Internet Engineering Task Force (IETF)                          M. Zhang
Request for Comments: 8249                                      X. Zhang
Updates: 6325, 7177, 7780                                D. Eastlake 3rd
Category: Standards Track                                         Huawei
ISSN: 2070-1721                                               R. Perlman
                                                               Dell EMC
                                                          S. Chatterjee
                                                                  Cisco
                                                         September 2017


        Transparent Interconnection of Lots of Links (TRILL):
                           MTU Negotiation

Abstract

  The base IETF TRILL (Transparent Interconnection of Lots of Links)
  protocol has a TRILL campus-wide MTU feature, specified in RFCs 6325
  and 7177, that assures that link-state changes can be successfully
  flooded throughout the campus while being able to take advantage of a
  campus-wide capability to support jumbo packets.  This document
  specifies recommended updates to that MTU feature to take advantage,
  for appropriate link-local packets, of link-local MTUs that exceed
  the TRILL campus MTU.  In addition, it specifies an efficient
  algorithm for local MTU testing.  This document updates RFCs 6325,
  7177, and 7780.

Status of This Memo

  This is an Internet Standards Track document.

  This document is a product of the Internet Engineering Task Force
  (IETF).  It represents the consensus of the IETF community.  It has
  received public review and has been approved for publication by the
  Internet Engineering Steering Group (IESG).  Further information on
  Internet Standards is available in Section 2 of RFC 7841.

  Information about the current status of this document, any errata,
  and how to provide feedback on it may be obtained at
  https://www.rfc-editor.org/info/rfc8249.











Zhang, et al.                Standards Track                    [Page 1]

RFC 8249                     MTU Negotiation              September 2017


Copyright Notice

  Copyright (c) 2017 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
  (https://trustee.ietf.org/license-info) 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
     1.1. Conventions Used in This Document ..........................3
  2. Link-Wide TRILL MTU Size ........................................4
     2.1. Operations .................................................5
  3. Testing Link MTU Size ...........................................6
  4. Refreshing Sz ...................................................8
  5. Relationship between Port MTU, Lz, and Sz .......................9
  6. LSP Synchronization ............................................10
  7. Recommendations for Traffic Link Testing of MTU Size ...........10
  8. Backward Compatibility .........................................11
  9. Security Considerations ........................................11
  10. Additions to Configuration ....................................12
     10.1. Per-RBridge Configuration ................................12
     10.2. Per-RBridge Port Configuration ...........................12
  11. IANA Considerations ...........................................12
  12. References ....................................................12
     12.1. Normative References .....................................12
     12.2. Informative References ...................................14
  Acknowledgements ..................................................14
  Authors' Addresses ................................................14














Zhang, et al.                Standards Track                    [Page 2]

RFC 8249                     MTU Negotiation              September 2017


1.  Introduction

  [RFC6325] describes the way RBridges agree on the campus-wide minimum
  acceptable inter-RBridge MTU (Maximum Transmission Unit) size (called
  "Sz") to ensure that link-state flooding operates properly and all
  RBridges converge to the same link state.  For the proper operation
  of TRILL (Transparent Interconnection of Lots of Links) IS-IS, all
  RBridges format their Link State Protocol Data Units (LSPs) to fit
  in Sz.

  [RFC7177] diagrams the state transitions of an adjacency.  If MTU
  testing is enabled, "Link MTU size is successfully tested" is part of
  an event (event A6) causing the transition from the "2-Way" state
  [RFC7177] to the "Report" state for an adjacency.  This means that
  the link MTU testing of size x succeeds, and x is greater than or
  equal to Sz [RFC6325].  If this link cannot support an MTU of Sz, it
  will not be reported as part of the campus topology.

  In this document, a new RECOMMENDED link-wide minimum inter-RBridge
  MTU size, "Lz", is specified.  As further discussed in Section 2, by
  calculating and using Lz as specified herein, link-scoped Protocol
  Data Units (PDUs) can be formatted greater than Sz, up to the
  link-wide minimum acceptable inter-RBridge MTU size, potentially
  improving the efficiency of link utilization and speeding link-state
  convergence.

  An optional TRILL MTU size-testing algorithm is specified in
  Section 3 as an efficient method to update the old MTU testing method
  described in Section 4.3.2 of [RFC6325] and in [RFC7177].  The new
  MTU size-testing method specified in this document is backward
  compatible with the old one.  Multicasting the MTU-probes is
  recommended when there are multiple RBridges on a link responding to
  the probing with an MTU-ack [RFC7177].  The testing method and rules
  of this document are devised in a way that minimizes the number of
  MTU-probes for testing, therefore reducing the number of multicast
  packets for MTU testing.

  This document updates RFCs 6325, 7177, and 7780.  The update to
  [RFC6325] and [RFC7177] is specified in Section 3.  The update to
  [RFC7780] is specified in Section 4.

1.1.  Conventions Used in This Document

  The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
  "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and
  "OPTIONAL" in this document are to be interpreted as described in
  BCP 14 [RFC2119] [RFC8174] when, and only when, they appear in all
  capitals, as shown here.



Zhang, et al.                Standards Track                    [Page 3]

RFC 8249                     MTU Negotiation              September 2017


2.  Link-Wide TRILL MTU Size

  This document specifies a new value "Lz" for the minimum acceptable
  inter-RBridge link MTU size on a local link.  Link-wide Lz is the
  minimum Lz supported and agreed upon amongst all RBridges on a
  specific link.  If the link is usable, Lz will be greater than or
  equal to Sz.

  Some TRILL IS-IS PDUs are exchanged only between neighbors instead of
  throughout the whole campus.  They are confined by the link-wide Lz
  instead of Sz.  Complete Sequence Number PDUs (CSNPs) and Partial
  Sequence Number PDUs (PSNPs) are examples of such PDUs.  These PDUs
  are exchanged only on the local link.  (While TRILL IS-IS Hellos are
  also link local, they are always limited to 1470 bytes for
  robustness.)

  [RFC7356] defines the PDUs that support flooding scopes in addition
  to area-wide scopes and domain-wide scopes.  As specified in
  [RFC8139], RBridges support the Extended L1 Circuit Scope (E-L1CS)
  Flooding Scope LSP (FS-LSP) [RFC7780].  The originatingSNPBufferSize
  for a port is the minimum of the following two quantities but
  not less than 1470 bytes: (1) the MTU of the port and (2) the maximum
  LSP size that the TRILL IS-IS implementation can handle.  They use
  that flooding to exchange their maximum supported value of "Lz".  The
  smallest value of the Lz advertised by the RBridges on a link, but
  not less than Sz, is the link-wide Lz.  An RBridge on a local link
  will be able to tell which other RBridges on that link support E-L1CS
  FS-LSPs because, as required by [RFC7780], all RBridges include the
  Scope Flooding Support TLV [RFC7356] in their TRILL Hellos.

  The maximum size for a level-1 link-local PDU (such as a PSNP or
  CSNP) that may be generated by a system is controlled by the value of
  the management parameter originatingL1SNPBufferSize.  This value
  determines Lz.  The TRILL APPsub-TLV shown in Figure 1 SHOULD be
  included in a TRILL GENINFO TLV [RFC7357] in an E-L1CS FS-LSP
  fragment zero.  If it is missing from an E-L1CS FS-LSP fragment zero
  or there is no E-L1CS FS-LSP fragment zero, it is assumed that its
  originating IS is implicitly advertising its originatingSNPBufferSize
  value as Sz octets.

  E-L1CS FS-LSPs are link local and can also be sent up to a size of Lz
  but, for robustness, E-L1CS FS-LSP fragment zero MUST NOT exceed
  1470 bytes.








Zhang, et al.                Standards Track                    [Page 4]

RFC 8249                     MTU Negotiation              September 2017


             +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
             | Type = 21                     |   (2 bytes)
             +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
             | Length = 2                    |   (2 bytes)
             +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
             | originatingSNPBufferSize      |   (2 bytes)
             +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

            Figure 1: The originatingSNPBufferSize APPsub-TLV

  Type: Set to the originatingSNPBufferSize APPsub-TLV (TRILL
     APPsub-TLV type 21).  Two bytes, because this APPsub-TLV appears
     in an extended TLV [RFC7356].

  Length: Set to 2.

  originatingSNPBufferSize: The local value of
     originatingL1SNPBufferSize as an unsigned integer, limited to the
     range from 1470 to 65,535 bytes.  (A value less than 1470 will be
     ignored.)

2.1.  Operations

  Lz MAY be reported using an originatingSNPBufferSize APPsub-TLV that
  occurs in fragment zero of the RBridge's E-L1CS FS-LSP.  An
  originatingSNPBufferSize APPsub-TLV occurring in any other fragment
  is ignored.  If more than one originatingSNPBufferSize APPsub-TLV
  occurs in fragment zero, the one advertising the smallest value for
  originatingSNPBufferSize, but not less than 1470 bytes, is used.

  Even if all RBridges on a specific link have reached consensus on the
  value of link-wide Lz based on advertised originatingSNPBufferSize,
  it does not mean that these RBridges can safely exchange PDUs between
  each other.  Figure 2 shows such a corner case.  RB1, RB2, and RB3
  are three RBridges on the same link and their Lz is 1800, so the
  link-wide Lz of this link is 1800.  There is an intermediate bridge
  (say B1) between RB2 and RB3 whose port MTU size is 1700.  If RB2
  sends PDUs formatted in chunks of size 1800, those PDUs will be
  discarded by B1.












Zhang, et al.                Standards Track                    [Page 5]

RFC 8249                     MTU Negotiation              September 2017


                        Lz:1800               Lz:1800
                         +---+         |         +---+
                         |RB1|(2000)---|---(2000)|RB2|
                         +---+         |         +---+
                                       |
                 Lz:1800               |
                  +---+               +--+
                  |RB3|(2000)---(1700)|B1|
                  +---+               +--+
                                       |

      Figure 2: Link-Wide Lz = 1800 vs. Tested Link MTU Size = 1700

  Therefore, the link MTU size SHOULD be tested.  After the link MTU
  size of an adjacency is successfully tested, those link-local PDUs,
  such as CSNPs, PSNPs, and E-L1CS FS-LSPs, will be formatted
  no greater than the tested link MTU size and will be safely
  transmitted on this link.

  As for Sz, RBridges continue to propagate their
  originatingL1LSPBufferSize across the campus through the
  advertisement of LSPs as defined in Section 4.3.2 of [RFC6325].  The
  smallest value of Sz advertised by any RBridge, but not less than
  1470, will be deemed as Sz.  Each RBridge formats their "campus-wide"
  PDUs -- for example, LSPs -- no greater than what they determine
  as Sz.

3.  Testing Link MTU Size

  [RFC7177] defines event A6 as indicating that the MTU test was
  successful if MTU testing is enabled.  As described in Section 4.3.2
  of [RFC6325], this is a combination of the following event and
  condition:

  o  Event: The link MTU size has been tested.

  o  Condition: The link can support Sz.

  This condition can be efficiently tested by the following "binary
  search algorithm" and rules.  This updates [RFC6325] and [RFC7177].

  x, lowerBound, and upperBound are local integer variables.  The
  MTU-probe and MTU-ack PDUs are specified in Section 3 of [RFC7176].
  It is RECOMMENDED that one Round-Trip Time (RTT) between the two
  adjacent RBridges be used as the minimum interval between two
  successive probes.  Note that RTT estimation is out of scope for this
  document.  If operators cannot estimate the RTT, the default value of
  5 milliseconds should be assumed.



Zhang, et al.                Standards Track                    [Page 6]

RFC 8249                     MTU Negotiation              September 2017


  Step 0: RB1 sends an MTU-probe padded to the size of link-wide Lz.

     1) If RB1 successfully receives the MTU-ack from RB2 to the probe
        of the value of link-wide Lz within k tries (where k is a
        configurable parameter whose default is 3), the link MTU size
        is set to the size of link-wide Lz.  Stop.

     2) RB1 tries to send an MTU-probe padded to 1470 bytes.

        a) If RB1 fails to receive an MTU-ack from RB2 after k tries
           (an MTU-ack should be considered to have failed two RTTs
           after the probe is sent out), RB1 sets the "failed minimum
           MTU test" flag for RB2 in RB1's Hello.  Stop.

        b) The link MTU size is set to 1470; lowerBound is set to 1470;
           upperBound is set to the link-wide Lz; x is set to
           [(lowerBound + upperBound) / 2], rounded down to the nearest
           integer.

  Step 1: RB1 tries to send an MTU-probe padded to the size x.

     1) If RB1 fails to receive an MTU-ack from RB2 after k tries:

        upperBound is set to x - 1; x is set to
        [(lowerBound + upperBound) / 2], rounded down to the nearest
        integer.

     2) If RB1 receives an MTU-ack to a probe of size x from RB2:

        The link MTU size is set to x; lowerBound is set to x; x is set
        to [(lowerBound + upperBound) / 2], rounded down to the nearest
        integer.  If lowerBound equals upperBound - 1, then x is set to
        upperBound.

     3) If lowerBound >= upperBound or Step 1 has been repeated n times
        (where n is a configurable parameter whose default value is 5),
        stop.

     4) Repeat Step 1.

  After the testing, the two connected RBridges agree on the value of
  the link MTU size.  MTU testing is only done in the Designated VLAN
  [RFC7177].  Since the execution of the above algorithm can be
  resource consuming, it is RECOMMENDED that the Designated RBridge
  (DRB) [RFC7177] take the responsibility to do the testing.  Multicast
  MTU-probes are used instead of unicast when multiple RBridges are





Zhang, et al.                Standards Track                    [Page 7]

RFC 8249                     MTU Negotiation              September 2017


  desired to respond with an MTU-ack on the link.  The binary search
  algorithm given here is a way to minimize the probing attempts; it
  reduces the number of multicast packets for MTU-probing.

  The following rules are designed to determine whether the
  aforementioned "Condition" holds.

  RBridges have figured out the upper bound and lower bound of the link
  MTU size from the execution of the above algorithm.  If Sz is smaller
  than the lower bound or greater than the upper bound, RBridges can
  directly judge whether the link supports Sz without MTU-probing.

  (a) If lowerBound >= Sz, this link can support Sz.

  (b) Else if upperBound <= Sz, this link cannot support Sz.

  Otherwise, RBridges SHOULD test whether the link can support Sz as in
  item (c) below.  If they do not, the only safe assumption will be
  that the link cannot support Sz.  This assumption, without testing,
  might rule out the use of a link that can, in fact, handle packets up
  to Sz.  In the worst case, this might result in unnecessary network
  partition.

  (c) lowerBound < Sz < upperBound.  RBridges probe the link with
      MTU-probe messages padded to Sz.  If an MTU-ack is received
      within k tries, this link can support Sz.  Otherwise, this link
      cannot support Sz.  Through this test, the lower bound and upper
      bound of the link MTU size can be updated accordingly.

4.  Refreshing Sz

  RBridges may join or leave the campus; this may change Sz.

  1) Joining

     a) When a new RBridge joins the campus and its
        originatingL1LSPBufferSize is smaller than the current Sz,
        reporting its originatingL1LSPBufferSize in its LSPs will cause
        other RBridges to decrease their Sz.  Then, any LSP greater
        than the reduced Sz MUST be split, and/or the LSP contents in
        the campus MUST be otherwise redistributed so that no LSP is
        greater than the new Sz.

     b) If the joining RBridge's originatingL1LSPBufferSize is greater
        than or equal to the current Sz, reporting its
        originatingL1LSPBufferSize will not change Sz.





Zhang, et al.                Standards Track                    [Page 8]

RFC 8249                     MTU Negotiation              September 2017


  2) Leaving

     a) From the specification of the Joining process, we know that if
        an RBridge's originatingL1LSPBufferSize is smaller than Sz,
        this RBridge will not join this campus.

     b) When an RBridge leaves the campus and its
        originatingL1LSPBufferSize equals Sz, its LSPs are purged from
        the remainder of the campus after reaching MaxAge [IS-IS].  Sz
        MAY be recalculated and MAY increase.  In other words, while in
        most cases RB1 ignores link-state information for IS-IS
        unreachable RBridge RB2 [RFC7780], originatingL1LSPBufferSize
        is meaningful.  Its value, even from IS-IS unreachable
        RBridges, is used in determining Sz.  This updates [RFC7780].

     c) When an RBridge leaves the campus and its
        originatingL1LSPBufferSize is greater than Sz, Sz will not be
        updated, since Sz is determined by another RBridge with a
        smaller originatingL1LSPBufferSize.

  Frequent LSP "resizing" is harmful to the stability of the TRILL
  campus, so, to avoid this, upward resizing SHOULD be dampened.  When
  an upward resizing event is noticed by an RBridge, it is RECOMMENDED
  that a timer be set at that RBridge via a configurable parameter --
  LSPresizeTime -- whose default value is 300 seconds.  Before this
  timer expires, all subsequent upward resizing will be dampened
  (ignored).  Of course, in a well-configured campus with all RBridges
  configured to have the same originatingL1LSPBufferSize, no resizing
  will be necessary.  It does not matter if different RBridges have
  different dampening timers or if some RBridges resize upward more
  quickly than others.

  If the refreshed Sz is smaller than the lower bound or greater than
  the upper bound of the tested link MTU size, the issue of resource
  consumption from testing the link MTU size can be avoided according
  to rule (a) or (b) as specified in Section 3.  Otherwise, RBridges
  test the link MTU size according to rule (c).

5.  Relationship between Port MTU, Lz, and Sz

  When the port MTU of an RBridge is smaller than the local
  originatingL1SNPBufferSize of an RBridge (an inconsistent
  configuration), that port SHOULD be disabled, since, in any case, an
  adjacency cannot be formed through such a port.  On the other hand,
  when an RBridge receives an LSP or E-L1CS FS-LSP with size greater
  than the link-wide Lz or Sz but not greater than its port MTU size,
  this LSP is processed normally.  If the size of an LSP is greater




Zhang, et al.                Standards Track                    [Page 9]

RFC 8249                     MTU Negotiation              September 2017


  than the MTU size of a port over which it is to be propagated, this
  LSP MUST NOT be sent over the port and an LSPTooLargeToPropagate
  alarm shall be generated [IS-IS].

6.  LSP Synchronization

  An RBridge participates in LSP synchronization on a link as soon as
  it has at least one adjacency on that link that has advanced to at
  least the 2-Way state [RFC7177].  On a LAN link, CSNPs and PSNPs are
  used for synchronization.  On a point-to-point link, only PSNPs are
  used.

  The CSNPs and PSNPs can be formatted in chunks of size (at most)
  link-wide Lz but are processed normally if received having a larger
  size.  Since the link MTU size may not have been tested in the 2-Way
  state, link-wide Lz may be greater than the supported link MTU size.
  In that case, a CSNP or PSNP may be discarded.  After the link MTU
  size is successfully tested, RBridges will begin to format these PDUs
  with a size no greater than that MTU; therefore, these PDUs will
  eventually get through.

  Note that the link MTU size is frequently greater than Sz.
  Link-local PDUs are limited in size by the link MTU size rather than
  Sz, which, when Lz is greater than Sz, promises a reduction in the
  number of PDUs and a faster LSP synchronization process.

7.  Recommendations for Traffic Link Testing of MTU Size

  Sz and link-wide Lz are used to limit the size of most TRILL IS-IS
  PDUs.  They are different from the MTU size restricting the size of
  TRILL Data packets.  The size of a TRILL Data packet is restricted by
  the physical MTU of the ports and links the packet traverses.  It is
  possible that a TRILL Data packet successfully gets through the
  campus but its size is greater than Sz or link-wide Lz values.

  The algorithm defined for testing the link MTU size can also be used
  in TRILL traffic MTU size testing; in that case, the link-wide Lz
  used in that algorithm is replaced by the port MTU of the RBridge
  sending MTU-probes.  The successfully tested size x MAY be advertised
  as an attribute of this link, using the MTU sub-TLV defined in
  [RFC7176].

  Unlike RBridges, end stations do not participate in the exchange of
  TRILL IS-IS PDUs; therefore, they cannot grasp the traffic link MTU
  size from a TRILL campus automatically.  An operator may collect
  these values using network management tools such as TRILL ping or
  TraceRoute.  Then, the path MTU can be set as the smallest tested




Zhang, et al.                Standards Track                   [Page 10]

RFC 8249                     MTU Negotiation              September 2017


  link MTU on this path, and end stations should not generate frames
  that -- when encapsulated as TRILL Data packets -- exceed this
  path MTU.

8.  Backward Compatibility

  There can be a mixture of Lz-ignorant and Lz-aware RBridges on a
  link.  This configuration will behave properly, although it may
  not be as efficient as it would be if all RBridges on the link are
  Lz aware.

  For an Lz-ignorant RBridge, TRILL IS-IS PDUs are always formatted
  no greater than Sz.  Lz-aware RBridges as receivers can handle these
  PDUs, since they cannot be greater than the link-wide Lz.

  For an Lz-aware RBridge, in the case that link-wide Lz is greater
  than Sz, larger link-local TRILL IS-IS PDUs can be sent out to
  increase efficiency.  Lz-ignorant RBridges as receivers will have
  no problem handling them, since the originatingL1LSPBufferSize value
  of these RBridges had been tested and the link-wide Lz is not greater
  than that value.

  An Lz-ignorant RBridge might not support the link MTU size-testing
  algorithm defined in Section 3 but could be using some algorithm just
  to test for the Sz MTU on the link.  In any case, if an RBridge per
  [RFC6325] receives an MTU-probe, it MUST respond with an MTU-ack
  padded to the same size as the MTU-probe.

9.  Security Considerations

  This document raises no significant new security issues for TRILL.
  In TRILL, RBridges are generally considered to be trusted devices.
  Protection against forged TRILL IS-IS PDUs, including forged Hellos
  containing originatingSNPBufferSize APPsub-TLVs, can be obtained
  through IS-IS PDU cryptographic authentication [RFC5310].  The worst
  that an RBridge can do by reporting an erroneous
  originatingSNPBufferSize is reduce Lz to Sz and thus make unavailable
  the optimization of being able to use link MTUs that exceed the
  campus-wide MTU for link-local TRILL IS-IS PDUs.

  For general and adjacency-related TRILL security considerations, see
  [RFC6325] and [RFC7177].









Zhang, et al.                Standards Track                   [Page 11]

RFC 8249                     MTU Negotiation              September 2017


10.  Additions to Configuration

  Implementation of the features specified in this document adds two
  RBridge configuration parameters, as follows:

10.1.  Per-RBridge Configuration

  Each RBridge implementing the RECOMMENDED LSP resizing damping
  strategy specified in Section 4 has an LSPresizeTime parameter that
  is an integer in the range of 0-65,535 and that defaults to 300.  It
  is the number of seconds for which an RBridge determines that Sz has
  increased before it will create any LSP or E-L1FS FS-LSP fragments.

10.2.  Per-RBridge Port Configuration

  Each RBridge port on which the calculation and use of Lz are
  implemented has an originatingL1SNPBufferSize parameter that is an
  integer in the range of 1470-65,535.  This parameter defaults to the
  minimum of the size that the port can accommodate and the link-local
  IS-IS PDU size that the TRILL implementation can accommodate.

11.  IANA Considerations

  IANA has assigned a new APPsub-TLV type for the TRILL
  originatingSNPBufferSize APPsub-TLV defined in Section 2 of this
  document.  This new type has been assigned from the range less than
  256 in the "TRILL APPsub-TLV Types under IS-IS TLV 251 Application
  Identifier 1" registry.  The entry is as follows:

     Type  Name                      Reference
     ----  ------------------------  ---------
     21    originatingSNPBufferSize  RFC 8249

12.  References

12.1.  Normative References

  [RFC2119]  Bradner, S., "Key words for use in RFCs to Indicate
             Requirement Levels", BCP 14, RFC 2119,
             DOI 10.17487/RFC2119, March 1997,
             <https://www.rfc-editor.org/info/rfc2119>.

  [RFC5310]  Bhatia, M., Manral, V., Li, T., Atkinson, R., White, R.,
             and M. Fanto, "IS-IS Generic Cryptographic
             Authentication", RFC 5310, DOI 10.17487/RFC5310,
             February 2009, <https://www.rfc-editor.org/info/rfc5310>.





Zhang, et al.                Standards Track                   [Page 12]

RFC 8249                     MTU Negotiation              September 2017


  [RFC6325]  Perlman, R., Eastlake 3rd, D., Dutt, D., Gai, S., and A.
             Ghanwani, "Routing Bridges (RBridges): Base Protocol
             Specification", RFC 6325, DOI 10.17487/RFC6325, July 2011,
             <https://www.rfc-editor.org/info/rfc6325>.

  [RFC7176]  Eastlake 3rd, D., Senevirathne, T., Ghanwani, A., Dutt,
             D., and A. Banerjee, "Transparent Interconnection of Lots
             of Links (TRILL) Use of IS-IS", RFC 7176,
             DOI 10.17487/RFC7176, May 2014,
             <https://www.rfc-editor.org/info/rfc7176>.

  [RFC7177]  Eastlake 3rd, D., Perlman, R., Ghanwani, A., Yang, H., and
             V. Manral, "Transparent Interconnection of Lots of Links
             (TRILL): Adjacency", RFC 7177, DOI 10.17487/RFC7177,
             May 2014, <https://www.rfc-editor.org/info/rfc7177>.

  [RFC7356]  Ginsberg, L., Previdi, S., and Y. Yang, "IS-IS Flooding
             Scope Link State PDUs (LSPs)", RFC 7356,
             DOI 10.17487/RFC7356, September 2014,
             <https://www.rfc-editor.org/info/rfc7356>.

  [RFC7357]  Zhai, H., Hu, F., Perlman, R., Eastlake 3rd, D., and O.
             Stokes, "Transparent Interconnection of Lots of Links
             (TRILL): End Station Address Distribution Information
             (ESADI) Protocol", RFC 7357, DOI 10.17487/RFC7357,
             September 2014, <https://www.rfc-editor.org/info/rfc7357>.

  [RFC7780]  Eastlake 3rd, D., Zhang, M., Perlman, R., Banerjee, A.,
             Ghanwani, A., and S. Gupta, "Transparent Interconnection
             of Lots of Links (TRILL): Clarifications, Corrections, and
             Updates", RFC 7780, DOI 10.17487/RFC7780, February 2016,
             <https://www.rfc-editor.org/info/rfc7780>.

  [RFC8174]  Leiba, B., "Ambiguity of Uppercase vs Lowercase in
             RFC 2119 Key Words", BCP 14, RFC 8174,
             DOI 10.17487/RFC8174, May 2017,
             <https://www.rfc-editor.org/info/rfc8174>.














Zhang, et al.                Standards Track                   [Page 13]

RFC 8249                     MTU Negotiation              September 2017


12.2.  Informative References

  [IS-IS]    International Organization for Standardization,
             "Information technology -- Telecommunications and
             information exchange between systems -- 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)", ISO/IEC 10589:2002, Second Edition,
             November 2002.

  [RFC8139]  Eastlake 3rd, D., Li, Y., Umair, M., Banerjee, A., and F.
             Hu, "Transparent Interconnection of Lots of Links (TRILL):
             Appointed Forwarders", RFC 8139, DOI 10.17487/RFC8139,
             June 2017, <https://www.rfc-editor.org/info/rfc8139>.

Acknowledgements

  The authors would like to thank Vishwas Manral for his comments and
  suggestions.

Authors' Addresses

  Mingui Zhang
  Huawei Technologies
  No. 156 Beiqing Rd. Haidian District
  Beijing  100095
  China

  Phone: +86-13810702575
  Email: [email protected]


  Xudong Zhang
  Huawei Technologies
  No. 156 Beiqing Rd. Haidian District
  Beijing  100095
  China

  Email: [email protected]











Zhang, et al.                Standards Track                   [Page 14]

RFC 8249                     MTU Negotiation              September 2017


  Donald Eastlake 3rd
  Huawei Technologies
  155 Beaver Street
  Milford, MA  01757
  United States of America

  Phone: +1-508-333-2270
  Email: [email protected]


  Radia Perlman
  Dell EMC
  505 1st Ave South
  Seattle, WA  98104
  United States of America

  Email: [email protected]


  Somnath Chatterjee
  Cisco Systems
  SEZ Unit, Cessna Business Park
  Outer Ring Road
  Bangalore  560087
  India

  Email: [email protected]
























Zhang, et al.                Standards Track                   [Page 15]