Switching Units. #2#2 Types of switching elements Telephone switches m switch samples Datagram routers m switch datagrams ATM switches m switch.

#1

Switching Units

#2

Types of switching elements

Telephone switches switch samples

Datagram routers switch datagrams

ATM switches switch ATM cells

INPUTS

OUTPUTS

#33

Look Inside a Router

Two key router functions: run routing algorithms/protocol (RIP, OSPF, BGP) switching datagrams from incoming to outgoing ports

#4

Repeaters, bridges, routers, and gateways Repeaters/Hubs: at physical level (L1) Bridges: at datalink level (L2)

based on MAC addresses discover attached stations by listening

Routers: at network level (L3) participate in routing protocols

Application level gateways: at application level (L7) treat entire network as a single hop

Gain functionality at the expense of forwarding speed for best performance, push functionality as low as

possible

#5

Types of services

Packet vs. circuit switches packets have headers and samples don’t

Connectionless vs. connection oriented connection oriented switches need a call

setup setup is handled in control plane by switch

controller connectionless switches deal with self-

contained datagrams Connectionless (router)

Connection-oriented (switching system)

Packet switch

Internet router ATM switching system

Circuit switch

?? Telephone switching system

#6

Other switching unit functions Participate in routing algorithms

to build routing tables Next Lecture!

Resolve contention for output trunks buffer scheduling Previous Lecture!

Admission control to guarantee resources to certain streams

#7

Requirements

Capacity of switch is the maximum rate at which it can move information, assuming all data paths are simultaneously active

Primary goal: maximize capacity subject to cost and reliability constraints

Circuit switch must reject call if can’t find a path for samples from input to output goal: minimize call blocking

Packet switch must reject a packet if it can’t find a buffer to store it awaiting access to output trunk goal: minimize packet loss

Subgoal: Don’t reorder packets

#8

Internal switching

In a circuit switch, path of a sample is determined at time of connection establishment No need for a sample header--position in frame is enough

In a packet switch, packets carry a destination field Need to look up destination port on-the-fly

Datagram lookup based on entire destination address

Cell lookup based on VCI – used as an index to a table

Other than that, switching units are very similar

#9

Blocking in packet switches

Can have both internal and output blocking

Internal no path to output Example: head of line blocking.

Output output link busy

If packet is blocked, must either buffer or drop it

#10

Dealing with blocking

Overprovisioning internal links much faster than inputs

Buffers at input or output

Backpressure if switch fabric doesn’t have buffers,

prevent packet from entering until path is available

Parallel switch fabrics increases effective switching capacity

#11

Three generations of packet switches Different trade-offs between cost and

performance Represent evolution in switching

capacity, rather than in technology With same technology, a later generation

switch achieves greater capacity, but at greater cost

All three generations are represented in current products

#12

First generation switch

Most Ethernet switches and cheap packet routers

Bottleneck can be CPU, host-adaptor or I/O bus, depending

computer

queues in memory

CPU

linecard linecard linecard

#13

Second generation switch

Port mapping intelligence in line cards Bottleneck is the bus (or ring)

bus

computer

front end processorsor line cards

#14

Third generation switches

Third generation switch provides parallel paths (fabric)

NxNpacketswitchfabric

OLC

OLC

OLC

IN

ILC

ILC

ILC

OUT

control

#15

Third generation (contd.)

Features self-routing fabric output buffer is a point of contention

• unless we arbitrate access to fabric potential for unlimited scaling,

• as long as we can resolve contention for output buffer

#16

Switching - Fabric

#17

Switching: abstract model

Number of connections: from few (4 or 8) to huge (100K)

#18

Multiplexors and demultiplexors Multiplexor: aggregates sessions

N input lines Output runs N times as fast as input

Demultiplexor: distributes sessions one input line and N outputs that run N

times slower Can cascade multiplexors

De-Mux

12

N

12

N

1 2 NMUX

#19

Time division switching

Key idea: when demultiplexing, position in frame determines output link

Time division switching interchanges sample position within a frame: Time slot interchange (TSI)

MUX

DEMUX

TSI

#20

Time Slot Interchange (TSI) : example

sessions: (1,3) (2,1) (3,4) (4,2)

4 3 2 1 3 1 4 2

1234

Read and write to shared memory in different order

1

2

3

4

2

4

1

3

#21

TSI

Simple to build. Multicast: easy (why?) Limit is the time taken to read and write to

memory For 120,000 telephone circuits

Each circuit reads and writes memory once every 125 ms.

Number of operations per second : 120,000 x 8000 x2 each operation takes around 0.5 ns => impossible with

current technology

Need to look to other techniques

#22

Space division switching

Each sample takes a different path through the switch, depending on its destination

Crossbar: Simplest possible space-division switch

Crosspoints can be turned on or off

inputs

outputs

#23

Crossbar - example

1

2

3

4

1 2 3 4

sessions: (1,2) (2,4) (3,1) (4,3)

inpu

ts

output

#24

Crossbar

Advantages: simple to implement simple control strict sense non-blocking Multicast

• Single source multiple destination ports

Drawbacks number of crosspoints, N2

large VLSI space vulnerable to single faults

#25

Time-space switching

Precede each input trunk in a crossbar with a TSI

Delay samples so that they arrive at the right time for the space division switch’s schedule Crosspoint: 4 (not 16)

memory speed : x2 (not x4)

2 1

4 3

MUX

MUX

1

2

3

4

TSI

TSI

1 2

4 3

DeMux DeMux

#26

Finding the schedule

Build a routing graph nodes - input links session connects an input and output

nodes. Feasible schedule Computing a schedule

compute perfect matching.12

34

12

34

#27

Time-Space: Example

2 1

4 3TSI

2 1

3 4

31

24

Internal speed = double link speed

time 1

time 2

TSI

#29

Internal Non-Blocking Types

Re-arrangeable Can route any permutation from inputs to outputs.

Strict sense non-blocking Given any current connections through the switch. Any unused input can be routed to any unused

output.

Wide sense non-blocking. There exists a specific routing algorithm, s.t., for any sequence of connections and releases, Any unused input can be routed to any unused

output, assuming all the sequence was served by the routing

algorithm.

#30

Circuit switching - Space division graph representation

transmitter nodes receiver nodes internal nodes

Feasible schedule edge disjoint paths.

cost function number of crosspoints (complexity of AxB is

AB) internal nodes

#31

Crossbar - example

1

2

3

4

1 2 3 4

#32

Another Examplein

pu

tsou

tputs

#33

Another Example

sessions: (1,3) (2,6) (3,1) (4,4) (5,2) (6,5)

inpu

tsou

tputs

#34

Clos Network

Clos(N, n , k) : N - inputs/outputs; cross-points: 2 (N/n)nk + k(N/n)2

nxk (N/n)x(N/n) kxn

N=6n=2k=2

3x3

3x3

2x2

2x2

2x2

2x2

2x2

2x2N

k N/nN/n

#35

Clos Network - strict sense non-blocking Holds for k 2n-1 Proof Methodology:

Recall: IF [A,B S and |A|+|B| > |S|] then A∩ B≠Ø S= The k middle switches A = middle switches reachable from the inputs B = middle switches reachable from the outputs Our case:

• |S|=k• |A| ≥ k-(n-1)• |B| ≥ k-(n-1)

#36

Clos Network - strict sense non-blocking Holds for k 2n-1 Proof:

Consider an idle input and output Input box connected to at most n-1 middle layer

switches output box connected to at most n-1 middle layer

switches There exists an ”unused" middle switch good for

both.

n x k

k x n

n-1

n-1

#37

Example

Clos(8,2,3)

N=8n=2k=3

4x4

4x4

3x2

3x2

3x2

2x3

2x3

2x3

2x3 4x4 3x2

Need to route a new call

#38

Clos Network

nxk (N/n)x(N/n) kxn

N=6n=2k=2

3x3

3x3

2x2

2x2

2x2

2x2

2x2

2x2

Why is k=n internally blocking?

#39

Clos Network - re-arrangable

Holds for k n Proof:

Consider the routing graph. find a perfect matching. route the perfect matching through a

single middle switch! remaining network is Clos(N-N/n,n-1,k-1)

summary: smaller circuit weaker guarantee

Multicast ?

12

34

12

34

#40

Recursive Construction: basis

The basic element:

The two states:

The dimension: r=0

#41

Recursive Construction: Benes Network

r-1 dimension N/2 size

r-1 dimension N/2 size

#42

Example 16x16

#43

Benes Networks

Symmetry Size:

F(N) = 2(N/2)*4 + 2F(N/2) = O(N log N) Rearrangable

Clos network with k=2 n=2 Proof I:

Build routing graph. Find 2 matchings route one in the upper Benes and the other

in the lower.

#44

Greedy permutation routing

Start with an arbitrary node i1 set i1 to upper.

At the output, o1 , a new constraint, set o2 to lower.

Continue until no new constraint. Completing a cycle.

Continue until done. Solve for the upper and lower Benes

recursively.

#45

Example: Benes Network for r=2

level 0 switches level 2r switches

I1

I2

12

34

56

78

#46

Example

level 0 switches level 2r switches

I1

I2

12

34

56

78

1 2 3 4 5 6 7 81 5 6 8 4 2 3 7

( )

#47

Example

level 0 switches level 2r switches

I1

I2

12

34

56

78

1 2 3 4 5 6 7 81 5 6 8 4 2 3 7

( )

#48

Example

level 0 switches level 2r switches

I1

I2

12

34

56

78

1 2 3 4 5 6 7 81 5 6 8 4 2 3 7

( )

#49

Example

level 0 switches level 2r switches

I1

I2

12

34

56

78

1 2 3 4 5 6 7 81 5 6 8 4 2 3 7

( )

#50

CRS-1 Switch Fabric Overview

1 of 8

2 of 8

8 of 8

1

2

8

1

2

16

40 Gbps

Line Card Line Card

50 Gbps136 Bytes cells Fabric

Chassis

100 Gbps/LC(2)(2.5X Speedup)

1296 x 1296 buffered non-blocking switchMulti-stage Interconnect—3 Stage Benes topology

S1 S2 S3

S1 S2 S3

S1 S2 S3

2 LEVELS OF PRIORITY

HP Low latency traffic

LP Best effort traffic

MULTICAST SUPPORT

1M multicast groups

#51

Basic Router Architecture3 Main components: Line cards,Switching mechanism,

Route Processor(s), Routing Applications

Control Components

Forwarding Component

Interconnect

#52

Strict Sense non-Blocking

N/2 x N/2

N/2 x N/2

.

.

.

.

.

.N/2 x N/2

#53

Properties

Size: F(N) = 2N*6 + 3F(N/2) = O( N1.58 )

strict sense non-blocking Clos network with k=3 n=2

Better parameters: n=sqrt{N}, k=2sqrt{N}-1 recursive size sqrt{N} x sqrt{N} Circuit size O(N log2.58 N)

#54

Cantor Networks

m copies of Benes network. For m = log N its strict sense non-

blocking Network size N log2 N Example

#55

Cantor Network

m=4

#56

Proof Sketch:

Benes network: • 2 log N -1 layers, • N/2 nodes in layer.• Middle layer= layer log N -1

Consider the middle layer of the Benes Networks. There are Nm/2 nodes in in all of them combined. Bound (from below) the number of nodes

reachable from an input and output. If the sum is more than Nm/2:

There is an intersection there has to be a route.

#57

Proof Sketch:

Let A(k) = number of nodes reachable at level k.

A(0)=m A(1)= 2A(0)-1 A(2)=2A(1)-2 A(k)=2A(k-1) - 2k-1 = 2k A(0) - k 2k-1

A(log N -1) = Nm/2 - (log N -1) N/4 Need that: 2A(log N -1) > Nm/2.

2[Nm/2 - (log N -1) N/4] > Nm/2. Hold for m> log N-1.

#58

Advanced constructions

There are networks of size O(N log N). the constants are huge!

Basic paradigm also applies to large packet switches.