Network Working Group                                              T. Li
Request for Comments: 5302                        Redback Networks, Inc.
Obsoletes: 2966                                                  H. Smit
Updates: 1195
Category: Standards Track                                  T. Przygienda
                                                                Z2 Sagl
                                                           October 2008


         Domain-Wide Prefix Distribution with Two-Level IS-IS

Status of This Memo

  This document specifies an Internet standards track protocol for the
  Internet community, and requests discussion and suggestions for
  improvements.  Please refer to the current edition of the "Internet
  Official Protocol Standards" (STD 1) for the standardization state
  and status of this protocol.  Distribution of this memo is unlimited.

Abstract

  This document describes extensions to the Intermediate System to
  Intermediate System (IS-IS) protocol to support optimal routing
  within a two-level domain.  The IS-IS protocol is specified in ISO
  10589, with extensions for supporting IPv4 (Internet Protocol)
  specified in RFC 1195.  This document replaces RFC 2966.

  This document extends the semantics presented in RFC 1195 so that a
  routing domain running with both level 1 and level 2 Intermediate
  Systems (IS) (routers) can distribute IP prefixes between level 1 and
  level 2, and vice versa.  This distribution requires certain
  restrictions to ensure that persistent forwarding loops do not form.
  The goal of this domain-wide prefix distribution is to increase the
  granularity of the routing information within the domain.

















Li, et al.                  Standards Track                     [Page 1]

RFC 5302            Domain-wide Prefix Distribution         October 2008


Table of Contents

  1.  Introduction . . . . . . . . . . . . . . . . . . . . . . . . .  3
    1.1.  Motivations for Domain-Wide Prefix Distribution  . . . . .  3
    1.2.  Scalability  . . . . . . . . . . . . . . . . . . . . . . .  5
    1.3.  Requirements Language  . . . . . . . . . . . . . . . . . .  6
  2.  Proposed Syntax and Semantics for L2->L1 Inter-Area Routes . .  6
    2.1.  Clarification of External Route-Type and External
          Metric-Type  . . . . . . . . . . . . . . . . . . . . . . .  7
    2.2.  Definition of External IP Prefixes in Level 1 LSPs . . . .  8
  3.  Types of IP Routes in IS-IS and Their Order of Preference  . .  8
    3.1.  Overview of All Types of IP Prefixes in IS-IS Link
          State PDUs . . . . . . . . . . . . . . . . . . . . . . . .  9
    3.2.  Order of Preference for all Types of IP Routes in IS-IS  . 11
    3.3.  Additional Notes on What Prefixes to Accept or
          Advertise  . . . . . . . . . . . . . . . . . . . . . . . . 12
  4.  Inter-Operability with Older Implementations . . . . . . . . . 12
  5.  Comparisons with Other Proposals . . . . . . . . . . . . . . . 13
  6.  Security Considerations  . . . . . . . . . . . . . . . . . . . 13
  7.  References . . . . . . . . . . . . . . . . . . . . . . . . . . 14
    7.1.  Normative References . . . . . . . . . . . . . . . . . . . 14
    7.2.  Informative References . . . . . . . . . . . . . . . . . . 14





























Li, et al.                  Standards Track                     [Page 2]

RFC 5302            Domain-wide Prefix Distribution         October 2008


1.  Introduction

  This document describes extensions to the Intermediate System to
  Intermediate System (IS-IS) protocol to support optimal routing
  within a two-level domain.  The IS-IS protocol is specified in
  [ISO-10589], with extensions for supporting IPv4 (Internet Protocol)
  specified in [RFC1195].

  This document replaces [RFC2966], which was an Informational
  document.  This document is on the standards track.  No other
  intentional substantive changes have been made.

  This document extends the semantics presented in RFC 1195 so that a
  routing domain running with both level 1 and level 2 Intermediate
  Systems (IS) (routers) can distribute IP prefixes between level 1 and
  level 2, and vice versa.  This distribution requires certain
  restrictions to ensure that persistent forwarding loops do not form.
  The goal of this domain-wide prefix distribution is to increase the
  granularity of the routing information within the domain.

  An IS-IS routing domain (a.k.a. an autonomous system running IS-IS)
  can be partitioned into multiple level 1 (L1) areas, and a level 2
  (L2) connected subset of the topology that interconnects all of the
  L1 areas.  Within each L1 area, all routers exchange link state
  information.  L2 routers also exchange L2 link state information to
  compute routes between areas.

  RFC 1195 defines the Type, Length, and Value (TLV) tuples that are
  used to transport IPv4 routing information in IS-IS.  RFC 1195 also
  specifies the semantics and procedures for interactions between
  levels.  Specifically, routers in an L1 area will exchange
  information within the L1 area.  For IP destinations not found in the
  prefixes in the L1 database, the L1 router should forward packets to
  the nearest router that is in both L1 and L2 (i.e., an L1L2 router)
  with the "attached bit" set in its L1 Link State Protocol Data Unit
  (LSP).

  Also per RFC 1195, an L1L2 router should be manually configured with
  a set of prefixes that summarizes the IP prefixes reachable in that
  L1 area.  These summaries are injected into L2.  RFC 1195 specifies
  no further interactions between L1 and L2 for IPv4 prefixes.

1.1.  Motivations for Domain-Wide Prefix Distribution

  The mechanisms specified in RFC 1195 are appropriate in many
  situations and lead to excellent scalability properties.  However, in
  certain circumstances, the domain administrator may wish to sacrifice
  some amount of scalability and distribute more specific information



Li, et al.                  Standards Track                     [Page 3]

RFC 5302            Domain-wide Prefix Distribution         October 2008


  than is described by RFC 1195.  This section discusses the various
  reasons why the domain administrator may wish to make such a
  tradeoff.

  One major reason for distributing more prefix information is to
  improve the quality of the resulting routes.  A well-known property
  of prefix summarization or any abstraction mechanism is that it
  necessarily results in a loss of information.  This loss of
  information in turn results in the computation of a route based upon
  less information, which will frequently result in routes that are not
  optimal.

  A simple example can serve to demonstrate this adequately.  Suppose
  that an L1 area has two L1L2 routers that both advertise a single
  summary of all prefixes within the L1 area.  To reach a destination
  inside the L1 area, any other L2 router is going to compute the
  shortest path to one of the two L1L2 routers for that area.  Suppose,
  for example, that both of the L1L2 routers are equidistant from the
  L2 source and that the L2 source arbitrarily selects one L1L2 router.
  This router may not be the optimal router when viewed from the L1
  topology.  In fact, it may be the case that the path from the
  selected L1L2 router to the destination router may traverse the L1L2
  router that was not selected.  If more detailed topological
  information or more detailed metric information was available to the
  L2 source router, it could make a more optimal route computation.

  This situation is symmetric in that an L1 router has no information
  about prefixes in L2 or within a different L1 area.  In using the
  nearest L1L2 router, that L1L2 is effectively injecting a default
  route without metric information into the L1 area.  The route
  computation that the L1 router performs is similarly suboptimal.

  Besides the optimality of the routes computed, there are two other
  significant drivers for the domain-wide distribution of prefix
  information.

  When a router learns multiple possible paths to external destinations
  via BGP, it will select only one of those routes to be installed in
  the forwarding table.  One of the factors in the BGP route selection
  is the IGP cost to the BGP next hop address.  Many ISP networks
  depend on this technique, which is known as "shortest exit routing".
  If a L1 router does not know the exact IGP metric to all BGP speakers
  in other L1 areas, it cannot do effective shortest exit routing.

  The third driver is the current practice of using the IGP (IS-IS)
  metric as part of the BGP Multi-Exit Discriminator (MED).  The value
  in the MED is advertised to other domains and is used to inform other
  domains of the optimal entry point into the current domain.  Current



Li, et al.                  Standards Track                     [Page 4]

RFC 5302            Domain-wide Prefix Distribution         October 2008


  practice is to take the IS-IS metric and insert it as the MED value.
  This tends to cause external traffic to enter the domain at the point
  closest to the exit router.  Note that the receiving domain MAY,
  based upon policy, choose to ignore the MED that is advertised.
  However, current practice is to distribute the IGP metric in this way
  in order to optimize routing wherever possible.  This is possible in
  current networks that only are a single area, but becomes problematic
  if hierarchy is to be installed into the network.  This is again
  because the loss of end-to-end metric information means that the MED
  value will not reflect the true distance across the advertising
  domain.  Full distribution of prefix information within the domain
  would alleviate this problem, as it would allow accurate computation
  of the IS-IS metric across the domain, resulting in an accurate value
  presented in the MED.

1.2.  Scalability

  The disadvantage to performing the domain-wide prefix distribution
  described above is that it has an impact on the scalability of IS-IS.
  Areas within IS-IS help scalability in that LSPs are contained within
  a single area.  This limits the size of the link state database,
  which in turn limits the complexity of the shortest path computation.

  Further, the summarization of the prefix information aids scalability
  in that the abstraction of the prefix information removes the sheer
  number of data items to be transported and the number of routes to be
  computed.

  It should be noted quite strongly that the distribution of prefixes
  on a domain-wide basis impacts the scalability of IS-IS in the second
  respect.  It will increase the number of prefixes throughout the
  domain.  This will result in increased memory consumption,
  transmission requirements, and computation requirements throughout
  the domain.

  It must also be noted that the domain-wide distribution of prefixes
  has no effect whatsoever on the first aspect of scalability, namely
  the existence of areas and the limitation of the distribution of the
  link state database.

  Thus, the net result is that the introduction of domain-wide prefix
  distribution into a formerly flat, single area network is a clear
  benefit to the scalability of that network.  However, it is a
  compromise and does not provide the maximum scalability available
  with IS-IS.  Domains that choose to make use of this facility should
  be aware of the tradeoff that they are making between scalability and
  optimality, and provision and monitor their networks accordingly.
  Normal provisioning guidelines that would apply to a fully



Li, et al.                  Standards Track                     [Page 5]

RFC 5302            Domain-wide Prefix Distribution         October 2008


  hierarchical deployment of IS-IS will not apply to this type of
  configuration.

1.3.  Requirements Language

  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 [RFC2119].

2.  Proposed Syntax and Semantics for L2->L1 Inter-Area Routes

  This document defines the syntax of how to advertise level 2 routes
  in level 1 LSPs.  The encoding is an extension of the encoding in RFC
  1195.

  To some extent, in IS-IS the level 2 backbone can be seen as a
  separate area itself.  RFC 1195 defines that L1L2 routers can
  advertise IP routes that were learned via L1 routing into L2.  These
  routes can be regarded as inter-area routes.  RFC 1195 defines that
  these L1->L2 inter-area routes must be advertised in L2 LSPs in the
  "IP Internal Reachability Information" TLV (TLV 128).  Intra-area L2
  routes are also advertised in L2 LSPs in an "IP Internal Reachability
  Information" TLV.  Therefore, L1->L2 inter-area routes are
  indistinguishable from L2 intra-area routes.

  RFC 1195 does not define L2->L1 inter-area routes.  A simple
  extension would be to allow an L1L2 router to advertise routes
  learned via L2 routing in its L1 LSP.  However, to prevent routing-
  loops, L1L2 routers MUST NOT advertise L2->L1 inter-area routes that
  they learn via L1 routing back into L2.  Therefore, there must be a
  way to distinguish L2->L1 inter-area routes from L1 intra-area
  routes.  [RFC5305] defines the "up/down bit" for this purpose in the
  extended IP reachability TLV (TLV 135).  RFC 1195 defines TLVs 128
  and 130 to contain IP routes.  TLVs 128 and 130 have a Metric field
  that consists of 4 Type of Service (TOS) metrics.  The first metric,
  the so-called "default metric", has the high-order bit reserved (bit
  8).  Routers must set this bit to zero on transmission, and ignore it
  on receipt.

  This document redefines this high-order bit in the default Metric
  field in TLVs 128 and 130 to be the up/down bit.  L1L2 routers MUST
  set this bit to one for prefixes that are derived from L2 routing and
  are advertised into L1 LSPs.  The bit MUST be set to zero for all
  other IP prefixes in L1 or L2 LSPs.  Prefixes with the up/down bit
  set that are learned via L1 routing MUST NOT be advertised by L1L2
  routers back into L2.





Li, et al.                  Standards Track                     [Page 6]

RFC 5302            Domain-wide Prefix Distribution         October 2008


2.1.  Clarification of External Route-Type and External Metric-Type

  RFC 1195 defines two TLVs for carrying IP prefixes.  TLV 128 is
  defined as "IP Internal Reachability Information", and should be used
  to carry IP prefixes that are directly connected to IS-IS routers.
  TLV 130 is defined as "IP External Reachability Information", and
  should be used to carry routes learned from outside the IS-IS domain.
  RFC 1195 documents TLV type 130 only for level 2 LSPs.

  RFC 1195 also defines two types of metrics.  Metrics of the internal
  metric-type should be used when the metric is comparable to metrics
  used to weigh links inside the IS-IS domain.  Metrics of the external
  metric-type should be used if the metric of an IP prefix cannot be
  directly compared to internal metrics.  The external metric-type can
  only be used for external IP prefixes.  A direct result is that
  metrics of the external metric-type should never be seen in TLV 128.

  To prevent confusion, this document states again that when a router
  computes IP routes, it MUST give the same preference to IP routes
  advertised in an "IP Internal Reachability Information" TLV and IP
  routes advertised in an "IP External Reachability Information" TLV.
  RFC 1195 states this quite clearly in the note in paragraph 3.10.2,
  item 2c).  This document does not alter this rule of preference.

  NOTE:  Internal routes (routes to destinations announced in the "IP
     Internal Reachability Information" field) and external routes
     using internal metrics (routes to destinations announced in the
     "IP External Reachability Information" field, with a metric of
     type "internal") are treated identically for the purpose of the
     order of preference of routes, and the Dijkstra calculation.

  However, IP routes advertised in "IP External Reachability
  Information" with the external metric-type MUST be given less
  preference than the same IP routes advertised with the internal
  metric-type, regardless of the value of the metrics.

  While IS-IS routers MUST NOT give different preference to IP prefixes
  learned via "IP Internal Reachability Information" and "IP External
  Reachability Information" when executing the Dijkstra calculation,
  routers that implement multiple IGPs are free to use this distinction
  between internal and external routes when comparing routes derived
  from different IGPs for inclusion in their global Routing Information
  Base (RIB).








Li, et al.                  Standards Track                     [Page 7]

RFC 5302            Domain-wide Prefix Distribution         October 2008


2.2.  Definition of External IP Prefixes in Level 1 LSPs

  RFC 1195 does not define the "IP External Reachability Information"
  TLV for L1 LSPs.  However, there is no reason why an IS-IS
  implementation could not allow for redistribution of external routes
  into L1.  Some IS-IS implementations already allow network
  administrators to do this.  This document loosens the restrictions in
  RFC 1195 and allows for the inclusion of the "IP External
  Reachability Information" TLV in L1 LSPs.

  RFC 1195 defines that IP routes learned via L1 routing must always be
  advertised in L2 LSPs in an "IP Internal Reachability Information"
  TLV.  Now that this document allows "IP External Reachability
  Information" TLVs in L1 LSPs and allows for the advertisement of
  routes learned via L2 routing into L1, the above rule needs an
  extension.

  When an L1L2 router advertises an L1 route into L2, where that L1
  route was learned via a prefix advertised in an "IP External
  Reachability Information" TLV, that L1L2 router SHOULD advertise that
  prefix in its L2 LSP within an "IP External Reachability Information"
  TLV.  L1 routes learned via an "IP Internal Reachability Information"
  TLV SHOULD still be advertised within an "IP Internal Reachability
  Information" TLV.  These rules should also be applied when
  advertising IP routes derived from L2 routing into L1.  Of course in
  this case, the up/down bit MUST be set also.

  RFC 1195 defines that if a router sees the same external prefix
  advertised by two or more routers with the same external metric, it
  must select the route that is advertised by the router that is
  closest to itself.  It should be noted that now that external routes
  can be advertised from L1 into L2, and vice versa, the router that
  advertises an external prefix in its LSP might not be the router that
  originally injected this prefix into the IS-IS domain.  Therefore, it
  is less useful to advertise external routes with external metrics
  into other levels.

3.  Types of IP Routes in IS-IS and Their Order of Preference

  RFC 1195 and this document define several ways of advertising IP
  routes in IS-IS.  There are four variables involved.

  1.  The level of the LSP in which the route is advertised.  There are
      currently two possible values: level 1 and level 2.

  2.  The route-type, which can be derived from the type of TLV in
      which the prefix is advertised.  Internal routes are advertised
      in IP Internal Reachability Information TLVs (TLV 128), and



Li, et al.                  Standards Track                     [Page 8]

RFC 5302            Domain-wide Prefix Distribution         October 2008


      external routes are advertised in IP External Reachability
      Information TLVs (TLV 130).

  3.  The metric-type: internal or external.  The metric-type is
      derived from the internal/external metric-type bit in the Metric
      field (bit 7).

  4.  The fact whether this route is leaked down in the hierarchy, and
      thus can not be advertised back up.  This information can be
      derived from the newly defined up/down bit in the default Metric
      field.

3.1.  Overview of All Types of IP Prefixes in IS-IS Link State PDUs

  The combination IP Internal Reachability Information and external
  metric-type is not allowed.  Also, the up/down bit MUST NOT be set in
  L2 LSPs.  This leaves us with 8 different types of IP advertisements
  in IS-IS.  However, there are more than 8 reasons for IP prefixes to
  be advertised in IS-IS.  The following list describes the types of IP
  prefixes and how they are encoded.

  L1 intra-area routes:  These are advertised in L1 LSPs, in TLV 128.
     The up/down bit is set to zero, metric-type is internal metric.
     These IP prefixes are directly connected to the advertising
     router.

  L1 external routes:  These are advertised in L1 LSPs, in TLV 130.
     The up/down bit is set to zero, metric-type is internal metric.
     These IP prefixes are learned from other IGPs, and are usually not
     directly connected to the advertising router.

  L2 intra-area routes:  These are advertised in L2 LSPs, in TLV 128.
     The up/down bit is set to zero, metric-type is internal metric.
     These IP prefixes are directly connected to the advertising
     router.  These prefixes cannot be distinguished from L1->L2 inter-
     area routes.

  L2 external routes:  These are advertised in L2 LSPs, in TLV 130.
     The up/down bit is set to zero, metric-type is internal metric.
     These IP prefixes are learned from other IGPs, and are usually not
     directly connected to the advertising router.  These prefixes
     cannot be distinguished from L1->L2 inter-area external routes.

  L1->L2 inter-area routes:  These are advertised in L2 LSPs, in TLV
     128.  The up/down bit is set to zero, metric-type is internal
     metric.  These IP prefixes are learned via L1 routing, and were
     derived during the L1 Shortest Path First (SPF) computation from




Li, et al.                  Standards Track                     [Page 9]

RFC 5302            Domain-wide Prefix Distribution         October 2008


     prefixes advertised in L1 LSPs in TLV 128.  These prefixes cannot
     be distinguished from L2 intra-area routes.

  L1->L2 inter-area external routes:  These are advertised in L2 LSPs,
     in TLV 130.  The up/down bit is set to zero, metric-type is
     internal metric.  These IP prefixes are learned via L1 routing,
     and were derived during the L1 SPF computation from prefixes
     advertised in L1 LSPs in TLV 130.  These prefixes cannot be
     distinguished from L2 external routes.

  L2->L1 inter-area routes:  These are advertised in L1 LSPs, in TLV
     128.  The up/down bit is set to one, metric-type is internal
     metric.  These IP prefixes are learned via L2 routing, and were
     derived during the L2 SPF computation from prefixes advertised in
     TLV 128.

  L2->L1 inter-area external routes:  These are advertised in L1 LSPs,
     in TLV 130.  The up/down bit is set to one, metric-type is
     internal metric.  These IP prefixes are learned via L2 routing,
     and were derived during the L2 SPF computation from prefixes
     advertised in L2 LSPs in TLV 130.

  L1 external routes with external metric:  These are advertised in L1
     LSPs, in TLV 130.  The up/down bit is set to zero, metric-type is
     external metric.  These IP prefixes are learned from other IGPs,
     and are usually not directly connected to the advertising router.

  L2 external routes with external metric:  These are advertised in L2
     LSPs, in TLV 130.  The up/down bit is set to zero, metric-type is
     external metric.  These IP prefixes are learned from other IGPs,
     and are usually not directly connected to the advertising router.
     These prefixes cannot be distinguished from L1->L2 inter-area
     external routes with external metric.

  L1->L2 inter-area external routes with external metric:  These are
     advertised in L2 LSPs, in TLV 130.  The up/down bit is set to
     zero, metric-type is external metric.  These IP prefixes are
     learned via L1 routing, and were derived during the L1 SPF
     computation from prefixes advertised in L1 LSPs in TLV 130 with
     external metrics.  These prefixes can not be distinguished from L2
     external routes with external metric.

  L2->L1 inter-area external routes with external metric:  These are
     advertised in L1 LSPs, in TLV 130.  The up/down bit is set to one,
     metric-type is external metric.  These IP prefixes are learned via
     L2 routing, and were derived during the L1 SPF computation from
     prefixes advertised in L2 LSPs in TLV 130 with external metrics.




Li, et al.                  Standards Track                    [Page 10]

RFC 5302            Domain-wide Prefix Distribution         October 2008


3.2.  Order of Preference for all Types of IP Routes in IS-IS

  Unfortunately, IS-IS cannot depend on metrics alone for route
  selection.  Some types of routes must always be preferred over
  others, regardless of the costs that were computed in the Dijkstra
  calculation.  One of the reasons for this is that inter-area routes
  can only be advertised with a maximum metric of 63.  Another reason
  is that this maximum value of 63 does not mean infinity (e.g., like a
  hop count of 16 in RIP denotes unreachable).  Introducing a value for
  infinity cost in IS-IS inter-area routes would introduce counting-
  to-infinity behavior via two or more L1L2 routers, which would have a
  bad impact on network stability.

  The order of preference of IP routes in IS-IS is based on a few
  assumptions.

  o  RFC 1195 defines that routes derived from L1 routing are preferred
     over routes derived from L2 routing.

  o  The note in RFC 1195, paragraph 3.10.2, item 2c) defines that
     internal routes with internal metric-type and external prefixes
     with internal metric-type have the same preference.

  o  RFC 1195 defines that external routes with internal metric-type
     are preferred over external routes with external metric-type.

  o  Routes derived from L2 routing are preferred over L2->L1 routes
     derived from L1 routing.

  Based on these assumptions, this document defines the following route
  preferences.

  1.  L1 intra-area routes with internal metric; L1 external routes
      with internal metric

  2.  L2 intra-area routes with internal metric; L2 external routes
      with internal metric; L1->L2 inter-area routes with internal
      metric; L1->L2 inter-area external routes with internal metric

  3.  L2->L1 inter-area routes with internal metric; L2->L1 inter-area
      external routes with internal metric

  4.  L1 external routes with external metric

  5.  L2 external routes with external metric; L1->L2 inter-area
      external routes with external metric





Li, et al.                  Standards Track                    [Page 11]

RFC 5302            Domain-wide Prefix Distribution         October 2008


  6.  L2->L1 inter-area external routes with external metric

3.3.  Additional Notes on What Prefixes to Accept or Advertise

  Section 3.1 enumerates all used IP route-types in IS-IS.  Besides
  these defined route-types, the encoding used would allow for a few
  more potential combinations.  One of them is the combination of "IP
  Internal Reachability Information" and external metric-type.  This
  combination SHOULD NOT be used when building an LSP.  Upon receipt of
  an IP prefix with this combination, routers MUST ignore this prefix.
  Another issue would be the usage of the up/down bit in L2 LSPs.
  Because IS-IS is currently defined with two levels of hierarchy,
  there should never be a need to set the up/down bit in L2 LSPs.
  However, if IS-IS would ever be extended with more than two levels of
  hierarchy, L2-only (or L1L2) routers will need to be able to accept
  L2 IP routes with the up/down bit set.  Therefore, it is RECOMMENDED
  that implementations ignore the up/down bit in L2 LSPs, and accept
  the prefixes in L2 LSPs regardless of whether the up/down bit is set.
  This will allow for simpler migration once more than two levels of
  hierarchy are defined.

  Another detail that implementors should be aware of is the fact that
  L1L2 routers SHOULD only advertise in their L2 LSP those L1 routes
  that they use for forwarding themselves.  They SHOULD NOT
  unconditionally advertise into L2 all prefixes from LSPs in the L1
  database.

  Not all prefixes need to be advertised up or down the hierarchy.
  Implementations might allow for additional manual filtering or
  summarization to further bring down the number of inter-area prefixes
  they advertise in their LSPs.  It is also RECOMMENDED that the
  default configuration of L1L2 routers not advertise any L2 routes
  into L1 (see also Section 4).

4.  Inter-Operability with Older Implementations

  The solution in this document is not fully compatible with RFC 1195.
  It is an extension to RFC 1195.  If routers do not use the new
  functionality of external L1 routes or L2->L1 inter-area routes,
  older implementations that strictly follow RFC 1195 will be
  compatible with newer implementations that follow this document.

  Implementations that do not accept the "IP External Reachability
  Information" TLV in L1 LSPs will not be able to compute external L1
  routes.  This could cause routing loops between L1-only routers that
  do understand external L1 routes for a particular destination, and
  L1-only routers that use the default route pointing to the closest
  attached L1L2 router for that destination.



Li, et al.                  Standards Track                    [Page 12]

RFC 5302            Domain-wide Prefix Distribution         October 2008


  Implementations that follow RFC 1195 SHOULD ignore bit 8 in the
  default Metric field when computing routes.  Therefore, even older
  implementations that do not know of the up/down bit should be able to
  accept the new L2->L1 inter-area routes.  These older implementations
  will install the new L2->L1 inter-area routes as L1 intra-area
  routes, but that in itself does not cause routing loops among L1-only
  routers.

  However, it is vital that the up/down bit is recognized by L1L2
  routers.  As has been stated before, L1L2 routers MUST NOT advertise
  L2->L1 inter-area routes back into L2.  Therefore, if L2 routes are
  advertised down into an L1 area, it is required that all L1L2 routers
  in that area run software that understands the new up/down bit.
  Older implementations that follow RFC 1195 and do not understand the
  new up/down bit will treat the L2->L1 inter-area routes as L1 intra-
  area routes, and they will advertise these routes back into L2.  This
  can cause routing loops, sub-optimal routing, or extra routing
  instability.  For this reason, it is RECOMMENDED that implementations
  by default not advertise any L2 routes into L1.  Implementations
  SHOULD force the network administrator to manually configure L1L2
  routers to advertise any L2 routes into L1.

5.  Comparisons with Other Proposals

  In [RFC5305], a new TLV is defined to transport IP prefix
  information.  This TLV format also defines an up/down bit to allow
  for L2->L1 inter-area routes.  RFC 5305 also defines a new TLV to
  describe links.  Both TLVs have wider metric space and have the
  possibility to define sub-TLVs to advertise extra information
  belonging to the link or prefix.  The wider metric space in IP prefix
  TLVs allows for more granular metric information about inter-area
  path costs.  To make full use of the wider metric space, network
  administrators must deploy both new TLVs at the same time.

  Deployment of RFC 5305 requires an upgrade of all routers in the
  network and a transition to the new TLVs.  Such a network-wide
  upgrade and transition might not be an easy task.  In this case, the
  solution defined in this document, which requires only an upgrade of
  L1L2 routers in selected areas, might be a good alternative to the
  solution defined in 5305.

6.  Security Considerations

  This document raises no new security issues for IS-IS; for general
  security considerations for IS-IS see [RFC5304].






Li, et al.                  Standards Track                    [Page 13]

RFC 5302            Domain-wide Prefix Distribution         October 2008


7.  References

7.1.  Normative References

  [ISO-10589]  ISO, "Intermediate System to Intermediate System
               intra-domain routeing information exchange protocol for
               use in conjunction with the protocol for providing the
               connectionless-mode network service (ISO 8473)",
               International Standard 10589:2002, Second Edition, 2002.

  [RFC1195]    Callon, R., "Use of OSI IS-IS for routing in TCP/IP and
               dual environments", RFC 1195, December 1990.

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

7.2.  Informative References

  [RFC2966]    Li, T., Przygienda, T., and H. Smit, "Domain-wide Prefix
               Distribution with Two-Level IS-IS", RFC 2966,
               October 2000.

  [RFC5304]    Li, T. and R. Atkinson, "IS-IS Cryptographic
               Authentication", RFC 5304, October 2008.

  [RFC5305]    Li, T. and H. Smit, "IS-IS Extensions for Traffic
               Engineering", RFC 5305, October 2008.
























Li, et al.                  Standards Track                    [Page 14]

RFC 5302            Domain-wide Prefix Distribution         October 2008


Authors' Addresses

  Tony Li
  Redback Networks, Inc.
  300 Holger Way
  San Jose, CA  95134
  USA

  Phone: +1 408 750 5160
  EMail: [email protected]


  Henk Smit

  EMail: [email protected]


  Tony Przygienda
  Z2 Sagl
  Via Tersaggio 20
  CH-6949 Comano
  Switzerland

  EMail: [email protected]



























Li, et al.                  Standards Track                    [Page 15]

RFC 5302            Domain-wide Prefix Distribution         October 2008


Full Copyright Statement

  Copyright (C) The IETF Trust (2008).

  This document is subject to the rights, licenses and restrictions
  contained in BCP 78, and except as set forth therein, the authors
  retain all their rights.

  This document and the information contained herein are provided on an
  "AS IS" basis and THE CONTRIBUTOR, THE ORGANIZATION HE/SHE REPRESENTS
  OR IS SPONSORED BY (IF ANY), THE INTERNET SOCIETY, THE IETF TRUST AND
  THE INTERNET ENGINEERING TASK FORCE DISCLAIM ALL WARRANTIES, EXPRESS
  OR IMPLIED, INCLUDING BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF
  THE INFORMATION HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED
  WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.

Intellectual Property

  The IETF takes no position regarding the validity or scope of any
  Intellectual Property Rights or other rights that might be claimed to
  pertain to the implementation or use of the technology described in
  this document or the extent to which any license under such rights
  might or might not be available; nor does it represent that it has
  made any independent effort to identify any such rights.  Information
  on the procedures with respect to rights in RFC documents can be
  found in BCP 78 and BCP 79.

  Copies of IPR disclosures made to the IETF Secretariat and any
  assurances of licenses to be made available, or the result of an
  attempt made to obtain a general license or permission for the use of
  such proprietary rights by implementers or users of this
  specification can be obtained from the IETF on-line IPR repository at
  http://www.ietf.org/ipr.

  The IETF invites any interested party to bring to its attention any
  copyrights, patents or patent applications, or other proprietary
  rights that may cover technology that may be required to implement
  this standard.  Please address the information to the IETF at
  [email protected].












Li, et al.                  Standards Track                    [Page 16]