t>
<t>For multiple domain deployment, the PCE, or the pool of PCEs <t>For multiple domain deployment, the PCE, or the pool of PCEs
responsible for these domains, needs only to control the edge router responsible for these domains, needs only to control the edge router
to build the multiple EBGP sessions; all other procedures are the same to build the multiple External BGP (EBGP) sessions; all other procedures are the same
as within one domain.</t> as within one domain.</t>
<t>The on-path router needs only to keep the specific policy routes <t>The on-path router needs only to keep the specific policy routes
for the BGP next-hop of the differentiated prefixes, not the specific for the BGP next hop of the differentiated prefixes, not the specific
routes to the prefixes themselves. This lessens the burden of the routes to the prefixes themselves. This lessens the burden of the
table size of policy based routes for the on-path routers; and has table size of policy-based routes for the on-path routers; and has
more expandability compared with BGP flowspec or Openflow solutions. more expandability compared with BGP Flowspec or OpenFlow solutions.
For example, if we want to differentiate 1000 prefixes from the normal For example, if we want to differentiate 1,000 prefixes from the normal
traffic, CCDR needs only one explicit peer route in every on-path traffic, CCDR needs only one explicit peer route in every on-path
router, whereas the BGP flowspec or Openflow solutions need 1000 router, whereas the BGP Flowspec or OpenFlow solutions need 1,000
policy routes on them.</t> policy routes on them.</t>
</section> </section>
<section>
<section title="High Availability"> <name>High Availability</name>
<t>The CCDR architecture is based on the use of the native IP <t>The CCDR architecture is based on the use of native IP.
protocol. If the PCE fails, the forwarding plane will not be impacted, If the PCE fails, the forwarding plane will not be impacted,
as the BGP sessions between all the devices will not flap and the as the BGP sessions between all the devices will not flap, and the
forwarding table remains unchanged.</t> forwarding table remains unchanged.</t>
<t>If one node on the optimal path fails, the priority traffic will <t>If one node on the optimal path fails, the priority traffic will
fall over to the best-effort forwarding path. One can even design fall over to the best-effort forwarding path. One can even design
several paths to load balance/hot-standby the priority traffic to meet several paths to load balance or to create a hot standby
a path failure situation.</t> of the priority traffic to meet a path failure situation.</t>
<t>For ensuring high availability of a PCE/SDN-controllers <t>For ensuring high availability of a PCE/SDN-controllers
architecture, an operator should rely on existing high availability architecture, an operator should rely on existing high availability
solutions for SDN controllers, such as clustering technology and solutions for SDN controllers, such as clustering technology and
deployment.</t> deployment.</t>
</section> </section>
<section>
<section title="Incremental deployment"> <name>Incremental Deployment</name>
<t>Not every router within the network needs to support the necessary <t>Not every router within the network needs to support the necessary
PCEP extension. For such situations, routers on the edge of a domain PCEP extension. For such situations, routers on the edge of a domain
can be upgraded first, and then the traffic can be prioritized between can be upgraded first, and then the traffic can be prioritized between
different domains. Within each domain, the traffic will be forwarded different domains. Within each domain, the traffic will be forwarded
along the best-effort path. A service provider can selectively upgrade along the best-effort path. A service provider can selectively upgrade
the routers on each domain in sequence.</t> the routers on each domain in sequence.</t>
</section> </section>
<section>
<section title="Loop Avoidance"> <name>Loop Avoidance</name>
<t>A PCE needs to assure calculation of the E2E path based on the <t>A PCE needs to assure calculation of the E2E path based on the
status of network and the service requirements in real-time.</t> status of network and the service requirements in real-time.</t>
<t>The PCE needs to consider the explicit route deployment order (for <t>The PCE needs to consider the explicit route deployment order (for
example, from tail router to head router) to eliminate any possible example, from tail router to head router) to eliminate any possible
transient traffic loop.</t> transient traffic loop.</t>
</section> </section>
<section>
<section title="E2E Path Performance Monitoring"> <name>E2E Path Performance Monitoring</name>
<t>It is necessary to deploy the corresponding E2E path performance <t>It is necessary to deploy the corresponding E2E path performance
monitoring mechanism to keep assure that the delay, jitter or packet monitoring mechanism to assure that the delay, jitter, or packet
loss index meet the original path performance aim. The performance loss index meets the original path performance aim. The performance
monitoring results should feedback to the PCE to let it accomplish the monitoring results should provide feedback to the PCE in order for it to
re-optimize process, send the update control message to related PCC if accomplish the
necessary. Traditional OAM methods(ping, trace) can be used.</t> re-optimization process and send the update control message to the relat
ed PCC if
necessary. Traditional OAM methods (ping, trace) can be used.</t>
</section> </section>
</section> </section>
<section>
<section title="Security Considerations"> <name>Security Considerations</name>
<t>The setup of BGP sessions, prefix advertisement, and explicit peer <t>The setup of BGP sessions, prefix advertisement, and explicit peer
route establishment are all controlled by the PCE. See <xref route establishment are all controlled by the PCE. See <xref target="RFC42
target="RFC4271"/> and <xref target="RFC4272"/> for BGP security 71"/> and <xref target="RFC4272"/> for BGP security
considerations. Security consideration part in <xref target="RFC5440"/> considerations. The Security Considerations found in <xref target="RFC5440
and <xref target="RFC8231"/> should be considered. To prevent a bogus " section="10"/>
and <xref target="RFC8231" section="10"/> should be considered. To prevent
a bogus
PCE sending harmful messages to the network nodes, the network devices PCE sending harmful messages to the network nodes, the network devices
should authenticate the validity of the PCE and ensure a secure should authenticate the validity of the PCE and ensure a secure
communication channel between them. Mechanisms described in <xref communication channel between them. Mechanisms described in <xref target="
target="RFC8253"/> should be used.</t> RFC8253"/> should be used.</t>
<t>The CCDR architecture does not require changes to the forwarding <t>The CCDR architecture does not require changes to the forwarding
behavior of the underlay devices. There are no additional security behavior of the underlay devices. There are no additional security
impacts on these devices.</t> impacts on these devices.</t>
</section> </section>
<section>
<section title="IANA Considerations"> <name>IANA Considerations</name>
<t>This document does not require any IANA actions.</t> <t>This document has no IANA actions.</t>
</section>
<section title="Acknowledgement">
<t>The author would like to thank Deborah Brungard, Adrian Farrel,
Vishnu Beeram, Lou Berger, Dhruv Dhody, Raghavendra Mallya , Mike
Koldychev, Haomian Zheng, Penghui Mi, Shaofu Peng, Donald Eastlake,
Alvaro Retana, Martin Duke, Magnus Westerlund, Benjamin Kaduk, Roman
Danyliw, Eric Vyncke, Murray Kucherawy, Erik Kline and Jessica Chen for
their supports and comments on this draft.</t>
</section> </section>
</middle> </middle>
<back> <back>
<references title="Normative References"> <references>
<?rfc include="reference.RFC.4271"?> <name>References</name>
<references>
<?rfc include="reference.RFC.4272"?> <name>Normative References</name>
<xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.R
<?rfc include="reference.RFC.4456"?> FC.4271.xml"/>
<xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.R
<?rfc include="reference.RFC.5440"?> FC.4272.xml"/>
<xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.R
<?rfc include="reference.RFC.7752"?> FC.4456.xml"/>
<xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.R
<?rfc include="reference.RFC.8231"?> FC.5440.xml"/>
<xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.R
<?rfc include="reference.RFC.8253"?> FC.7752.xml"/>
<xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.R
<?rfc include="reference.RFC.8283"?> FC.8231.xml"/>
</references> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.R
FC.8253.xml"/>
<references title="Informative References"> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.R
<?rfc include="reference.RFC.4655"?> FC.8283.xml"/>
</references>
<?rfc include="reference.RFC.8735"?> <references>
<name>Informative References</name>
<xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.R
FC.4655.xml"/>
<xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.R
FC.8735.xml"/>
</references>
</references> </references>
<section numbered="false">
<name>Acknowledgments</name>
<t>The author would like to thank <contact fullname="Deborah Brungard"/>,
<contact fullname="Adrian Farrel"/>,
<contact fullname="Vishnu Beeram"/>, <contact fullname="Lou Berger"/>, <co
ntact fullname="Dhruv Dhody"/>, <contact fullname="Raghavendra Mallya"/>, <conta
ct fullname="Mike Koldychev"/>, <contact fullname="Haomian Zheng"/>, <cont
act fullname="Penghui Mi"/>, <contact fullname="Shaofu Peng"/>, <contact fullnam
e="Donald Eastlake"/>,
<contact fullname="Alvaro Retana"/>, <contact fullname="Martin Duke"/>, <c
ontact fullname="Magnus Westerlund"/>, <contact fullname="Benjamin Kaduk"/>, <co
ntact fullname="Roman Danyliw"/>, <contact fullname="Éric Vyncke"/>, <cont
act fullname="Murray Kucherawy"/>, <contact fullname="Erik Kline"/>, and <contac
t fullname="Jessica Chen"/> for
their supports and comments on this document.</t>
</section>
</back> </back>
</rfc> </rfc>
 End of changes. 110 change blocks. 
385 lines changed or deleted 329 lines changed or added

This html diff was produced by rfcdiff 1.48. The latest version is available from http://tools.ietf.org/tools/rfcdiff/