TRILL Tutorial - Ethernet Technology Summit

TRILL Tutorial 

Donald E. Eastlake, 3 rd 

Huawei R&D USA 

 

San Jose, CA USA February 

2012 1

The Speaker 


u Principal Engineer, Huawei R&D USA 

Co-Chair, IETF TRILL Working Group 

Voting member of IEEE 802.1 Working Group 

Chair, IETF PPPEXT Working Group 

Author of over 50 IETF RFCs 

Note: 

§ 

This tutorial is a high level overview. It is not possible to include 

all the details in a presentation of this length. 


2012 2

CONTENTS 

§ Introduction 

• What is TRILL? 

• Least Cost Paths 

• Unicast Multi-Pathing 

• TRILL History 

• TRILL Features 

§ TRILL Examples 

§ TRILL Availability 

§ Technical Details 

§ TRILL Protocol Layer 

§ Example Processing 

§ Multi-Destination Traffic 

§ TRILL Support of DCB 

§ TRILL Comparisons 

§ References 


2012 3

What is TRILL? 

§ TRansparent Interconnection of Lots of Links 

• TRILL WG Charter 

– http://www.ietf.org/dyn/wg/charter/trill-charter.html 

• A standard specified by the IETF TRILL WG co-chaired by 

– Donald E. Eastlake 3 rd , Huawei Technologies 

– Erik Nordmark, Cisco Systems 

§ TRILL Switch / RBridge (Routing Bridge) 

• Devices that implements TRILL 

§ TRILL/RBridge Campus – 

• A network of RBridges, links, and any intervening 

bridges, bounded by end stations / layer 3 routers. 


2012 4


§ A Compatible Protocol 

• Attached end nodes just think it is Ethernet. 

§ The more classic bridges you convert to 

TRILL Switches, the better your network’s 

stability and bandwidth utilization. 


2012 5


§ Basically a simple idea: 

• Encapsulate native frames in a transport header 

providing a hop count 

• Route the encapsulated frames using IS-IS 

• Decapsulate native frames before delivery 

§ Provides 

• Least cost paths with zero/minimal configuration 

• Equal Cost Multi-Pathing of unicast 

• Multi-paths of multi-destination 


2012 6

Least Cost Paths 

B1 

= end sta-on 

B3 

B2 

A three bridge network 


2012 7


B1 

= end sta-on 

B3 

B2 

Spanning tree eliminates loops 

by disabling ports 


2012 8


RB1 

= end sta-on 

RB3 

RB2 

A three RBridge network: better performance 

using all facilities 


2012 9

Unicast Multi-Pathing 

B1 

B3 

B5 

B2 

B4 

= end sta)on 

Bridges limit traffic to one path 


2012 10

Unicast Multi-Pathing 

RB1 

RB3 

RB5 

RB2 

RB4 

RBridges support 

multi-path for higher throughput 

= end sta)on 


2012 11

TRILL History up to 2009 

l 

l 

l 

l 

l 

l 

l 

l 

l 

l 

l 

l 

1964: Packet switching/routing invented by Paul Baran. 

1973: Ethernet invented by Robert Metcalfe 

1979: Link State Routing invented by John McQuillan. 

1985: Radia Perlman invents the Spanning Tree Protocol. 

1987: DECnet Phase V / IS-IS designed by Radia Perlman. 

2002: Beth Israel Deaconess Hospital network in Boston melts 

down due to deficiencies in the Spanning Tree Protocol. 

2004: TRILL presented by inventor Radia Perlman at Infocom. 

2005: TRILL presented to IEEE 802 by Radia Perlman, rejected. 

2005: IETF Charters the TRILL Working Group. 

2008: MTU problem delays protocol while fix is incorporated. 

2009: RFC 5556 “TRILL: Problem and Applicability Statement” 

2009: TRILL Protocol passed up to IESG for Standards Approval. 


2012 12

TRILL in 2010/2011 

l 2010: TRILL approved as an IETF Standard (March 15, 2010) 

l 

l 

l 

l Ethertypes, Multicast addresses & NLPID assigned 

2010: Successful TRILL control plane interop at UNH IOL 

2011: TRILL Protocol base document set published: 

l 

l 

l 

l 

l 

RFC 6325: “RBridges: TRILL Base Protocol 

Specification” (Includes TRILL over Ethernet) 

RFC 6326: “TRILL Use of IS-IS” 

RFC 6327: “RBridges: Adjacency” 

RFC 6361: “TRILL over PPP” 

RFC 6439: “RBridges: Appointed Forwarders” 

2011: TRILL Working Group Re-Chartered to do further 

development of the TRILL protocol 


2012 13

TRILL Features 

Bridges 

TRILL 

Switch 

Routers 

§ Transparency 

§ Plug & Play 

§ Virtual LANs 

• Multi-tenant support 

§ Frame Priorities 

§ Data Center Bridging 

§ Virtualization Support 

§ Multi-pathing 

§ Optimal Paths 

§ Rapid Fail Over 

§ The safety of a TTL 

• Implemented in data plane 

§ Extensions 


2012 14

Yet More TRILL Features 

§ Breaks up spanning tree for greater stability. 

§ Unicast forwarding tables at transit RBridges scale 

with the number of RBridges, not the number of end 

stations. Transit RBridges do not learn end station 

addresses. 

§ Compatible with existing IP Routers. TRILL switches 

are as transparent to IP routers as bridges are. 

§ Jumbo frame support including jumbo routing frames 

§ Has a poem (see next slide) 

• The only other bridging or routing protocol with a poem is 

Spanning Tree. (see Algorhyme on the web) 


2012 15

Algorhyme V2 

§ 

§ 

§ 

§ 

§ 

§ 

§ 

§ 

§ 

§ 

§ 

§ 

§ 

I hope that we shall one day see 

A graph more lovely than a tree. 

A graph to boost efficiency 

While still configuration-free. 

A network where RBridges can 

Route packets to their target LAN. 

The paths they find, to our elation, 

Are least cost paths to destination! 

With packet hop counts we now see, 

The network need not be loop-free! 

RBridges work transparently, 

Without a common spanning tree. 

- By Ray Perlner 


2012 16

CONTENTS 



• Acme Power Plant 

• Acme Data Center 








§ References 


2012 17

A 

A A A A A A A A A A A 

Acme Power Plant 

Rapid Spanning 

Tree Protocol 

Domain 

GB GB GB GB 

GB 

GB 

A A A A A 

CB 

A 

CB 

Bridged Process 

Control Network 

A = Access Bridge 

GB = aGgrega-on 

Bridge 

CB = Core Bridge 


2012 18

A 



Rapid Spanning 

Tree Protocol 

Domain 

GB GB GB GB 

GB 

GB 

A A A A A 

CB 

A 

CB 

Spanning Tree 

Eliminates 

Loops by 

Disabling Ports 



Bridge 

CB = Core Bridge 


2012

A 



RSTP 

Domain 

RSTP 

Domain 

GB GB GB GB 

GB 

CRB 

GB 

A A A A A A 

CRB 

RSTP 

Domain 

Process Control 

Network with RBridge 

Core breaking up 

spanning tree 



Bridge 

CRB = Core RBridge 


2012

A A A A A A 

A 

A A A A A 


GRB GRB GRB GRB 

GRB 

GRB 

A A A A A A 

Process Control 

Network with RBridge 

Mesh eliminating 

spanning tree 


GRB = aGgrega-on 

RBridge 


2012

Acme Data Center 

1:1 Backup 

Distribution Bridges 

must be able to handle 

100% of the load. Only 1 

path available between 

any pair of “B”s. 

Wan 

Router 

Dist. 

Bridge 

Wan 

Router 

Dist. 

Bridge 

Data Center 

Network with 

Bridges 

B = Head of Rack 

Bridge 

B B B B B B B B B B B B B B B 


2012

Acme Data Center 

N:1 Backup 

Distribution Bridges need to 

handle only 25% of the load. 

Multiple available paths 

between “H”s. 

H = Head of Rack 

RBridge 

Dist. 

RBridge 

Dist. 

RBridge 

Wan 

Router 

Dist. 

RBridge 

Wan 

Router 

Data Center 

Network with 

RBridges 

Dist. 

RBridge 

Dist. 

RBridge 

H H H H H H H H H H H H H H H 


2012

CONTENTS 










§ References 


2012 24

TRILL Products 

§ “Pre-standard” products 

– Both to be convertible to standard TRILL 

• Cisco FabricPath 

• Brocade VCS 

§ IBM / Blade Networks RackSwitch G8264 

support for the TRILL Standard. 

§ Many additional TRILL standard supporting 

products to be announced in 2012. 


2012 25

TRILL Silicon Availability 

• Here are six publicly known independent silicon 

implementations of the TRILL Fast Path. In some cases there 

are multiple different chips. 

• Broadcom – merchant silicon 

• Brocade – VCS product 

• Cisco – FabricPath product 

• Fulcrum – merchant silicon 

• Marvell – merchant silicon 

• Mellanox – merchant silicon 


2012 26

Open Source TRILL 

• Oracle: TRILL for Solaris 

• http://hub.opensolaris.org/bin/view/Project+rbridges/WebHome 

• TRILL Port to Linux (in process): 

National University of Sciences and Technology 

(NUST), 

§ 

§ 

Dr. Ali Khayam 

Islamabad, Pakistan 

• http://www.wisnet.seecs.nust.edu.pk/people/~khayam/index.php 


2012 27

CONTENTS 










§ References 


2012 28

TRILL is Based on IS-IS 

§ TRILL switches (RBridges) use IS-IS link state 

routing 

• Neighbor RBridges find each other by exchanging Hellos 

• This information is flooded so all RBridges in the campus 

know about all adjacencies. Then all RBridges can 

– calculate the topology for least cost unicast forwarding, 

including Equal Cost Multi-Pathing 

– Calculate the same distribution trees for multi-destination 

frames 

• Other flooded information supports nicknames (see later 

slide), optimization of multicast distribution based on VLAN 

attachment and multicast listeners, etc. 


2012 29

TRILL is Based on IS-IS 

§ The IS-IS (Intermediate System to Intermediate 

System) link state routing protocol was chosen for 

TRILL over IETF OSPF (Open Shortest Path First), 

the only plausible alternative, for the following 

reasons: 

• IS-IS runs directly at Layer 2. Thus no IP addresses are 

needed, as they are for OSPF, and IS-IS can run with zero 

configuration. 

• IS-IS uses a TLV (type, length, value) encoding which 

makes it easy to define and carry new types of data. 


2012 30

TRILL Nicknames 

§ TRILL Switches are identified by IS-IS System ID but 

also by 2-bytes nicknames. 

§ Nicknames can be configured but by default are autoallocated. 

In case of collisions, the lower priority 

RBridge must select a new nickname. 

§ Nicknames: 

• Saves space in headers. 

• An RBridge can hold more than one nickname so that 

– It can be the root of more than one different distribution tree. 

– May be used to distinguish frames following traffic engineered 

routes versus least cost routes. 


2012 31

MAC Address Learning 

§ By IS-IS all TRILL Switches in the campus learn 

about and can reach each other but what about 

reaching end station MAC addresses? 

• By default, TRILL Switches at the edge (directly connected 

to end stations) learn attached VLAN/MAC addresses from 

data. 

• Optionally, MAC addresses can be passed through the 

control plane. 

• MAC addresses can be statically configured. 

• Transit TRILL Switches do not learn end station addresses. 


2012 32

RBridges on an Access Link 

§ You can have multiple RBridges on a link with one or 

more end stations. 

§ One is elected to be in charge of the link and to 

handle end station traffic. But to load split, it can each 

VLAN to another RBridge on the link. 

RB1 RB2 RB3 

B1 B2 B3 


2012 33

TRILL Encapsulation 

§ When an edge TRILL Switch receives an end station 

(native) frame, it encapsulates it with a TRILL Header 

and then adds a link header and forwards it. 

• On Ethernet, the local link header is addressed 

from the local source RBridge to the next hop 

RBridge for unicast frames or to the All-RBridges 

multicast address for multi-destination frames. 

• The TRILL Header specifies the ingress RBridge 

and either the egress RBridge for unicast frames 

or the distribution tree for multi-destination frames. 


2012 34

TRILL Encapsulation 

§ Summary of reasons for encapsulation: 

• Provides a hop count to mitigate loop issues 

• To hide the original source address to avoid confusing any 

bridges present as might happen if multi-pathing were in use 

• To direct unicast frames toward the egress RBridge so that 

forwarding tables in transit RBridges need only be sized with 

the number of RBridges in the campus, not the number of 

end stations 

• To provide a separate outer VLAN tag, when necessary, for 

forwarding traffic between RBridges, independent of the 

original VLAN of the frame 


2012 35

TRILL Details 

§ TRILL Header 

TRILL Ethertype 

V 

R 

M 

ExtLng 

Hop 

Egress RBridge Nickname 

Ingress RBridge Nickname 

• Nicknames – auto-configured 16-bit campus local names for 

TRILL Switches (RBridges) 

• V = Version (2 bits) 

• R = Reserved (2 bits) 

• M = Multi-Destination (1 bit) 

• ExtLng = Length of TRILL Header Extensions 

• Hop = Hop Limit (6 bits) 


2012 36

TRILL Over Ethernet 

Data: 

DA 

SA 

VLAN * 


Ethertype 

0x22F3 


Header 

Payload Frame 

(DA, SA, VLAN/Tenant, Data) 

FCS 

Ethernet Link 

Transport Header 

Original Frame with 

VLAN/Tenant Label 

IS-IS: 

DA 

SA VLAN * L2-IS-IS 

Ethertype 

0x22F4 

IS-IS PDU 

FCS 

RBridge 

One 

Ethernet 

Cloud 

RBridge 

Two 

* Link Transport VLAN only needed for VLAN sensitive link. 

RBridge 

Three 


2012 37

TRILL Over PPP 

Data: 

PPP TRILL 

Data Protocol 

0x005D 


Header 

Payload Frame 

(DA, SA, VLAN/Tenant, Data) 

PPP 

FCS 

PPP Link 

Transport Header 

Original Frame with 

VLAN/Tenant Label 

IS-IS: 

PPP TRILL 

IS-IS Protocol 

0x405D 

IS-IS PDU 

PPP 

FCS 

RBridge 

One 

PPP 

RBridge 

Two 


2012 38

CONTENTS 










§ References 


2012 39

TRILL Protocol Layer 

§ TRILL operates at Layer 2 ½ 

§ Layer 3 

§ TRILL 

Routers 

(plus servers and other end stations) 

TRILL Switches 

§ Layer 2 

Bridges 

Hubs/Repeaters 


2012 40


§ Direct Connection at the same protocol layer 

Device 

Peers 

Connection 

Device 


2012 41


§ Former Situation 

Router / 

End 

Station 

Peers 

Router / 

End 

Station 

Bridge 

Bridge 

Non-Peers 


2012 42


§ Former Situation 

• or maybe 

Router / 

End 

Station 

Peers 

Router / 

End 

Station 

Customer 

Bridge 

Customer 

Bridge 

Peers 

Customer 

Bridge 

Non-Peers 

Provider 

Bridge(s) 

Provider 

Bridge(s) 


2012 43


§ With TRILL Switches 

Router / 

End 

Station 

Peers 

Router / 

End 

Station 


Switch 

Peers 


Switch 

Peers 


Switch 

Non-Peers 

Bridge(s) 

Bridge(s) 


2012 44

CONTENTS 










§ References 


2012 45

Input Port Processing 

§ Detailed example of unicast frame TRILL 

routing on an Ethernet link 

Input Na)ve Frame on link: 

Dest MAC 

Src MAC 

Data 

FCS 

VLAN 

• Input port adds VLAN-ID and priority if frame untagged 

Input Na)ve Frame a7er input port: 

Dest MAC 

Src MAC 

VLAN 

Data 

FCS 


2012 46

TRILL Unicast Ingress 

Input Na)ve Frame: 

Dest MAC 

Src MAC 

VLAN 

Data 

FCS 

Look Up Egress, Next 

Hop DA & Output Port 

Output TRILL Data Frame: 

DA 

SA VLAN 1 

TTL=n 

Egress 

TRILL Header 

Ingress 

Payload Frame 

New 

FCS 

Link Transport 

Header 

Ingressing 

RBridge 

Original Frame with 

VLAN or Tenant ID 

1 

Outer VLAN tag is a transport artifact and only needed if RBridges are connected by a 

bridged LAN or carrier Ethernet requiring a VLAN tag or the like. 


2012 47

TRILL Unicast Transit 

DA 

Incoming Link 

Transport Header 

SA 

VLAN 1 

TRILL Hdr 

Egress 

TTL=n 

Ingress 

Input TRILL Data Frame: 

Payload Frame 

FCS 

Transit RBridge 

Look Up Next 

DA & Output Port 

DA 

SA VLAN 1 

Outgoing Link 

Transport Header 


Egress 

TTL=n-1 

Ingress 

Payload Frame 

New 

FCS 

Output TRILL Data Frame: 

1 

Input and output Outer VLANs can differ. The true VLAN or 

Tenant ID of the data is inside the payload frame. Outer 

VLAN is only needed if link is VLAN sensitive. 


2012 48

TRILL Unicast Egress 

Link Transport 

Header 

DA 

SA 

VLAN 1 


Egress 

Egressing 

RBridge 

Ingress 

Input TRILL Data Frame: 

Payload Frame 

FCS 

Output Na)ve Frame: 

Dest MAC 

Src MAC VLAN 2 

Data 

New FCS 

Look Up Output Port 

1 

Outer VLAN only needed if RBridges are connected by a bridged 

LAN or carrier Ethernet requiring a VLAN tag or the like 

2 

Final native frame VLAN tag may be omitted depending on 

RBridge output port configuration. 


2012 49

Output Port Processing 

Output Na)ve Frame before output port: 

Dest MAC 

Src MAC 

VLAN 

Data 

New FCS 

§ Output port may be configured to output 

untagged and will do so by default for the port 

VLAN ID 

Dest MAC 

Src MAC 

Data 

New FCS 


2012 50

CONTENTS 










§ References 


2012 51

Multi-Destination Traffic 

§ Multi-destination frames are send on a bidirectional 

distribution tree. 

• The root of a tree is a TRILL Switch or a link 

(pseudo-node) determined by a separate priority 

and represented by nickname. 

• The ingress RBridge picks the tree, puts the tree 

root nickname in the “egress nickname” slot, and 

sets the M bit in the TRILL Header. 

§ All the RBridges in a campus calculate the 

same trees. 


2012 52


§ Multi-destination frames are more dangerous 

than unicast because they can multiply at fork 

points in the distribution tree. 

• So, in addition to the Hop Count, a Reverse Path 

Forwarding Check is performed. This discards the 

frame if, for the ingress and tree, it seems to be 

arriving on the wrong port. 

• To reduce the RPFC state, ingress routers can 

announce which tree or trees they will use. 


2012 53


§ As a frame is propagated on a distribution 

tree, its distribution can be pruned by VLAN 

and by multicast group since it is not useful to 

send a frame down a tree branch if 

• There are no end stations downstream in the 

VLAN of the frame, or 

• The frame is multicast and there is no multicast 

listener or multicast router downstream. 


2012 54


§ A single RBridge can even be the root of multiple 

least cost trees which may be useful if it is the source 

or sink of lots of multi-destination traffic. 

RB5 RB6 RB7 

RB1 

RB2 

RB3 

RB4 

= end sta)on 


2012 55

CONTENTS 










§ References 


2012 56

TRILL Support of DCB 

§ The goal is “loss-less” Ethernet. That is, no 

loss due to queue overflow 

§ Basic Ethernet PAUSE “works”, but is a very 

blunt instrument 

• Interference with loss dependent flow control such 

as TCP 

• Blocking of high priority control frames 

• Congestion spreading 


2012 57


§ Answer 1: 

• Consider different frame priorities as different pipes 

– 802.1Qbb: Separate PAUSE per priority 

– 802.1Qaz: Allocate bandwidth between priorities 

§ Answer 2: 

• Provide back pressure on the origin of congesting flows 

– 802.1Qau: Congestion Notification (CN) 

§ Combining them may work best 


2012 58


§ Answer 1: Consider different frame priorities as 

different pipes 

• 802.1Qbb: Separate PAUSE per priority 

– Don’t enable for priorities where urgent control frames are sent 

or where loss dependent flow control is in use 

– Enable for priorities where loss-less flow is more important. 

• 802.1Qaz: Ability to allocate bandwidth between these pipes 

– Highest priority frames not restricted 

– Remainder of bandwidth can be carved up and frames can be 

selected in preference to “higher priority” frames if they have not 

used the allocation for their pipe. 

• The above are implemented in port queuing. 


2012 59


§ For PFC and ETS, all stations are peers: 

Router / 

End 

Station 

Peers 

Router / 

End 

Station 

Peers 

RBridge 

Peers 

RBridge 

Bridge 

Bridge 


2012 60


§ Answer 2: Congestion Notification (CN): Provide back 

pressure on the origin of congesting flows 

• When queue depth exceeds a bound, send a Congestion 

Notification Message CNM back to source MAC address in 

the congesting frame’s VLAN 

• Enabled per priority. (CNM itself usually priority 6.) 

• Frames can be labeled with a CN tag for more fine grained 

flows 

• Mostly implemented in port logic 

§ In TRILL a CN tag, if present, goes inside the 

encapsulated frame and a CNM is just a native 

frame, except for one corner case. 


2012 61


Link 

Header 


Header 

Orig. 

DA 

Orig. 

SA 

Orig. 

VLAN 

CN 

Tag 

Original 

Payload 

FCS 

Encapsulated Frame 

§ However, RBridges may have to handle 

CNMs generated by TRILL ignorant bridges 

between RBridges. Such a CNM will be 

initially addressed to the previous hop 

RBridge, not the original end station. 


2012 62


Origin 

RBridge 

Bridge(s) 

RBridge 

Adjusted CNM 

CNM 

§ Previous hop RBridge has to adjust the CNM 

so that it goes back to the origin end station. 

§ Note: All of the Data Center Bridging facilities depend 

on appropriate engineering, limited delay bandwidth 

product, etc., to actually provide “loss-less” service. 


2012 63

CONTENTS 










§ References 


2012 64

TRILL Comparisons 

§ TRILL versus MPLS 

• MPLS is an older, more mature technology with 

better Quality of Service features, etc. 

• MPLS is more configuration intensive. TRILL can 

be auto-configuring. 

• TRILL provides easier support of multicast 

• TRILL can usually scale better because 

– MPLS requires a label entry at each LSR (Label Switched 

Router) for each MPLS path through that LSR 

– TRILL requires a nickname entry at each RBridge for 

each TRILL switch in the campus 


2012 65


§ TRILL versus IP 

• IP is an older, more mature technology 

• IP chips are cheaper than TRILL chips because of 

greater volume but TRILL forwarding is simpler so, 

with equal volume, TRILL would be cheaper 

• TRILL supports VM mobility. Changing subnets 

changes IP Address, breaking TCP connections 

• TRILL is better at multicast because 

– IP requires a complex protocol like PIM to do multicast 

– TRILL has simple multicast distribution, with pruning for 

optimization, designed in from the start 


2012 66


§ IETF TRILL versus IEEE 802.1aq 

• 802.1aq (Shortest Path Bridging) is better integrated with the 

previous bridging standards, for example it uses 802.1 OAM 

• TRILL breaks up Spanning Tree into smaller regions as 

much as it can. 802.1aq, although it provides shortest paths 

like TRILL, still propagates Spanning Tree Protocol, 

connecting Spanning Trees into bigger trees. 

• TRILL supports both point-to-point and multi-access links 

and efficient encoding on non-Ethernet links. 802.1aq is 

designed only for point-to-point Ethernet links. 

• Unicast routing computation for N switches is O(N*(log N)) 

for TRILL and O(N 2 *(log N)) for 802.1aq 


2012 67


§ Overall: 

• All of the protocols are being extended or 

related projects being started, so all are 

being improved. 

• Different technologies have different 

strengths – you should use the best one for 

your situation. 


2012 68

CONTENTS 










§ References 


2012 69

TRILL IETF RFCs 

§ The core TRILL protocol RFCs 

• RFC 6325, “RBridges: TRILL Base Protocol 

Specification” 

• RFC 6326, “TRILL Use of IS-IS” 

• RFC 6327, “RBridges: Adjacency” 

• RFC 6361, “TRILL over PPP” 

• RFC 6439, “RBridges: Appointed Forwarders” 

§ Related RFC 

• RFC 5556, “TRILL Problem and Applicability” 


2012 70

Other References 

§ Introductory Internet Protocol Journal Article 

• http://www.cisco.com/web/about/ac123/ac147/ 

archived_issues/ipj_14-3/143_trill.html 

§ MIB 

• https://datatracker.ietf.org/doc/draft-ietf-trill-rbridge-mib/ 

§ BFD Support 

• https://datatracker.ietf.org/doc/draft-ietf-trill-rbridge-bfd/ 

§ Multi-Tenant Support 

• https://datatracker.ietf.org/doc/draft-ietf-trill-fine-labeling 


2012 71

END 


Huawei R&D USA 

 


2012 72

TRILL Tutorial - Ethernet Technology Summit

Create successful ePaper yourself

Delete template?

Save as template?