INTERNET-DRAFT Mingui Zhang
Intended Status: Standards Track Xudong Zhang
Updates: 6325 Donald Eastlake
Expires: August 17, 2014 February 13, 2014
TRILL IS-IS MTU Negotiation
The base IETF TRILL protocol has a TRILL campus wide MTU feature,
specified in RFC 6325 and RFC6327bis, that assures that link status
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 optional 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. It updates RFC 6325.
Status of this Memo
This Internet-Draft is submitted to IETF in full conformance with the
provisions of BCP 78 and BCP 79.
Internet-Drafts are working documents of the Internet Engineering
Task Force (IETF), its areas, and its working groups. Note that
other groups may also distribute working documents as
Internet-Drafts are draft documents valid for a maximum of six months
and may be updated, replaced, or obsoleted by other documents at any
time. It is inappropriate to use Internet-Drafts as reference
material or to cite them other than as "work in progress."
The list of current Internet-Drafts can be accessed at
The list of Internet-Draft Shadow Directories can be accessed at
Mingui Zhang, et al Expires August 17, 2014 [Page 1]INTERNET-DRAFT MTU Negotiation February 13, 2014Copyright and License Notice
Copyright (c) 2014 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
(http://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 IS-IS MTU Size . . . . . . . . . . . . . . . . 3
3. Link MTU Size Testing . . . . . . . . . . . . . . . . . . . . . 5
4. Refreshing Campus-Wide Sz . . . . . . . . . . . . . . . . . . . 7
5. Relationship between Port MTU, Lz and Sz . . . . . . . . . . . 8
6. LSP Synchronization . . . . . . . . . . . . . . . . . . . . . . 8
7. Recommendations for Traffic Link MTU Size Testing . . . . . . . 8
8. Backwards Compatibility . . . . . . . . . . . . . . . . . . . . 9
9. Security Considerations . . . . . . . . . . . . . . . . . . . . 9
10. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 10
11. References . . . . . . . . . . . . . . . . . . . . . . . . . . 10
11.1. Normative References . . . . . . . . . . . . . . . . . . . 10
11.2. Informative References . . . . . . . . . . . . . . . . . . 10
Author's Addresses . . . . . . . . . . . . . . . . . . . . . . . . 12
Mingui Zhang, et al Expires August 17, 2014 [Page 2]INTERNET-DRAFT MTU Negotiation February 13, 20141. Introduction
[RFC6325] describes the way how RBridges agree on the campus-wide
minimum acceptable inter-RBridge MTU (Maximum Transmission Unit) size
- the campus-wide Sz to ensure that link state flooding operates
properly and all RBridges converge to the same link state. For the
proper operation of TRILL IS-IS, all RBridges MUST format their LSPs
not greater than the campus-wide Sz. [RFC6327bis] defines the diagram
of state transitions of an adjacency. "Link MTU size is successfully
tested" is part of an event (A6) causing the transition from "2-way"
state to "Report" state for an adjacency. If MTU testing is enabled,
this part means the link MTU testing of size X succeeds, and X is
greater than or equal to the campus-wide Sz [RFC6325]. In other
words, if this link cannot support an MTU of the campus-wide Sz, it
will not be reported as part of the campus topology.
This document specifies a new value, "link-wide Lz" to represent the
link-wide minimum acceptable inter-RBridge MTU size for a specific
link. There are PDUs which are valid only to a local link, such as
CSNPs and PSNPs. These PDUs should be formatted not greater than the
link-wide Lz. Since link-wide Lz is normally greater than the campus-
wide Sz, link scope PDUs can therefore be optionally formatted
greater than the campus-wide Sz up to Lz.
An optional TRILL IS-IS MTU size testing algorithm is specified in
Section 3 to detail the MTU testing method described in Section 4.3.2
of [RFC6325] and in [RFC6327bis]. It's recommended to multicast the
MTU-probes when there are multiple RBridges on a link responding to
the probing with MTU-ack [RFC6327bis]. The testing method and rules
of this draft are devised in a way to minimize the tries of MTU-
probing for testing, which therefore reduces the number of multicast
packets for MTU testing.
1.1. Conventions used in this document
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 RFC 2119 [RFC2119].
2. Link-Wide TRILL IS-IS MTU Size
This draft specifies a new value "Lz" to represent the acceptable
inter-RBridge link MTU size on the local link. Link-wide Lz is the
minimum Lz supported by all RBridges on a specific link. If the link
is usable, Lz will be greater than or equal to the campus wide Sz
MTU. Some TRILL IS-IS PDUs are exchanged only between neighbors
instead of the whole campus. They should be confined by the link-wide
Lz instead of the campus-wide Sz. CSNPs and PSNPs are examples of
Mingui Zhang, et al Expires August 17, 2014 [Page 3]INTERNET-DRAFT MTU Negotiation February 13, 2014
such PDUs. They are exchanged just on the link as part of LSP
[FS-LSP] defines the PDUs which support flooding scopes in addition
to area wide scope and domain wide scope. RBridges on a local link
that support Lz greater than Sz MUST support the L1 Circuit Scoped
(L1CS) flooding. They use that flooding to exchange their maximally
supportable value of "Lz". The smallest value of the Lz collected on
a link, but not less than Sz, is the link-wide Lz.
The maximum sized level 1 link-local PDU, such as PSNP or CSNP, which
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 2.1 SHOULD be
included in a GENINFO TLV [RFC6823] in an L1CS-LSP number zero. If it
is missing from a fragment zero L1CS-LSP or there is no fragment zero
L1CS-LSP, it is assumed that its originating IS is implicitly
advertising its originatingSNPBufferSize value as Sz octets.
| Type | (2 byte)
| Length | (2 byte)
| originatingSNPBufferSize | (2 byte)
Figure 2.1: Lz is reported in the originatingSNPBufferSize TLV.
Type: set to originatingSNPBufferSize subTLV (TRILL APPsub-TLV type
0x0002). Two bytes because this APPsub-TLV appears in an Extended TLV
Length: set to 2.
originatingSNPBufferSize: the local value of
originatingL1SNPBufferSize, limited to 1470 to 65,535 bytes.
Mingui Zhang, et al Expires August 17, 2014 [Page 4]INTERNET-DRAFT MTU Negotiation February 13, 2014
+---+ | +---+
+---+ | +---+
Figure 2.2: Link-wide Lz = 1800 v.s. tested link MTU size = 1700
Even if all RBridges on a specific link have reached consensus on the
value of link-wide Lz, it does not mean that these RBridges can
safely exchange PDUs between each other. Figure 2.2 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 chunk of size 1800, it
will be discarded by B1.
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
CSNP, PSNP and also L1CS-LSP will be formatted no greater than the
tested link MTU size and will be safely transmitted on this link.
As for campus-wide 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 the campus-wide Sz. Each RBridge should
format their "campus-wide" PDUs, for example LSPs, not greater than
what they believe to be the campus-wide Sz.
3. Link MTU Size Testing
[RFC6327bis] defines the event A6 as including "MTU test is
successful" if the MTU testing is enabled. As described in Section
4.3.2 of [RFC6325], this is a combination of the following event and
Event: The link MTU size has been tested.
Condition: The link can support the campus-wide Sz.
This condition can be efficiently tested by the following "Binary
Search Algorithm" and rules. The MTU-probe and MTU-ack PDUs are
Mingui Zhang, et al Expires August 17, 2014 [Page 5]INTERNET-DRAFT MTU Negotiation February 13, 2014
specified in Section 3 of [RFC6326bis].
X, X1, and X2 are local integer variables.
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), then link MTU size is
set to the size of link-wide Lz and stop.
2) RB1 tries to send an MTU-probe padded to the size 1470.
a) If RB1 fails to receive an MTU-ack from RB2 after k tries, RB1
sets the "failed minimum MTU test" flag for RB2 in RB1's Hello
b) Link MTU size <-- 1470, X1 <-- 1470, X2 <-- link-wide Lz, X <--
[(X1 + X2)/2] (Operation "[...]" returns the fraction-rounded-
up integer.). Repeat Step 1.
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, then:
X2 <-- X and X <-- [(X1 + X2)/2]
2) If RB1 receives an MTU-ack to a probe of size X from RB2 then:
link MTU size <-- X, X1 <-- X and X <-- [(X1 + X2)/2]
3) If X1 >= X2 or Step 1 has been repeated n times (where n is a
configurable parameter whose default value is 5), stop. Else
repeat Step 1.
MTU testing is only done in the Designated VLAN [RFC6327bis]. Since
the execution of the above algorithm can be resource consuming, it is
recommended that the DRB takes the responsibility to do the testing.
Multicast should be used instead of unicast when multiple RBridges
are desired to respond with MTU-ack on the link. The Binary Search
Algorithm is proposed here to minimize the tries of probing therefore
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 (X2) and lower bound (X1)
for the link MTU size from the execution of the above algorithm. If
Mingui Zhang, et al Expires August 17, 2014 [Page 6]INTERNET-DRAFT MTU Negotiation February 13, 2014
the campus-wide Sz is smaller than the lower bound or greater than
the upper bound, RBridges can directly judge whether the link
supports the campus-wide Sz without MTU-probing.
(a) If X1 >= campus-wide Sz. This link can support campus-wide Sz.
(b) Else if X2 <= campus-wide Sz. This link cannot support campus-
Otherwise, RBridges need to test whether the link can support campus-
(c) X1 < campus-wide Sz < X2. RBridges need probe the link with MTU-
probe messages padded to campus-wide Sz. If an MTU-ack is
received within k tries, this link can support campus-wide Sz.
Otherwise, this link cannot support campus-wide Sz. Through this
test, the lower bound and upper bound of link MTU size can be
4. Refreshing Campus-Wide Sz
RBridges may join in or leave the campus, which may change the
campus-wide Sz. The following recommendations are specified for
refreshing the campus-wide Sz.
1) When a new RBridge joins in the campus and its
originatingL1LSPBufferSize is smaller than current campus-wide Sz,
reporting its originatingL1LSPBufferSize in its LSPs will cause
other RBridges decrease their campus-wide Sz. Then the LSPs in the
campus will be resized to be no greater than the new campus-wide
2) When an RBrige leaves the campus and its
originatingL1LSPBufferSize is equal to the campus-wide Sz, its
LSPs are purged from the remaining campus after reaching MaxAge
[ISO10589]. The campus-wide Sz may be recalculated and may be
increased. In other words, while RB1 normally ignores link state
information for any IS-IS unreachable [TRILLClear] RBridge RB2,
originatingL1LSPBufferSize is an exception. Its value, even from
IS-IS unreachable RBridges, is used in determining Sz.
Frequent LSP "resizing" is harmful to the stability of the TRILL
campus, so it should be dampened. Within the two kinds of resizing
actions, only the upward resizing will be dampened. When an upward
resizing event happens, a timer is set (this is a configurable
parameter whose default value is 300 seconds). Before this timer
expires, all subsequent upward resizing will be dampened. Of course,
in a well-configured campus with all RBridges configured to have the
Mingui Zhang, et al Expires August 17, 2014 [Page 7]INTERNET-DRAFT MTU Negotiation February 13, 2014
same originatingL1LSPBufferSize, no resizing will be necessary.
If the refreshed campus-wide Sz is smaller than the lower bound or
greater than the upper bound of the tested link MTU size, the
resource consuming link MTU size testing can be avoided according to
rule (a) or (b) specified in Section 3. Otherwise, RBridges need to
test the link MTU size according to rule (c). But it's unnecessary to
perform the link MTU size testing algorithm all over again.
5. Relationship between Port MTU, Lz and Sz
When port MTU size is smaller than the local
originatingL1SNPBufferSize of an RBridge (sort of a wrong
configuration), this port should be explicitly disabled from the
TRILL campus. On the other hand, when an RBridge receives an LSP or
L1CS-LSP with size greater than the link-wide Lz or the campus-wide
Sz but not greater than its port MTU size, this LSP should be
processed rather than discarded. If the size of an LSP is greater
than the MTU size of a port over which it is to be propagated, no
attempt shall be made to propagate this LSP over the port and an
LSPTooLargeToPropagate alarm shall be generated [ISO10589].
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 [RFC6327bis]. On a LAN link, CSNP and PSNP PDUs
are used for synchronization. On a point-to-point link, only PSNP are
The CSNPs and PSNPs MUST be formatted in chunks of size at most the
link-wide Lz but are processed normally if received larger than that.
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 in the
size no greater than it, therefore these PDUs will finally
successfully get through.
Note that the link MTU size is probably greater than the campus-wide
Sz. Link local PDUs are formatted in the size of link MTU size rather
than the campus-wide Sz, which promises a reduction in the number of
PDUs and a faster LSP synchronization process.
7. Recommendations for Traffic Link MTU Size Testing
Campus-wide 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
Mingui Zhang, et al Expires August 17, 2014 [Page 8]INTERNET-DRAFT MTU Negotiation February 13, 2014
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 the campus-wide Sz or
link-wide Lz values.
The algorithm defined in link MTU size testing can also be used in
TRILL traffic MTU size testing, only that the link-wide Lz used in
that algorithm should be replaced by the port MTU of the RBridge
sending MTU probes. The successfully tested size X can be advertised
as an attribute of this link using MTU sub-TLV defined in
Unlike RBridges, end stations do not participate in the exchange of
ISIS PDUs of TRILL, therefore they can not 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 is set as the smallest tested link MTU
on this path and end stations should not generate frames that, when
encapsulated as TRILL Data packets, may exceed this path MTU.
8. Backwards Compatibility
There can be a mixture of Lz-ignorant and Lz-aware RBridges on a
link. This will act properly although it will not be as efficient as
it would be if all RBridges on the link are Lz-aware.
At the side of an Lz-aware RBridge, in case that link-wide Lz is
greater than campus-wide Sz, larger link-local TRILL IS-IS PDUs can
be sent out to gain efficiencies. Lz-ignorant RBridges as receivers
will have no problem to handle them since the
originatingL1LSPBufferSize value of these RBridges had been reported
and the link-wide Lz is not greater than that value.
At the side of an Lz-ignorant RBridge, TRILL IS-IS PDUs are always
formatted not greater than the campus-wide Sz. Lz-aware RBridges as
receivers can handle these PDUs since they cannot be greater than the
An Lz-ignorant RBridge does not support the link MTU testing
algorithm defined in Section 3 but may be using some algorithm just
to test for 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. So the extension of TRILL
MTU negotiation with Lz, as specified in this document, is fully
9. Security Considerations
Mingui Zhang, et al Expires August 17, 2014 [Page 9]INTERNET-DRAFT MTU Negotiation February 13, 2014
This document raises no new security issues for TRILL. For general
and adjacency related TRILL security considerations, see [RFC6325]
10. IANA Considerations
Similar as Section 7.2 of [ESADI], IANA is requested to create a new
subregistry (2 suggested) under the Generic Information TLV (#251)
[RFC6823] for TRILL originatingSNPBufferSize sub-TLV defined in
Section 2 of this document.
11. References 11.1. Normative References
[RFC6325] R. Perlman, D. Eastlake, et al, "RBridges: Base Protocol
Specification", RFC 6325, July 2011.
[RFC6327bis] D. Eastlake, R. Perlman, et al, "TRILL: Adjacency",
draft-ietf-trill-rfc6327bis-04.txt, in RFC Editor's
[RFC6326bis] D. Eastlake, T. Senevirathne, et al, "Transparent
Interconnection of Lots of Links (TRILL) Use of IS-IS",
draft-ietf-isis-rfc6326bis-03.txt, in RFC Editor's
[FS-LSP] L. Ginsberg, S. Previdi and Y. Yang, "IS-IS Flooding
Scope LSPs", draft-ietf-isis-fs-lsp-01.txt, work in
[TRILLClear] D. Eastlake, M. Zhang, et al., "TRILL: Clarifications,
Corrections, and Updates", draft-ietf-trill-clear-
correct-06.txt, in RFC Editor's queue.
[RFC6823] Ginsberg, L., Previdi, S., and M. Shand, "Advertising
Generic Information in IS-IS", RFC 6823, December 2012.
11.2. Informative References
[ISO10589] ISO, "Intermediate system to Intermediate system
routeing information exchange protocol for use in
conjunction with the Protocol for providing the
Connectionless-mode Network Service (ISO 8473)," ISO/IEC
[ESADI] H. Zhai, F. Hu, et al., "TRILL (Transparent
Interconnection of Lots of Links): ESADI (End Station
Mingui Zhang, et al Expires August 17, 2014 [Page 10]INTERNET-DRAFT MTU Negotiation February 13, 2014
Address Distribution Information) Protocol", draft-ietf-
trill-esadi-05.txt, work in progress.
Mingui Zhang, et al Expires August 17, 2014 [Page 11]INTERNET-DRAFT MTU Negotiation February 13, 2014Author's Addresses
No.156 Beiqing Rd. Haidian District,
Beijing 100095 P.R. China
No.156 Beiqing Rd. Haidian District,
Beijing 100095 P.R. China
Donald E. Eastlake, 3rd
155 Beaver Street
Milford, MA 01757 USA
19111 Pruneridge Ave.
Cupertino, CA 95014 USA
Mingui Zhang, et al Expires August 17, 2014 [Page 12]