Chapter 5

Network Layer:
The Control
Plane

Computer
Networking: A Top
Down Approach
7th Edition, Global Edition
Jim Kurose, Keith Ross
Pearson
April 2016
Network Layer: Control 5-1
Plane
Chapter 5: network layer
control plane
chapter goals: understand principles
behind network control plane
 traditional routing algorithms
 SDN controlllers
 Internet Control Message Protocol
 network management

and their instantiation, implementation in

the Internet:
 OSPF, BGP, OpenFlow, ODL and ONOS
controllers, ICMP, SNMP
Network Layer: Control 5-2
Plane
Chapter 5: outline
5.5 The
5.1 introduction
SDN control plane
5.2 ICMP:
5.6 routingThe
protocols
Internet Control Message Protocol
5.7link
Network
state management and SNMP
 distance vector
5.3 intra-AS routing in the Internet: OSPF
5.4 routing among the ISPs: BGP

Network Layer: Control 5-3

Plane
 Network-layer functions
Recall: two network-layer
functions:
forwarding: move
packets from router’s data plane
input to appropriate
router output
 routing: determine
route taken by packets control plane
from source to
destination
wo approaches to structuring network control plane:
per-router control (traditional)
logically centralized control (software defined networking)

Network Layer: Control 5-4

Plane
Per-router control plane
Individual routing algorithm components in each and every
router interact with each other in control plane to compute
forwarding tables

Routing
Algorithm
control
plane

data
plane

Network Layer: Control 5-5

Plane
Logically centralized control plane
A distinct (typically remote) controller interacts with local
control agents (CAs) in routers to compute forwarding tables

Remote Controller

control
plane

data
plane

CA
CA CA CA CA

Network Layer: Control 5-6

Plane
Chapter 5: outline
5.5 The
5.1 introduction
SDN control plane
5.2 ICMP:
5.6 routingThe
protocols
Internet Control Message Protocol
5.7link state management and SNMP
Network
 distance vector
5.3 intra-AS routing in the Internet: OSPF
5.4 routing among the ISPs: BGP

Network Layer: Control 5-7

Plane
Routing protocols

Routing 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! Network Layer: Control 5-8
Plane
Graph abstraction of the
network
5
3
v w 5
2
u 2 1 z
3
1 2
x 1
y
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) }

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

Network Layer: Control 5-9

Plane
 Graph abstraction: costs
5
c(x,x’) = cost of link (x,x’)
3 e.g., c(w,z) = 5
v w 5
2
u cost could always be 1, or
2
3
1 z inversely related to bandwidth,
1 2 or inversely related to
x 1
y
congestion

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

key question: what is the least-cost path between u and z ?

outing algorithm: algorithm that finds that least cost path

Network Layer: Control 5-10

Plane
 Routing algorithm
classification
Q: global or decentralized Q: static or dynamic?
information?
static:
global:
 routes change slowly
 all routers have complete
over time
topology, link cost info
 “link state” algorithms dynamic:
 routes change more
decentralized: quickly
 router knows physically-
• periodic update
connected neighbors, link
costs to neighbors • in response to link
 iterative process of cost changes
computation, exchange of
info with neighbors
 “distance vector”
algorithms
Network Layer: Control 5-11
Plane
Chapter 5: outline
5.5 The
5.1 introduction
SDN control plane
5.2 ICMP:
5.6 routingThe
protocols
Internet Control Message Protocol
5.7link state management and SNMP
Network
 distance vector
5.3 intra-AS routing in the Internet: OSPF
5.4 routing among the ISPs: BGP

Network Layer: Control 5-12

Plane
 A link-state routing algorithm
Dijkstra’s algorithm notation:
 net topology, link  c(x,y): link cost from node
costs known to all x to y; = ∞ if not direct
nodes neighbors
• accomplished via “link  D(v): current value of cost
state broadcast” of path from source to
• all nodes have same dest. v
info  p(v): predecessor node
 computes least cost along path from source to
paths from one node v
(‘source”) to all other  N': set of nodes whose
nodes least cost path definitively
• gives forwarding table known
for that node
 iterative: after k
iterations, know least Network Layer: Control 5-13
Plane
Dijsktra’s algorithm
1 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'

Network Layer: Control 5-14

Plane
Dijkstra’s algorithm: example
D(v) D(w) D(x) D(y) D(z)
Step N' p(v) p(w) p(x) p(y) p(z)
0 u 7,u 3,u 5,u ∞ ∞
1 uw 6,w 5,u 11,w ∞
2 uwx 6,w 11,w 14,x
3 uwxv 10,v 14,x
4 uwxvy 12,y
5 uwxvyz x
9

notes: 5
4
7
 construct shortest path
8
tree by tracing
predecessor nodes u
3 w y z
 ties can exist (can be 2
broken arbitrarily) 3
7 4
v
Network Layer: Control 5-15
Plane
 Dijkstra’s algorithm: another
example
Step N' D(v),p(v) D(w),p(w) D(x),p(x) D(y),p(y) D(z),p(z)
0 u 2,u 5,u 1,u ∞ ∞
1 ux 2,u 4,x 2,x ∞
2 uxy 2,u 3,y 4,y
3 uxyv 3,y 4,y
4 uxyvw 4,y
5 uxyvwz
5
3
v w 5
2
u 2 1 z
3
1 2
x 1
y

* Check out the online interactive exercises for more

examples: http://gaia.cs.umass.edu/kurose_ross/interactive/ Network Layer: Control 5-16
Plane
 Dijkstra’s algorithm: example
(2)
resulting shortest-path tree from u:

v w
u z
x y

resulting forwarding table in u:

destination link
v (u,v)
x (u,x)
y (u,x)
w (u,x)
z (u,x)
Network Layer: Control 5-17
Plane
 Dijkstra’s algorithm,
discussion
algorithm 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:
 e.g., support link cost equals amount of
A 1+e
1 carried A A 2+e A
traffic:
2+e 0 0 2+e 0
D 0 0 B D 1+e 1 B D B D 1+e 1 B
0 0
0 e 0 0
1
C C 0 1
C 1+e C 0
1
e
given these costs, given these costs, given these costs,
initially find new routing…. find new routing….find new routing….
resulting in new costs resulting in new cost
resulting in new costs
Network Layer: Control 5-18
Chapter 5: outline
5.5 The
5.1 introduction
SDN control plane
5.2 ICMP:
5.6 routingThe
protocols
Internet Control Message Protocol
5.7link state management and SNMP
Network
 distance vector
5.3 intra-AS routing in the Internet: OSPF
5.4 routing among the ISPs: BGP

Network Layer: Control 5-19

Plane
Distance vector algorithm
Bellman-Ford equation (dynamic
programming)

let
dx(y) := cost of least-cost path from x to
y
then v

dx(y) = min {c(x,v) +neighbor

cost from dv(y) }v to destination
cost to neighbor v

min taken over all neighbors v of x

Network Layer: Control 5-20
Plane
 Bellman-Ford example
5
3
clearly, dv(z) = 5, dx(z) = 3, dw(z) = 3
v w 5
2
u 2 1 z B-F equation says:
3
1 2 du(z) = min { c(u,v) + dv(z),
x 1
y
c(u,x) + dx(z),
c(u,w) + dw(z) }
= min {2 + 5,
1 + 3,
5 + 3} = 4
ode achieving minimum is next hop in shortest path
sed in forwarding table
Network Layer: Control 5-21
Plane
Distance vector algorithm
 Dx(y) = estimate of least cost from x to
y
• x maintains distance vector Dx = [Dx(y): y є
N]
 node x:
• knows cost to each neighbor v: c(x,v)
• maintains its neighbors’ distance
vectors. For each neighbor v, x
maintains
Dv = [Dv(y): y є N ]

Network Layer: Control 5-22

Plane
Distance vector algorithm
key idea:
 from time-to-time, each node sends its
own distance vector estimate to
neighbors
 when x receives new DV estimate from
neighbor, it updates its own DV using B-
(y) ← minv{c(x,v) + Dv(y)} for each node y ∊ N
FDequation:
x

 under minor, natural conditions, the

estimate Dx(y) converge to the actual
least cost dx(y)

Network Layer: Control 5-23

Plane
Distance vector algorithm
iterative, each node:
asynchronous:
each local iteration
caused by: wait for (change in local link
cost or msg from neighbor)
 local link cost change
 DV update message
from neighbor recompute estimates
distributed:
 each node notifies
if DV to any dest has
neighbors only when
its DV changes changed, notify neighbors
• neighbors then notify
their neighbors if
necessary

Network Layer: Control 5-24

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

from
from

y ∞∞ ∞ y 2 0 1
z ∞∞ ∞ z 7 1 0

node y cost to
table x y z y
2 1
x ∞ ∞ ∞
x z
from

y 2 0 1 7
z ∞∞ ∞

node z cost to
table x y z
x ∞∞ ∞
from

y ∞∞ ∞
z 7 1 0
time
Network Layer: Control 5-25
Plane
 Dx(z) = min{c(x,y) +
Dx(y) = min{c(x,y) + Dy(y), c(x,z) + Dz(y)}
= min{2+0 , 7+1} = 2 Dy(z), c(x,z) + Dz(z)}
= min{2+1 , 7+0} = 3
node x cost to cost to cost to
table x y z x y z x y z
x 0 2 7 x 0 2 3 x 0 2 3

from
from

y ∞∞ ∞ y 2 0 1 y 2 0 1

from
z ∞∞ ∞ z 7 1 0 z 3 1 0
node y cost to cost to cost to
table x y z x y z x y z y
2 1
x ∞ ∞ ∞ x 0 2 7 x 0 2 3 x z
from

y 2 0 1 y 2 0 1 7
from

y 2 0 1

from
z ∞∞ ∞ z 7 1 0 z 3 1 0

node z cost to cost to cost to

table x y z x y z x y z

x ∞∞ ∞ x 0 2 7 x 0 2 3
from

from

y 2 0 1 y 2 0 1
from

y ∞∞ ∞
z 7 1 0 z 3 1 0 z 3 1 0
time
Network Layer: Control 5-26
Plane
 Distance vector: link cost
changes
link cost changes: 1
 node detects local link cost y
4 1
change
 updates routing info,
x z
50
recalculates
distance vector
 if DV changes, notify
“good t0 : y detects link-cost change, updates its DV, informs its
neighbors
news neighbors.
travels t1 : z receives update from y, updates its table, computes new
fast” least cost to x , sends its neighbors its DV.

t2 : y receives z’s update, updates its distance table. y’s least costs
do not change, so y does not send a message to z.

* Check out the online interactive exercises for more

examples: http://gaia.cs.umass.edu/kurose_ross/interactive/ Network Layer: Control 5-27
Plane
Distance vector: link cost
changes
link cost changes: 60
 node detects local link cost y
4 1
change
 bad news travels slow -
x z
50
“count to infinity” problem!
 44 iterations before
algorithm stabilizes: see
poisoned reverse:
text (routing loop)
 If Z routes through Y to get to X :
 Z tells Y its (Z’s) distance to X is infinite (so Y
won’t route to X via Z)
 will this completely solve count to infinity
problem?

Network Layer: Control 5-28

Plane
 Comparison of LS and DV
algorithms
message complexity robustness: what
 LS: with n nodes, E links, happens if router
O(nE) msgs sent malfunctions?
 DV: exchange between LS:
neighbors only
• node can advertise
• convergence time varies incorrect link cost
speed of convergence • each node computes
only its own table
 LS: O(n2) algorithm requires
O(nE) msgs DV:
• may have oscillations • DV node can advertise
 DV: convergence time incorrect path cost
varies • each node’s table used
• may be routing loops by others
• error propagate thru
• count-to-infinity problem
network

Network Layer: Control 5-29

Plane
Chapter 5: outline
5.5 The
5.1 introduction
SDN control plane
5.2 ICMP:
5.6 routingThe
protocols
Internet Control Message Protocol
5.7link
Network
state management and SNMP
 distance vector
5.3 intra-AS routing in the Internet: OSPF
5.4 routing among the ISPs: BGP

Network Layer: Control 5-30

Plane
Making routing
scalable
our routing study thus far -
idealized
 all routers identical
 network “flat”
… not true in practice
scale: with billions administrative
of destinations: autonomy
 can’t store all  internet = network of
destinations in networks
routing tables!  each network admin
 routing table may want to control
exchange would routing in its own
swamp links! network
Network Layer: Control 5-31
Plane
Internet approach to scalable
routing
aggregate routers into regions known as
“autonomous systems” (AS) (a.k.a.
“domains”)
intra-AS routing inter-AS routing
 routing among hosts,  routing among AS’es
routers in same AS  gateways perform
(“network”) inter-domain routing
 all routers in AS must run (as well as intra-
same intra-domain domain routing)
protocol
 routers in different AS
can run different intra-
domain routing protocol
 gateway router: at
“edge” of its own AS, has Network Layer: Control 5-32
Plane
Interconnected ASes

3c
3a 2c
3b 2a
AS3 2b
1c AS2
1a 1b AS1
1d  forwarding table
configured by both
intra- and inter-AS
Intra-AS Inter-AS routing algorithm
Routing
algorithm
Routing
algorithm
• intra-AS routing
determine entries
Forwarding
table
for destinations
within AS
• inter-AS & intra-AS
determine entries
for external
Network Layer: Control 5-33
Plane
 Inter-AS tasks
 suppose router in AS1 must:
AS1 receives 1. learn which dests
datagram destined are reachable
outside of AS1: through AS2, which
• router should through AS3
forward packet to 2. propagate this
gateway router, reachability info to
but which one? all routers in AS1
job of inter-AS routing!
3c
3a
3b
AS3 2c other
1c 2a networks
other 1a 2b
networks 1b AS2
AS1 1d

Network Layer: Control 5-34

Plane
Intra-AS Routing
 also known as interior gateway
protocols (IGP)
 most common intra-AS routing
protocols:
• RIP: Routing Information Protocol
• OSPF: Open Shortest Path First (IS-IS
protocol essentially same as OSPF)
• IGRP: Interior Gateway Routing
Protocol (Cisco proprietary for decades,
until 2016)

Network Layer: Control 5-35

Plane
OSPF (Open Shortest Path
First)
 “open”: publicly available
 uses link-state algorithm
• link state packet dissemination
• topology map at each node
• route computation using Dijkstra’s algorithm
 router floods OSPF link-state
advertisements (LSA) to all other routers
in entire AS
• carried in OSPF messages directly over IP
(rather than TCP or UDP) (protocol # 89)
• link state: for each attached link
 IS-IS routing protocol: nearly identical to
OSPF
Network Layer: Control 5-36
Plane
OSPF “advanced” features
 security: all OSPF messages authenticated
(to prevent malicious intrusion)
 multiple same-cost paths allowed (only
one path in RIP)
 for each link, multiple cost metrics for
different ToS (e.g., satellite link cost set
low for best effort ToS; high for real-time
ToS)
 integrated uni- and multi-cast support:
• Multicast OSPF (MOSPF) uses same
topology data base as OSPF - 224.0.0.5/
224.0.0.6
 hierarchical OSPF in large domains.
Network Layer: Control 5-37
Plane
Hierarchical
OSPF
boundary router (ASBR)
backbone router

backbone
area border area 0
routers (ABR)

designated
router (DR) area 3
DR
DR

internal
routers
area 1
area 2

Network Layer: Control 5-38

Plane
Hierarchical
OSPF
 two-level hierarchy: local area, backbone.
• link-state advertisements only in area
• each nodes has detailed area topology;
only know direction (shortest path) to nets
in other areas.
 area border routers: “summarize” distances
to nets in own area, advertise to other ABRs.
 backbone routers: run OSPF routing limited
to backbone.
 boundary routers: connect to other AS’es.

Network Layer: Control 5-39

Plane
Chapter 5: outline
5.5 The
5.1 introduction
SDN control plane
5.2 ICMP:
5.6 routingThe
protocols
Internet Control Message Protocol
5.7link
Network
state management and SNMP
 distance vector
5.3 intra-AS routing in the Internet: OSPF
5.4 routing among the ISPs: BGP

Network Layer: Control 5-40

Plane
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:
• eBGP: obtain subnet reachability
information from neighboring ASes
• iBGP: propagate reachability information to
all AS-internal routers.
• 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”
Network Layer: Control 5-41
Plane
eBGP, iBGP connections

2b

2a 2c
∂
1b 3b
2d
1a 1c ∂
3a 3c
AS 2
1d 3d

AS 1 eBGP connectivity AS 3
iBGP connectivity

1c gateway routers run both eBGP and iBGP protocols

Network Layer: Control 5-42

Plane
BGP basics
 BGP session: two BGP routers (“peers”) exchange
BGP messages over semi-permanent TCP
connection:
• advertising paths to different destination network
prefixes
 when (BGP is arouter
AS3 gateway “path3a
vector” protocol)
advertises path AS3,X
to AS2 gateway router 2c:
• AS3 promises to AS2 it will forward datagrams
towards X
AS 3 3b
AS 1 1b
3a 3c
1a 1c
AS 2 2b 3d X
1d
BGP advertisement:
2a 2c AS3, X

2d
Network Layer: Control 5-43
Plane
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
• NEXT-HOP: indicates specific internal-AS router
to next-hop AS
 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

Network Layer: Control 5-44

Plane
BGP path advertisement
AS3 3b
AS1 1b
3a 3c
1a 1c
AS2 2b 3d X
1d AS3,X
AS2,AS3,X
2a 2c

2d

 AS2 router 2c receives path advertisement AS3,X (via

eBGP) from AS3 router 3a
 Based on AS2 policy, AS2 router 2c accepts path AS3,X,
propagates (via iBGP) to all AS2 routers
 Based on AS2 policy, AS2 router 2a advertises (via
eBGP) path AS2, AS3, X to AS1 router 1c
Network Layer: Control 5-45
Plane
BGP path advertisement
AS3 3b
AS1 1b AS3,X
3a 3c
1a 1c
AS2 2b 3d X
1d AS3,X
AS2,AS3,X
2a 2c

2d

gateway router may learn about multiple paths to

destination:
 AS1 gateway router 1c learns path AS2,AS3,X from 2a
 AS1 gateway router 1c learns path AS3,X from 3a
 Based on policy, AS1 gateway router 1c chooses path
AS3,X, and advertises path within AS1 via iBGP
Network Layer: Control 5-46
Plane
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

Network Layer: Control 5-47

Plane
 BGP, OSPF, forwarding table
entries
Q: how does router set forwarding table entry to distant prefix?

AS3 3b
AS1 1b AS3,X
1 AS3,X
3a 3c
1a 2 1c
local link AS2 2b 3d X
interfaces 2 1d 1 AS3,X
at 1a, 1d AS2,AS3,X
2a 2c
physical link
2d

dest interface  recall: 1a, 1b, 1c learn about dest X

… … via iBGP from 1c: “path to X goes
X 1  through
1d: OSPF1c”intra-domain routing: to get
… … to 1c, forward over outgoing local
interface 1

Network Layer: Control 5-48

Plane
BGP, OSPF, forwarding table
entries
Q: how does router set forwarding table entry to distant prefix?

AS3 3b
AS1 1b
1
3a 3c
1a 2 1c
AS2 2b 3d X
1d
2a 2c

2d

dest interface  recall: 1a, 1b, 1c learn about dest X

… … via iBGP from 1c: “path to X goes
X 2  through
1d: OSPF1c”intra-domain routing: to get
… … to 1c, forward over outgoing local
 interface 1
1a: OSPF intra-domain routing: to
get to 1c, forward over outgoing
local interface 2 Network Layer: Control 5-49
Plane
BGP route selection
 router may learn about more than one
route to destination AS, selects route
based on:
1. local preference value attribute: policy
decision
2. shortest AS-PATH
3. closest NEXT-HOP router: hot potato
routing
4. additional criteria

Network Layer: Control 5-50

Plane
Hot Potato Routing
AS3 3b
AS1 1b
3a 3c
1a 1c
AS2 2b 3d X
1d 112
AS3,X
152
AS1,AS3,X 2a 263 2c
201
OSPF link weights
2d

 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 (e.g., 2d chooses 2a, even
though more AS hops to X): don’t worry about
inter-domain cost!
Network Layer: Control 5-51
Plane
BGP: achieving policy via
advertisements
legend: provider
B network
X
W A
customer
C network:

Suppose an ISP only wants to route traffic to/from its customer

networks (does not want to carry transit traffic between other ISPs)
 A advertises path Aw to B and to C
 B chooses not to advertise BAw to C:
 B gets no “revenue” for routing CBAw, since none of C,
A, w are B’s customers
 C does not learn about CBAw path
 C will route CAw (not using B) to get to w
Network Layer: Control 5-52
Plane
BGP: achieving policy via
advertisements
legend: provider
B network
X
W A
customer
C network:

Suppose an ISP only wants to route traffic to/from its customer

networks (does not want to carry transit traffic between other ISPs)

 A,B,C are provider networks

 X,W,Y are customer (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 (or Y)
Network Layer: Control 5-53
Plane
Why different Intra-, Inter-AS
routing ?
policy:
 inter-AS: admin wants control over how its
traffic routed, who routes through its net.
 intra-AS: single admin, so no policy
decisions needed
scale:
 hierarchical routing saves table size,
reduced update traffic – important for
Inter-AS
performance:
 intra-AS: can focus on performance
 inter-AS: policy may dominate over
performance Network Layer: Control 5-54
Plane
Chapter 5: outline
5.5 The
5.1 introduction
SDN control plane
5.2 ICMP:
5.6 routingThe
protocols
Internet Control Message Protocol
5.7link
Network
state management and SNMP
 distance vector
5.3 intra-AS routing in the Internet: OSPF
5.4 routing among the ISPs: BGP

Network Layer: Control 5-55

Plane
Software defined networking (SDN)
 Internet network layer: historically has
been implemented via distributed, per-
router approach
• monolithic router contains switching hardware,
runs proprietary implementation of Internet
standard protocols (IP, RIP, IS-IS, OSPF, BGP) in
proprietary router OS (e.g., Cisco IOS)
• different “middleboxes” for different network
layer functions: firewalls, load balancers, NAT
boxes, ..

 ~2005: renewed interest in rethinking

network control plane
Network Layer: Control 5-56
Plane
Recall: per-router control plane
Individual routing algorithm components in each and every
router interact with each other in control plane to compute
forwarding tables

Routing
Algorithm
control
plane

data
plane

Network Layer: Control 5-57

Plane
Recall: logically centralized control plane
A distinct (typically remote) controller interacts with local
control agents (CAs) in routers to compute forwarding tables

Remote Controller

control
plane

data
plane

CA
CA CA CA CA

Network Layer: Control 5-58

Plane
Software defined networking (SDN)

Why a logically centralized control

plane?
 easier network management: avoid
router misconfigurations, greater
flexibility of traffic flows
 table-based forwarding (recall OpenFlow
API) allows “programming” routers
• centralized “programming” easier: compute
tables centrally and distribute
• distributed “programming”: more difficult:
compute tables as result of distributed
algorithm (protocol) implemented in each and
every router
 open (non-proprietary) implementation of
Network Layer: Control 5-59
Plane
 Analogy: mainframe to PC evolution *

Ap Ap Ap Ap Ap Ap Ap Ap Ap Ap
Specialized App
p p p p p p p p p p
Application
Open Interface
s
Specialized Windows Mac
Operating or Linux or OS
(OS)
System
Open Interface
Specialized
Hardware
Microprocessor

Vertically integrated Horizontal

Closed, proprietary Open interfaces
Slow innovation Rapid innovation
Small industry Huge industry
* Slide courtesy: N. McKeown Network Layer: Control 5-60
Traffic engineering: difficult
traditional routing
5
3
2 v w 5

u 2
3
1 z
1
2
x 1 y

Q: what if network operator wants u-to-z traffic to flow along

uvwz, x-to-z traffic to flow xwyz?
A: need to define link weights so traffic routing algorithm
computes routes accordingly (or need a new routing algorithm)!

Link weights are only control “knobs”: wrong!

Network Layer: Control 5-61
Traffic engineering:
difficult
5
3
2 v w 5

u 2
3
1 z
1
2
x 1 y

Q: what if network operator wants to split u-to-z

traffic along uvwz and uxyz (load balancing)?
A: can’t do it (or need a new routing algorithm)

Network Layer: Control 5-62

Plane
Traffic engineering: Networking 401

difficult
5
3
v
v
w
w
2 5

zz
u 2 1
3
1
2
xx yy
1

Q: what if w wants to route blue and red traffic differently?

A: can’t do it (with destination based forwarding, and LS,

DV routing)

Network Layer: Control 5-63

Plane
Software defined networking (SDN)
4. programmable 3. control plane
control routing
access
control … load
balance functions
applications external to data-
plane switches
Remote Controller

control
plane

data
plane

CA 2. control,
data plane
CA CA CA CA separation

1: generalized“ flow-
based” forwarding
(e.g., OpenFlow)
Network Layer: Control 5-64
Plane
SDN perspective: data plane switches
Data plane switches network-control
applications
 fast, simple, commodity …
routing
switches implementing
generalized data-plane access load
control balance
forwarding (Section 4.4) in
hardware control
plane
northbound API
 switch flow table
computed, installed by SDN Controller
controller (network operating system)
 API for table-based switch
control (e.g., OpenFlow) southbound API
• defines what is controllable
and what is not
data
 protocol for plane
communicating with
controller (e.g., OpenFlow) SDN-controlled switches
Network Layer: Control 5-65
Plane
SDN perspective: SDN controller
SDN controller (network network-control
applications
OS): routing
…
 maintain network state
access load
information control balance
 interacts with network
control
control applications northbound API plane
“above” via northbound
API SDN Controller
 interacts with network (network operating system)
switches “below” via
southbound API southbound API

 implemented as
distributed system for data
plane
performance, scalability,
fault-tolerance,
robustness
Network Layer: Control 5-66
SDN-controlled switches

Plane
SDN perspective: control applications
network-control apps: network-control
applications
 “brains” of control: …
routing
implement control
functions using lower- access load
control balance
level services, API
provided by SND control
plane
controller northbound API

 unbundled: can be
SDN Controller
provided by 3rd party: (network operating system)
distinct from routing
vendor, or SDN southbound API
controller
data
plane

SDN-controlled switches
Network Layer: Control 5-67
Plane
Components of SDN controller
routing access load
control balance
Interface layer
to network Interface, abstractions for network control apps
control apps:
abstractions API
network
graph
RESTful
API
… intent

Network-wide
state statistics … flow tables
management
Network-wide distributed, robust state management
SDN
layer: state of
controller
networks links,
switches,
Link-state info host info … switch info

services: a
communication
distributed OpenFlow … SNMP
layer:
database Communication to/from controlled devices
communicate
between SDN
controller and
controlled
switches
Network Layer: Control 5-68
Plane
OpenFlow protocol

OpenFlow Controller  operates between

controller, switch
 TCP used to exchange
messages
• optional encryption
 three classes of
OpenFlow messages:
• controller-to-switch
• asynchronous (switch to
controller)
• symmetric (misc)

Network Layer: Control 5-69

Plane
OpenFlow: controller-to-switch
messages
Key controller-to-switch
messages OpenFlow Controller
 features: controller
queries switch features,
switch replies
 configure: controller
queries/ sets switch
configuration parameters
 modify-state: add, delete,
modify flow entries in the
OpenFlow tables
 packet-out: controller can
send this packet out of Network Layer: Control 5-70
Plane
OpenFlow: switch-to-controller
messages
Key switch-to-controller
messages OpenFlow Controller
 packet-in: transfer packet
(and its control) to controller.
See packet-out message
from controller
 flow-removed: flow table
entry deleted at switch
 port status: inform controller
of a change on a port.
Fortunately, network operators don’t “program”
switches by creating/sending OpenFlow messages
directly. Instead use higher-level abstraction at
controller Network Layer: Control 5-71
Plane
SDN: control/data plane interaction
example
Dijkstra’s link-state 1 S1, experiencing link
Routing failure using OpenFlow
port status message to
4 5 notify controller
network
graph
RESTful
API
… intent 2 SDN controller receives
OpenFlow message,
statistics
3
… flow tables
updates link status info
3 Dijkstra’s routing
Link-state info host info … switch info algorithm application
2 has previously
OpenFlow … SNMP
registered to be called
when ever link status
changes. It is called.
4 Dijkstra’s routing
6 algorithm access
1
network graph info, link
s2 state info in controller,
computes new routes
s1
s4
s3
Network Layer: Control 5-72
Plane
SDN: control/data plane interaction
example
Dijkstra’s link-state
Routing

4 5
network
graph
RESTful
API
… intent 5 link state routing app
interacts with flow-table-
statistics
3
… flow tables
computation component
in SDN controller, which
Link-state info host info … switch info
computes new flow
tables needed
2 6 Controller uses
OpenFlow … SNMP
OpenFlow to install new
tables in switches that
need updating
6
1

s2
s1
s4
s3
Network Layer: Control 5-73
Plane
OpenDaylight (ODL)
controller…
Traffic
Engineering  ODL Lithium
controller
REST API  network apps
Network Basic Network Service Functions
may be
service apps contained
Access
topology switch stats
manager
within, or be
manager manager
Control external to SDN
forwarding host controller
manager manager
 Service
Service Abstraction Layer (SAL) Abstraction
Layer:
OpenFlow 1.0
… SNMP OVSDB
interconnects
internal,
external
applications and
services
Network Layer: Control 5-74
Plane
ONOS controller
Network …
control apps  control apps
northbound separate from
abstractions,
REST API Intent
protocols
controller
 intent framework:
hosts paths flow rules topology high-level
specification of
ONOS
devices links statistics distributed service: what
core rather than how
 considerable
device link host flow packet southbound emphasis on
abstractions,
OpenFlow Netconf OVSDB protocols
distributed core:
service reliability,
replication
performance
scaling
Network Layer: Control 5-75
Plane
SDN: selected challenges
 hardening the control plane:
dependable, reliable, performance-
scalable, secure distributed system
• robustness to failures: leverage strong
theory of reliable distributed system for
control plane
• dependability, security: “baked in” from day
one?
 networks, protocols meeting mission-
specific requirements
• e.g., real-time, ultra-reliable, ultra-secure
 Internet-scaling
Network Layer: Control 5-76
Plane
Chapter 5: outline
5.5 The
5.1 introduction
SDN control plane
5.2 ICMP:
5.6 routingThe
protocols
Internet Control Message Protocol
5.7link
Network
state management and SNMP
 distance vector
5.3 intra-AS routing in the Internet: OSPF
5.4 routing among the ISPs: BGP

Network Layer: Control 5-77

Plane
ICMP: internet control message
protocol
 used by hosts &
Type Code description
routers to
0 0 echo reply (ping)
communicate 3 0 dest. network unreachable
network-level 3 1 dest host unreachable
information 3 2 dest protocol unreachable
• error reporting: 3 3 dest port unreachable
unreachable host, 3 6 dest network unknown
network, port, 3 7 dest host unknown
protocol 4 0 source quench (congestion
• echo request/reply control - not used)
(used by ping) 8 0 echo request (ping)
 network-layer 9 0 route advertisement
“above” IP: 10 0 router discovery
11 0 TTL expired
• ICMP msgs carried in
12 0 bad IP header
IP datagrams
(protocol # 1)
 ICMP message: type, Network Layer: Control 5-78
Plane
Traceroute and ICMP
 source sends series of
 when ICMP message
UDP segments to
destination arrives, source records
• first set has TTL =1 RTTs
• second set has TTL=2, stopping criteria:
etc.  UDP segment
• unlikely port number eventually arrives at
 when datagram in nth destination host
set arrives to nth  destination returns
router: ICMP “port
• router discards unreachable”
datagram and sends
source ICMP message message (type 3,
(type 11, code 0) code 3)
• ICMP message include
3 probes 3 probes  source stops
name of router & IP
address3 probes
Network Layer: Control 5-79
Plane
Chapter 5: outline
5.5 The
5.1 introduction
SDN control plane
5.2 ICMP:
5.6 routingThe
protocols
Internet Control Message Protocol
5.7link
Network
state management and SNMP
 distance vector
5.3 intra-AS routing in the Internet: OSPF
5.4 routing among the ISPs: BGP

Network Layer: Control 5-80

Plane
What is network
management?
 autonomous systems (aka “network”): 1000s of
interacting hardware/software components
 other complex systems requiring monitoring,
control:
• jet airplane
• nuclear power plant
• others?
"Network management includes the deployment, integration
and coordination of the hardware, software, and human
elements to monitor, test, poll, configure, analyze, evaluate,
and control the network and element resources to meet the
real-time, operational performance, and Quality of Service
requirements at a reasonable cost."

Network Layer: Control 5-81

Plane
 Infrastructure for network
management
definitions:
managing entity
agent data
managing
data managed device
managed devices
entity
contain managed
agent data
objects whose data
network is gathered into a
management
protocol agent data
managed device Management
managed device
Information Base
(MIB)
agent data
agent data
managed device
managed device

Network Layer: Control 5-82

Plane
 SNMP protocol
Two ways to convey MIB info, commands:

managing managing
entity entity

request
trap msg
response

agent data agent data

managed device managed device

request/response mode trap mode

Network Layer: Control 5-83
Plane
SNMP protocol: message
types
Message type Function
GetRequest
GetNextRequest manager-to-agent: “get me data”
GetBulkRequest (data instance, next data in list, block of data)

InformRequest manager-to-manager: here’s MIB value

SetRequest manager-to-agent: set MIB value

Response Agent-to-manager: value, response to

Request

Trap Agent-to-manager: inform manager

of exceptional event

Network Layer: Control 5-84

Plane
SNMP protocol: message
formats
Get/set header Variables to get/set
PDU Error
Request Error
type Status Name Value Name Value ….
ID Index
(0-3) (0-5)

PDU Trap
type Enterprise Agent Type
Specific Time
Name Value ….
Addr code stamp
4 (0-7)
Trap header Trap info

SNMP PDU

More on network management: see earlier editions of text!

Network Layer: Control 5-85

Plane
Chapter 5: summary
we’ve learned a lot!
 approaches to network control plane
• per-router control (traditional)
• logically centralized control (software defined
networking)
 traditional routing algorithms
• implementation in Internet: OSPF, BGP
 SDN controllers
• implementation in practice: ODL, ONOS
 Internet Control Message Protocol
 network management
next stop: link layer!
Network Layer: Control 5-86
Plane