Introduction to Computer Networking Guy Leduc Chapter 5 ... · 5 Network Layer Control Plane 5-9 Chapter 5: outline 5.1 Introduction 5.2 ICMP: The Internet Control Message Protocol

1

Network Layer Control Plane 5-1

Introduction to Computer Networking

Guy Leduc

Chapter 5 Network Layer: The Control Plane

Computer Networking: A Top Down Approach, 7th edition. Jim Kurose, Keith RossAddison-Wesley, April 2016.

© From Computer Networking, by Kurose&Ross Network Layer Control Plane 5-2

Chapter 5: Network Layer Control Plane

Chapter goals: ❒  understand principles behind network control

plane :❍  traditional routing algorithms

❒  and their instantiation, implementation in the Internet:❍ OSPF, RIP, BGP

2


Chapter 5: outline

❒  5.1 Introduction❒  5.2 ICMP: The Internet Control Message Protocol❒  5.3 Intra-domain routing algorithms

❍  Graph abstraction and routing metrics❍  Link state routing

•  OSPF❍  Distance Vector routing

•  RIP

❒  5.4 Inter-domain routing❍  BGP


Key Network-Layer Functions❒  forwarding: move packets

from router’s input to appropriate router output

❒  routing: determine route taken by packets from source to destination

data plane

control plane

Two approaches to structuring network control plane:!  per-router control (traditional)!  logically centralized control (software defined networking)(Only the first one will be studied in this course)

3

Per-router control plane

RoutingAlgorithm

Individual routing algorithm components in each and every router interact in the control plane

dataplane

controlplane

4.1 • OVERVIEW OF NETWORK LAYER 309

tables. In this example, a routing algorithm runs in each and every router and both forwarding and routing functions are contained within a router. As we’ll see in Sec-tions 5.3 and 5.4, the routing algorithm function in one router communicates with the routing algorithm function in other routers to compute the values for its forward-ing table. How is this communication performed? By exchanging routing messages containing routing information according to a routing protocol! We’ll cover routing algorithms and protocols in Sections 5.2 through 5.4.

The distinct and different purposes of the forwarding and routing functions can be further illustrated by considering the hypothetical (and unrealistic, but technically feasible) case of a network in which all forwarding tables are configured directly by human network operators physically present at the routers. In this case, no routing protocols would be required! Of course, the human operators would need to interact with each other to ensure that the forwarding tables were configured in such a way that packets reached their intended destinations. It’s also likely that human configu-ration would be more error-prone and much slower to respond to changes in the net-work topology than a routing protocol. We’re thus fortunate that all networks have both a forwarding and a routing function!

Values in arrivingpacket’s header

1

23

Local forwardingtable

header

0100011001111001

1101

3221

output

Control plane

Data plane

Routing algorithm

Figure 4.2 ♦ Routing algorithms determine values in forward tables

M04_KURO4140_07_SE_C04.indd 309 11/02/16 3:14 PM

5-5Network Layer: Data Plane

1

2

0111

values in arriving packet header

3

© From Computer Networking, by Kurose&Ross


Chapter 5: outline




•  RIP


4


ICMP: Internet Control Message Protocol

❒  used by hosts & routers to communicate network-level information❍  error reporting: unreachable

host, network, port, protocol❍  echo request/reply (used by

ping)❒  network-layer “above” IP:

❍  ICMP messages carried in IP datagrams

❒  ICMP message: type, code plus first 8 bytes of IP datagram causing error

Type Code description0 0 echo reply (ping)3 0 dest. network unreachable3 1 dest host unreachable3 2 dest protocol unreachable3 3 dest port unreachable3 6 dest network unknown3 7 dest host unknown4 0 source quench (congestion control - not used)8 0 echo request (ping)9 0 route advertisement10 0 router discovery11 0 TTL expired12 0 bad IP header


Traceroute and ICMP❒  source sends series of UDP

segments to dest❍  first set has TTL=1❍  second set has TTL=2, etc.❍  unlikely port number

❒  when nth set of datagrams arrives to nth router:❍  router discards datagrams❍  and sends source ICMP

messages (type 11, code 0)❍  ICMP messages include name

of router & IP address

❒  when ICMP messages arrive, source records RTTs

stopping criterion:"  UDP segment eventually

arrives at destination host"  destination returns ICMP “port

unreachable” message (type 3, code 3)

"  source stops

3 probes

3 probes

3 probes

5


Chapter 5: outline




•  RIP


Routing protocolsRouting protocol goal: determine “good” paths (equivalently, routes), from sending hosts to receiving host, through network of routers

#  path: sequence of routers packets will traverse in going from given initial source host to given final destination host

#  “good”: least “cost”, “fastest”, “least congested”#  routing: a “top-10” networking challenge!

5-10Network Layer Control Plane© From Computer Networking, by Kurose&Ross

6


u

yx

wv

z2

21

3

1

1

2

53

5

Graph: G = (N,E)

N = set of routers = { u, v, w, x, y, z }

E = set of links ={ (u,v), (u,x), (v,x), (v,w), (x,w), (x,y), (w,y), (w,z), (y,z) }

Graph abstraction of the network

aside: graph abstraction is useful in other network contexts,e.g., P2P, where N is set of peers and E is set of TCP connections


Graph abstraction: costs

u

yx

wv

z2

21

3

1

1

2

53

5 c(x,x’) = cost of link (x,x’) e.g., c(w,z) = 5

cost could always be 1, or inversely related to bandwidth,or related to congestion

Cost of path (x1, x2, x3,…, xp) = c(x1,x2) + c(x2,x3) + … + c(xp-1,xp)

Question: What’s the least-cost path between u and z ?

Routing algorithm: algorithm that finds least-cost path

7


How to set link costs?❒  Many possible ways to define routing metrics,

i.e. to set link costs❒  To achieve minimum hop routing:

❍  Set all link costs to 1❍  Will also minimize (average) link load (and node processing)

•  See next slide❍  Does not necessarily minimize delay, nor congestion!

❒  Changing link costs will surely change the least-cost paths!❒  Link costs can be engineered to optimize the network to

some extent❍  But this usually requires to know the traffic matrix (TM)❍  What is a TM?

•  For every pair of nodes (i,j), TM(i,j) is the amount of traffic entering the network at node i and exiting the network at node j


Minimum hop routing minimises the average link load (for any TM)

❒  Minimizing the average link load (in bps) is equivalent to minimizing the sum of all the link loads. So, remove denominator N from the score.

❒  Routing a new flow of rate R along a given path P will increase the score:

❒  Therefore, minimizing the average link load is equivalent to setting the path P so as to minimize the number of hops

❒  So to achieve this, each link will simply get the static cost = 1❒  Does not depend on Traffic Matrix (why?)

€

Score = Avg _link _ load =

loadii∈links

∑

N

€

Score _ increase =i∈P

∑ R = R ×nb _hops(P )

8


InvCap routing minimises the average link utilisation (for any TM)

❒  Equivalent to minimizing the sum of all the link utilisations❒  Routing a new flow of rate R along a given path P will increase the

score:

❒  Therefore, minimizing the average link utilisation is equivalent to setting the path P so as to minimize

(for all P)

❒  So, to achieve this, each link will simply get the static cost = 1/Ci❒  InvCap routing = link cost is proportional to the inverse of its capacity❒  Does not depend on Traffic Matrix (why?)

€

Score = Avg _link _util =

utilii∈links

∑

N=

loadi

capacity ii∈links

∑

N

€

Score _ increase =i∈P

∑ RCi

= R ×i∈P

∑ 1Ci

€

i∈P

∑ 1Ci


❒  Link delay metric❍  Minimizes delay❍  But delay has several components:

•  Propagation delay•  Transmission delay (= packet_size / link_capacity)•  Queuing delay (variable, depends on the load, difficult to take into

account)❒  Administrative link cost

❍  Any link cost computed so as to optimise a given network score❍  For example to better balance the network load

•  but is traffic matrix dependent in general!❒  Basically any summable quantity

❍  Summable = cost of a path is the sum of the costs of all the links composing this path

Other possible routing metrics

9


B

A CD E

FG

HI

J

KL

M

N

O

B

A CD E

FG

HI

J

KL

M

N

O

A network A sink tree for router B based onthe hop count metric

❒  If a router J is on the optimal path from router I to router K, then the optimal path from J to K also falls along the same route

❒  Consequence: the set of optimal routes from all sources to a destination form a tree rooted at the destination

❒  Similarly: the set of optimal routes from one source to all destinations form a tree rooted at the source

Optimality principle

From Computer Networks, by Tanenbaum © Prentice Hall


Routing Algorithm classificationQ: global or decentralized

information?Global:❒  all routers have complete

topology, link cost info❒  “link state” algorithmsDecentralized: ❒  router knows physically-

connected neighbors, link costs to neighbors

❒  iterative process of computation, exchange of info with neighbors

❒  “distance vector” algorithms

Q: Static or dynamic?Static: ❒  routes change slowly over

timeDynamic: ❒  routes change more

quickly❍  periodic update❍  in response to link cost

changes

10


Chapter 5: outline




•  RIP



A Link-State Routing Algorithm

Principle❒ Graph topology discovery:

❍  “link state broadcast”, see next slides❍ So, all nodes have the same complete topological info

❒  Least-cost path computation❍ Every node computes least-cost paths to all other nodes❍  It uses Dijkstra’s algorithm (see later)❍ This gives forwarding table for that node

11


❒  Link State Packets are composed of❍  the source node, a sequence number and an age (see later)❍  a distance vector limited to the neighbours

Source ofpackets

Neighboursof the source


Building link state packets


The packet buffer for router BSend flags ACK flags

Source Seq. Age A C F A C F DataA 21 60 0 1 1 1 0 0F 21 60 1 1 0 0 0 1E 21 59 0 1 0 1 0 1C 20 60 1 0 1 0 1 0D 21 59 1 0 0 0 1 1

2

875

34

1 6A

B C

D

E F

❒  Packets are flooded selectively❍  Packets are not forwarded on the links they arrived on❍  Duplicate (or older) packets are detected by the sequence number

❒  Packets are acknowledged

Packet received from D via C and F

Clearly, routers do not forward the received Link State packets immediately but put them for a short while in a packet buffer (holding area). Why?

Distributing Link State Packets


12


❒  What if the sequence number wraps around?❍  Solution: choose 32-bit sequence number❍  Needs 137 years to wrap around if one link state packet is sent

every second (in practice one packet per e.g. 10 seconds)❒  What if a router crashes?

❍  It restarts with sequence number 0 and its packets are ignored until the sequence number reaches the previous value

❍  Solution: the age field is decremented by 1 every second and the entry removed when age hits 0

❒  What if a sequence number is corrupted?❍  Same consequence, same solution

Potential problems



A Link-State Routing Algorithm

Dijkstra’s algorithm❒  net topology, link costs

known to all nodes❍  accomplished via “link

state broadcast” ❍  all nodes have same info

❒  computes least-cost paths from one node (“source”) to all other nodes❍  gives forwarding table for

that node❒  iterative: after k iterations,

know least-cost path to k destinations

Notation:❒  c(x,y): link cost from node x

to y; = ∞ if not direct neighbors

❒  D(v): current value of cost of path from source to destination v; = ∞ if v not reachable

❒  p(v): predecessor node along path from source to v

❒  N': set of nodes whose least-cost path definitively known

13


Dijkstra’s Algorithm1 Initialization: 2 N' = {u} 3 for all nodes v 4 if v adjacent to u 5 then D(v) = c(u,v) 6 else D(v) = ∞ 7 8 Loop 9 find w not in N' such that D(w) is a minimum 10 add w to N' 11 update D(v) for all v adjacent to w and not in N' : 12 D(v) = min( D(v), D(w) + c(w,v) ) 13 /* new cost to v is either old cost to v or known 14 shortest path cost to w plus cost from w to v */ 15 until all nodes in N'

(as executed in node u)


w3

4

v

x

u

5

37 4

y

8

z2

7

9

Step N'D(v)

p(v)012345

D(w)p(w)

D(x)p(x)

D(y)p(y)

D(z)p(z)

u ∞ ∞ 7,u 3,u 5,uuw ∞ 11,w 6,w 5,u

14,x 11,w 6,wuwxuwxv 14,x 10,v

uwxvy 12,y

notes:#  construct shortest path tree

by tracing predecessor nodes

#  ties can exist (can be broken arbitrarily)

uwxvyz

Dijkstra’s algorithm: example

14


Dijkstra’s algorithm: example (2) Resulting shortest-path tree from u:

vwxyz

wwxww

destination next hop

Resulting forwarding table in u:

w3

4

v

x

u

5

37 4

y

8

z2

7

9


Dijkstra’s algorithm, discussionAlgorithm complexity: n nodes❒  each iteration: need to check all nodes, w, not in N❒  n(n+1)/2 comparisons: O(n2)❒  more efficient implementations possible: O(nlogn)Oscillations possible when link cost are traffic dependent:❒  e.g., suppose link cost = amount of carried traffic

AD

CB

1 1+ε

ε0

ε

1 1

0 0

initially

AD

CB

given these costs,find new routing….

resulting in new costs

2+ε 0

001+ε 1

AD

CB



0 2+ε

1+ε10 0

AD

CB



2+ε 0

001+ε 1

2+ε

15


Chapter 5: outline




•  RIP



OSPF (Open Shortest Path First)❒  “open”: publicly available

❒  uses Link State (LS) algorithm ❍  LS packet dissemination: router floods OSPF messages to all

other routers in entire domain❍  topology map at each node❍  route computation using Dijkstra’s algorithm

❒  OSPF messages (called LSAs, Link-State Advertisements) are carried directly over IP

16


OSPF “advanced” features❒  designated router: when several routers are connected to the same

subnet, one of them can be designated to exchange LSAs with the others (the latter do not exchange LSAs between themselves, to reduce traffic)

❒  security: LSAs can be authenticated (to prevent malicious intrusion) ❒  multiple same-cost paths allowed

❍  ECMP = Equal Cost MultiPath❒  for each link, multiple cost metrics can be defined for different ToS

(e.g., satellite link cost set “low” for best effort ToS; high for real time ToS)

❒  integrated uni- and multicast support: ❍  Multicast OSPF (MOSPF) uses same topology data base as

OSPF❒  hierarchical OSPF in large domains


Hierarchical OSPFboundary router

backbone router

area 1area 2

area 3

backbone(area 0)

areaborderrouters

internalrouters

17


Hierarchical OSPF

❒  two-level hierarchy: local area, backbone❍  link-state advertisements (LSA) only flooded in area ❍  each node has detailed area topology; only knows

direction (shortest path) to nets in other areas❒  area border routers: “summarize” distances to nets in

own area, advertise to other Area Border routers❒  backbone routers: run OSPF routing limited to backbone❒  boundary routers: connect to other domains


Chapter 5: outline




•  RIP


18


Distance vector algorithm

Bellman-Ford equation (dynamic programming)

let dx(y) := cost of least-cost path from x to ythen dx(y) = min {c(x,v) + dv(y) } v

cost to neighbor v

min taken over all neighbors v of x

cost from neighbor v to destination y


Bellman-Ford example

u

yx

wv

z2

21

3

1

1

2

53

5Knowing dv(z) = 5, dx(z) = 3, dw(z) = 3

du(z) = min { c(u,v) + dv(z), c(u,x) + dx(z), c(u,w) + dw(z) } = min {2 + 5, 1 + 3, 5 + 3} = 4

Node achieving minimum is next hop in shortest path, used in forwarding table

B-F equation says:

19


Distance Vector Algorithm

❒ Dx(y) = estimate of least cost from x to y❒ Node x maintains its distance vector

Dx = [Dx(y): y є N ]❒ Node x:

❍  knows cost to each neighbor v: c(x,v)❍  also maintains its neighbors’ distance vectors:

for each neighbor v, x maintains Dv = [Dv(y): y є N ]


Distance vector algorithm

Key idea: ❒  From time-to-time, each node sends its own distance

vector estimate to neighbors❍  Asynchronous

❒  When node x receives new DV estimate from neighbor, it updates its own DV using B-F equation (thus erasing its previous vector):

Dx(y) ← minv{c(x,v) + Dv(y)} for each node y ∊ N

❒  Under minor, natural conditions, the estimate Dx(y) converges to the actual least cost dx(y)

20


Distance Vector Algorithm

Iterative, asynchronous:each local iteration caused by: ❒  local link cost change ❒  DV update message from

neighbor

Distributed:each node notifies neighborsonly when its DV changes

❍  neighbors then notify their neighbors if necessary

wait for (change in local link cost or msg from neighbor)

recompute estimates

if DV to any dest has changed, notify neighbors

Each node:


x y zxyz

0 2 7∞ ∞ ∞∞ ∞ ∞

from

cost to

from

from

x y zxyz

0

from

cost to

x y zxyz

∞ ∞

∞ ∞ ∞

cost to

x y zxyz

∞ ∞ ∞7 1 0

cost to

∞2 0 1

∞ ∞ ∞

2 0 17 1 0

time

x z12

7

y

node xtable

node ytable

node ztable

Dx(y) = min {c(x,y) + Dy(y), c(x,z) + Dz(y)} = min {2+0, 7+1} = 2

Dx(z) = min {c(x,y) + Dy(z), c(x,z) + Dz(z)} = min {2+1, 7+0} = 3

32

21


x y zxyz

0 2 7∞ ∞ ∞∞ ∞ ∞

from

cost to

from

from

x y zxyz

0 2 3

from

cost tox y z

xyz

0 2 3

from

cost to

x y zxyz

∞ ∞

∞ ∞ ∞

cost tox y z

xyz

0 2 7

from

cost tox y z

xyz

0 2 3

from

cost to

x y zxyz

0 2 3

from

cost tox y z

xyz

0 2 7

from

cost tox y z

xyz

∞ ∞ ∞7 1 0

cost to

∞2 0 1

∞ ∞ ∞

2 0 17 1 0

2 0 17 1 0

2 0 13 1 0

2 0 13 1 0

2 0 1

3 1 02 0 1

3 1 0

time

x z12

7

y

Dx(y) = min {c(x,y) + Dy(y), c(x,z) + Dz(y)} = min {2+0, 7+1} = 2

Dx(z) = min {c(x,y) + Dy(z), c(x,z) + Dz(z)} = min {2+1, 7+0} = 3

node xtable

node ytable

node ztable


Distance Vector: link cost changes

Link cost changes:❒  node detects local link cost change ❒  updates routing info, recalculates

distance vector❒  if DV changes, notify neighbors

“goodnews travelsfast”

x z14

50

y1

At time t0, y detects the link-cost change, updates its DV (Dy(x) ← 1), and informs its neighbors

At time t1, z receives the update from y. It computes a new least cost to x (Dz(x) ← 2) and sends its neighbors its DV

At time t2, y receives z’s update and updates its distance table. y’s least costs do not change and hence y does not send any message to z

22


Distance Vector: link cost changesLink cost changes:❒  good news travels fast ❒  bad news travels slow -

“count to infinity” problem!

❒  Other example (see book):❍  44 iterations before algorithm

stabilizes

A B

1335577!

∞

C

2244668!

∞

D

3335577!

∞

E

4444668!

∞

InitiallyAfter 1 exchangeAfter 2 exchangesAfter 3 exchangesAfter 4 exchangesAfter 5 exchangesAfter 6 exchanges

Propagation of bad news(link A-B is down, c(B,A) = ∞)

x z14

50

y60


Distance to Ain nr. of hops

Dy(x) = 4, 6, 8, … 48, 50, 51Dz(x) = 5, 7, 9, … 49, 50


A patch: Poisoned reverse❒  Also called “split-horizon”❒  If C routes through B to get to A, then C lies to B, namely

❍  C tells B its (C’s) distance to A is infinite•  (so B won’t route to A via C)

❒  With poisoned reverse, we get (when link A-B goes down):

❒  Q: Will this completely solve count-to-infinity problem?

A B

1∞∞∞∞

C

22∞∞∞

D

333∞∞

E

4444∞

InitiallyAfter 1 exchangeAfter 2 exchangesAfter 3 exchangesAfter 4 exchanges

23


A B

C

D

Router

Suppose link CD fails

Evolution of the routing tables in A, B and C for destination D:A B C

(2,C) (2,C) (1,D) Initially(2,C) (2,C) (∞, –) After C’s recomputation, C will notify A and B (3,B) (3,A) (∞, –) After A and B’s recomputations(∞, –) (∞, –) (4,A) After 3 exchanges(∞, –) (5,C) (∞, –) After 4 exchanges(6,B) (∞, –) (∞, –) After 5 exchanges

… … … …


Poisoned reverse is not a panacea

Distance to D Next hopto reach D


Chapter 5: outline




•  RIP


24


RIP (Routing Information Protocol)❒  included in BSD-UNIX distribution in 1982❒  distance vector algorithm

❍  distance metric: # hops (max = 15 hops), each link has cost 1❍  DVs exchanged with neighbors every 30 sec in response

message (aka advertisement)❍  each advertisement: list of up to 25 destination subnets (in IP

addressing sense)

DC

BA

u vw

x

yz

subnet hops u 1 v 2 w 2 x 3 y 3 z 2

From router A to destination subnets:


RIP: example

destination subnet next router # hops to dest w A 2

y B 2 z B 7

x - 1… … …

routing table in router D (≠ the forwarding table)

w x yz

A

C

D B

25


w x yz

A

C

D B

destination subnet next router # hops to dest w A 2

y B 2 z B 7

x - 1… … ...

routing table in router D

A 5

RIP: example A-to-D advertisement (contains also the next router)

dest next router # hops w - 1 x - 1 y D 3 z C 4 … … ...

entry discarded by D that sees itself as next-hop, equivalent to “Poisoned reverse”


Comparison of LS and DV algorithms

Message complexity❒  LS: with n nodes, E links, O(nE) messages sent ❒  DV: exchange between neighbors only, but

❍  depends on number of iterations

Speed of Convergence❒  LS: O(nlogn) algorithm to compute least-cost paths to all other

nodes by Dijkstra❒  DV: number of BF updates depends on number of iterations

❍  possible count-to-infinity

26


Intra-domain Routing Protocols❒  are also known as Interior Gateway Protocols (IGP)❒  most common Intra-domain routing protocols:

❍ OSPF: Open Shortest Path First •  Link State Protocol, recommended by IETF, deployed

❍  IS-IS: Intermediate System to Intermediate System •  Nearly identical to OSPF, standardized by ISO, deployed

❍ RIP: Routing Information Protocol•  Distance Vector Protocol, IETF, only used in small networks

❍  (E)IGRP: (Enhanced) Interior Gateway Routing Protocol•  Distance Vector Protocol, improves RIP, from Cisco


Chapter 5: outline




•  RIP


27


Making routing scalable

scale: with billions of destinations:

❒  can’t store all destinations in routing tables!

❒  routing table exchange would swamp links!

administrative autonomy❒  internet = network of networks❒  each network admin may want

to control routing in its own network

Our routing study thus far - idealization ❒  all routers identical❒  network “flat” (not considering areas in OSPF)… not true in practice

aggregate routers into regions known as “autonomous systems” (AS) (aka “domains”)

inter-AS routing❒  routing among ASes❒  gateway routers perform

inter-domain routing (as well as intra-domain routing)

Internet approach to scalable routing

intra-AS routing#  routing among hosts, routers in

same AS (“network”)#  all routers in AS must run same

intra-domain protocol#  routers in different AS can run

different intra-domain routing protocol

#  gateway router: at “edge” of its own AS, has link(s) to router(s) in other ASes


28


3b

1d

3a

1c2aAS3

AS1AS21a

2c2b

1b

Intra-ASRouting algorithm

Inter-ASRouting algorithm

Forwardingtable

3c

Interconnected ASes

❒  forwarding table configured by both intra- and inter-AS routing algorithms❍  intra-AS sets entries for

internal destinations❍  inter-AS and intra-AS set

entries for external destinations


Inter-AS tasks❒  suppose router in AS1

receives datagram destined outside of AS1❍  router should forward

packet to gateway router, but which one?

AS1 must:1.  learn which destinations

reachable through AS2, which through AS3

2.  propagate this reachability info to all routers in AS1

Job of inter-AS routing!

AS3

AS2

3b

3c3a

AS1

1c1a

1d1b

2a2c

2bothernetworks

othernetworks

29


Chapter 5: outline




•  RIP


Internet inter-AS routing: BGP❒  BGP (Border Gateway Protocol): the de facto inter-

domain routing protocol❍  “glue that holds the Internet together”

❒  BGP provides each AS a means to:❍  obtain subnet reachability information from neighboring

ASes (eBGP, external BGP)❍  propagate reachability information to all AS-internal routers

(iBGP, internal BGP) ❍  determine “good” routes to other networks based on

reachability information and policy❒  allows subnet to advertise its existence to rest of

Internet: “I am here”


30

eBGP, iBGP connections

eBGP connectivityiBGP connectivity

1b

1d

1c1a

2b

2d

2c2a3b

3d

3c3a

AS 2

AS 3AS 1

5-59Network Layer Control Plane

1c

∂

∂

gateway routers run both eBGP and iBGP protocols

© From Computer Networking, by Kurose&Ross

iBGP connectivity ≠ physical connectivity in AS

BGP basics

!  when AS3 gateway router 3a advertises path AS3,X to AS2 gateway router 2c:❍ AS3 promises to AS2 it will forward datagrams towards subnet X

!  BGP session: two BGP routers (“peers”) exchange BGP messages over semi-permanent TCP connection:•  advertising paths to different destination network prefixes (BGP is a “path vector” protocol)

•  Prefix = IP address prefix = range of IP addresses

1b

1d

1c1a2b

2d

2c2a

3b

3d

3c3a

AS 2

AS 3AS 1

BGP advertisement:AS3, X


X

31

Path attributes and BGP routes❒  advertised prefix includes BGP attributes

❍  prefix + attributes = “route”❒  two important attributes:

❍  AS-PATH: list of ASes through which prefix advertisement has passed, e.g, AS 67, AS 17

•  BGP is a path vector protocol (AS-PATH are propagated)•  In a DV protocol, only the distance is propagated

❍  NEXT-HOP: the IP address of the (internal-AS) gateway router leading to first AS in AS-PATH

❒  Policy-based routing:❍  gateway receiving route advertisement uses import policy to

accept/decline path (e.g., never route through AS Y).❍  AS policy also determines whether to advertise path to other

neighboring ASes


BGP path advertisement

!  Based on AS2 policy, AS2 router 2c accepts path AS3,X, propagates (via iBGP) to all AS2 routers

1b

1d

1c1a2b

2d

2c2a

3b

3d

3c3a

AS2

AS3AS1

AS3,X AS2,AS3,X

!  AS2 router 2c receives path advertisement AS3,X (via eBGP) from AS3 router 3a

!  Based on AS2 policy, AS2 router 2a advertises (via eBGP) path AS2, AS3, X to AS1 router 1c


X

32

BGP path advertisement

!  AS1 gateway router 1c learns path AS2,AS3,X from 2a

1b

1d

1c1a2b

2d

2c2a

3b

3d

3c3a

AS2

AS3AS1

AS3,X AS2,AS3,X

Gateway router may learn about multiple paths to destination:

AS3,X

!  AS1 gateway router 1c learns path AS3,X from 3a!  Based on policy, AS1 gateway router 1c chooses e.g. path

AS3,X, and advertises path within AS1 via iBGP5-63Network Layer Control Plane© From Computer Networking, by Kurose&Ross

X


BGP messages

❒  BGP messages exchanged between peers over TCP connection

❒  BGP messages:❍ OPEN: opens TCP connection to remote BGP peer

and authenticates sending BGP peer❍ UPDATE: advertises new path (or withdraws old)❍ KEEPALIVE keeps connection alive in absence of

UPDATES; also ACKs OPEN request❍ NOTIFICATION: reports errors in previous msg; also

used to close connection

33

BGP, OSPF, forwarding table entries

!  recall: 1a, 1b, 1d learn about dest X via iBGP from 1c: “path to X goes through 1c”

1b

1d

1c1a2b

2d

2c2a

3b

3d

3c3a

AS2

AS3AS1

AS3,X AS2,AS3,X

AS3,X

!  1d: OSPF intra-domain routing: to get to 1c, forward over outgoing local interface 1

AS3,X

Q: how does router set forwarding table entry to distant prefix?

12

1

2

dest interface…

…X

…

…1

physical link

local link interfacesat 1a, 1d


X

BGP, OSPF, forwarding table entries

!  recall: 1a, 1b, 1c learn about dest X via iBGP from 1c: “path to X goes through 1c”

1b

1d

1c1a2b

2d

2c2a

3b

3d

3c3a

AS2

AS3AS1

!  1d: OSPF intra-domain routing: to get to 1c, forward over outgoing local interface 1

Q: how does router set forwarding table entry to distant prefix?

dest interface…

…X

…

…2

!  1a: OSPF intra-domain routing: to get to 1c, forward over outgoing local interface 2

1

2


X

34


BGP route selection (simplified)

❒  router may learn about more than one route to some destination prefix, selects route based on (in decreasing preference order):

1.  Local preference value attribute: policy decisionprefer routes through “Customer AS” over routes through “Peer AS” over routes through “Provider AS” (economical viewpoint)

2.  Shortest AS-PATH 3.  Closest gateway router (= BGP NEXT-HOP):

so-called hot potato routing (see next slide)4.  Additional criteria

❒  2d learns (via iBGP) it can route to X via 2a or 2c❒  hot potato routing: choose local gateway that has

least intra-domain cost (2d chooses 2a, because 201 < 263): don’t worry about inter-domain cost!

1b

1d

1c1a2b

2d

2c2a

3b

3d

3c3a

AS2

AS4

AS1

AS3,AS4,X AS1,AS4,X

OSPF link cost

201

152112

263

5-68Network Layer Control Plane

4d

4c4aAS3

X

Hot Potato Routing

35


AS1AS2 AS3 AS4

AS5

AS6 AS7

AS8

AS9 AS10

Information AS6 receives from its neighbor ASes to reach prefixes in AS4:

AS2 uses AS2-AS3-AS4AS7 uses AS7-AS3-AS4AS5 uses AS5-AS6-AS7-AS3-AS4AS9 uses AS9-AS5-AS6-AS7-AS3-AS4

❒  If an AS sees itself in the AS-PATH advertised by a neighbor AS, it discards it, otherwise it would create a loop

❒  More powerful than poisoned reverse (used in DV)❍  Made possible by the presence of the AS-PATH in the advertisements

Discarded by AS6, because these AS-PATHs use AS6 as an intermediate AS

Path Vector protocols avoid routing loops


BGP: achieving policy via advertisements

Figure 4.5-BGPnew: a simple BGP scenario

A

B

C

WX

Y

legend:

customer network:

providernetwork

❒  A,B,C are provider networks❒  X,W,Y are customers (of provider networks)❒  X is dual-homed: attached to two networks

❍  Policy to enforce: X does not want to route from B to C via X ❍  ... so X will not advertise to B a route to C

==

=

€€

€

€ Customer-provider relationship

Peering relationship

Suppose an ISP only wants to route traffic to/from its customer networks (does not want to carry transit traffic between other ISPs)

36


BGP routing policy (2)

❒  A advertises path AW to B and to C❒  B advertises path BAW to X❒  B chooses not to advertise path BAW to C:

❍  B gets no “revenue” for routing CBAW since none of W, C, A are B’s customers

❍  C does not learn about CBAW❍  C will route CAW (not using B) to get to W

A

B

C

WX

Y

==

=

€€

€

€

legend:

customer network:

providernetwork


BGP Policies (Valley-Free routing)

Advertisements received ( ) only forwarded along links, to avoid costly transit traffic

Customer

Provider

Local ASPeer

Advertisement from provider:

Customer

Provider

Peer

XXX

Customer

Provider

Local ASPeer

Advertisement from peer:

Customer

Provider

Peer

XX

X

Customer

Provider

Local ASPeer

Advertisement from customer:

Customer

Provider

Peer

37


Why different Intra- and Inter-AS routing? Policy: ❒  Inter-AS: admin wants control over how its traffic is

routed, who routes through its net❒  Intra-AS: single admin, so no policy decisions neededScale:❒  hierarchical routing saves table size, reduced update

trafficPerformance: ❒  Intra-AS: can focus on performance❒  Inter-AS: policy may dominate over performance


Chapter 5: Network Layer Control Plane❒  ICMP❒  Routing protocols:

❍  Intra-domain routing•  LS - Link State (OSPF)

–  Send local knowledge to all–  Selective flooding–  Dijkstra

•  DV - Distance Vector (RIP)–  Send global knowledge to

neighbors–  Bellman-Ford–  Count to ∞ problem–  Poisoned reverse

❍  Inter-domain routing (BGP)•  Path Vector Protocol•  Policies

❒  For further study:o  Centralized control (SDN)o  Traffic engineeringo  Network management

# Next stop: link layer!

Introduction to Computer Networking Guy Leduc Chapter 5 ... · 5 Network Layer Control Plane 5-9 Chapter 5: outline 5.1 Introduction 5.2 ICMP: The Internet Control Message Protocol

Documents