cs 4284 systems capstone
DESCRIPTION
CS 4284 Systems Capstone. Networking. Godmar Back. Summary. Basics of Network Layer Routing (path selection) vs Forwarding (switching) Service models Datagram Networks vs VC Networks Basics of routers Packet classification, Packet Scheduling, AQM Next: Routing Algorithms. - PowerPoint PPT PresentationTRANSCRIPT
CS 4284Systems Capstone
Godmar Back
Networking
CS 4284 Spring 2013
Summary
• Basics of Network Layer– Routing (path selection) vs Forwarding
(switching)– Service models– Datagram Networks vs VC Networks– Basics of routers
• Packet classification, Packet Scheduling, AQM• Next: Routing Algorithms
CS 4284 Spring 2013
The Network Layer• transports segment from
sending to receiving host • on sending side
encapsulates segments into datagrams
• on receiving side, delivers segments to transport layer
• network layer protocols in every host, router
• router examines header fields in all IP datagrams passing through it
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
application
transportnetworkdata linkphysical
application
transportnetworkdata linkphysical
CS 4284 Spring 2013
Key Network-Layer Functions
• forwarding: move packets from router’s input to appropriate router output
• routing: determine route taken by packets from source to dest.
– Routing algorithms
analogy:
• routing: process of planning trip from source to dest
• forwarding: process of getting through single interchange
CS 4284 Spring 2013
1
23
0111
value in arrivingpacket’s header
routing algorithm
local forwarding tableheader value output link
0100010101111001
3221
Interplay between Routing and Forwarding
Routing Algorithms
CS 4284 Spring 2013
u
yx
wv
z2
21 3
1
12
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
CS 4284 Spring 2013
Graph Abstraction: Costs
u
yx
wv
z2
21 3
1
12
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 inversely 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 ?
CS 4284 Spring 2013
Routing Algorithm ClassificationGlobal 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
Static or dynamic?Static: • routes change slowly over
timeDynamic: • routes change more quickly
– periodic update– in response to link cost
changes
Note: Global/decentralized classification does not say where routing computation is performed, it says what information one has when computing routes
CS 4284 Spring 2013
Link State Routing• Routers periodically broadcast link state
information– link state messages or LSA (link state advertisement)– broadcast via sequence-number controlled flooding
• All routers acquire complete connectivity + link cost information of entire network
• Each router computes routing table from that obtained topology– Dijkstra’s shortest path to other routers, from
computing router this becomes forwarding table
CS 4284 Spring 2013
Link State Broadcast
• Goal: every node obtains a copy of this table– and updates as conditions change
u
yx
wv
z2
21 3
1
12
53
5u v w x y zu 0 2 5 1
v 2 0 3 2
w 5 3 0 3 1 5x 1 2 3 0 1
y 1 1 0 2z 5 2 0
CS 4284 Spring 2013
Link State Broadcast (2)• Idea:
– Send n’th row of own adjacency matrix – (conceptually, at least)
– With sender id n– Sequence number k
• On receipt of msg from node n, node m updates n’th row with msg content– update adj[m][n] to note connection
to neighbor with link cost• Controlled flooding, based on
sequence number• First round, everybody sends
u
yx
wv
z2
21 3
1
12
53
5
*,0
from node_idseq number
CS 4284 Spring 2013
Link State Broadcast (3)• u’s view after first broadcast
u
yx
wv
z2
21 3
1
12
53
5
u v w x y z
u 0 2 5 1
v 0
w 0
x 0
y 0
z 0
CS 4284 Spring 2013
Link State Broadcast (4)• v’s view after first broadcast
u
yx
wv
z2
21 3
1
12
53
5
u v w x y z
u 0
v 2 0 3 2
w 0
x 0
y 0
z 0
CS 4284 Spring 2013
Link State Broadcast (5)• Msgs exchanged in
second round
u
yx
wv
z2
21 3
1
12
53
5
u,1 0 2 5 1
v,1 2 0 3 2
w,1 5 3 0 3 1 5
x,1 1 2 3 0 1
y,1 1 1 0 2
z,1 5 2 0
CS 4284 Spring 2013
Link State Broadcast (6)
• everybody’s view after 2nd broadcast– Link changes will cause
new broadcasts
u
yx
wv
z2
21 3
1
12
53
5
u v w x y z
u 0 2 5 1
v 2 0 3 2
w 5 3 0 3 1 5
x 1 2 3 0 1
y 1 1 0 2
z 5 2 0
CS 4284 Spring 2013
Dijkstra’s algorithmNotation:• 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 dest. v
• p(v): predecessor node along path from source to v
• N': set of nodes whose least cost path definitively known
• Idea:– Keep track of nodes to whom
least cost path is known– Keep track of complement:
“candidate set” – for which you have estimates of cost
• Start with one node– Create estimates for adjacent
nodes• Repeat
– Choose best candidate, update estimates for other candidates based on candidate’s neighbors
– Remove candidate from candidate set and add to known set
• Until done
CS 4284 Spring 2013
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'
CS 4284 Spring 2013
2, u 4, x 2, x 2, u 3, y 4, y 3, y 4, y 4, y
2, u 5, u 1, u
Dijkstra’s Algorithm Example
u
yx
wv
z2
21 3
1
12
53
5
Step N’ D(v), p(v) D(w), p(w) D(x), p(x) D(y), p(y) D(z), p(z)v w x y z0 u1 u2 ux3 uxy
5 uxyvw4 uxyv
Done!
CS 4284 Spring 2013
Implementing Link State Routing• “HELLO” neighbor protocol
– Get to know your neighbors– Learn if neighbor moves out (link goes down)
• LSA broadcast protocol– Let the entire world know who your neighbors are
• As soon as you learn who yours are• Whenever your knowledge changes
– Help others broadcast information about their neighbors to entire world (until everybody learns entire topology)
– Maintain sequence numbers to prevent circular forwarding
• The two can be combined (design choice)
CS 4284 Spring 2013
Dijkstra’s algorithm, discussionAlgorithm complexity: n nodes, assume |E| << n2
• 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:• If, for instance, link cost = amount of carried traffic
AD
CB
1 1+e
e0
e1 1
0 0A
DC
B2+e 0
001+e1
AD
CB
0 2+e
1+e10 0
AD
CB
2+e 0
001+e1
initially … recomputerouting
… recompute … recompute
CS 4284 Spring 2013
Distance Vector Algorithm (1)
Bellman-Ford Equation (dynamic programming)Definedx(y) := cost of least-cost path from x to y
Then
dx(y) = min {c(x,v) + dv(y) }
where min is taken over all neighbors of x
CS 4284 Spring 2013
Bellman-Ford Example (2)
u
yx
wv
z2
21 3
1
12
53
5Clearly, 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 that achieves minimum is nexthop in shortest path forwarding table➜
B-F equation says:
CS 4284 Spring 2013
Distance Vector Algorithm (3)• Dx(y) = estimate of least cost from x to y• Distance vector: Dx = [Dx(y): y N ]• Node x knows cost to each neighbor v:
c(x,v)• Node x maintains Dx = [Dx(y): y N ]• Node x also maintains its neighbors’
distance vectors (at least temporarily)– For each neighbor v, x maintains Dv = [Dv(y): y N ]
CS 4284 Spring 2013
Distance Vector Algorithm (4)Basic Idea: • Each node periodically sends its own distance
vector estimate to neighbors• When a node x receives new DV estimate from
neighbor, it updates its own DV using B-F equation:
Dx(y) ← minv{c(x,v) + Dv(y)} for each node y ∊ N
• Under minor, natural conditions, the estimate Dx(y) converges to actual least cost dx(y)
CS 4284 Spring 2013
Distance Vector Algorithm (5)Iterative,
asynchronous: each local iteration caused by:
• local link cost change • DV update message
from neighborDistributed:• each node notifies
neighbors only when its DV changes– neighbors then notify
their neighbors if necessary
wait for (change in local link cost of msg from neighbor)
recompute estimates
if DV to any dest has changed, notify neighbors
Each node:
CS 4284 Spring 2013
Example DV Routing
- 1 2 5
1 - 4
2 4 - 1
5 1 -
12
5 4
1
0
1
2
3
CS 4284 Spring 2013
Example (cont’d)
- 1 2 5
1 - 4
2 4 - 1
5 1 -
1
2
5
4
1
- 1 2 3- 1 2 3
1 - 3 41 - 3 5
2 3 - 12 3 - 1
3 5 1 -
time
3 4 1 -
Dx(y) ← minv{c(x,v) + Dv(y)} for each node y ∊ N
CS 4284 Spring 2013
Example (cont’d)• Converged after
3 iterations
• What if link cost changes?
0 1 2 30 - 1 2 31 1 - 3 42 2 3 - 13 3 4 1 -
CS 4284 Spring 2013
Distance Vector: Link Cost ChangesLink 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, and informs its neighbors.
At time t1, z receives the update from y and updates its table. It computes a new least cost to x 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.
CS 4284 Spring 2013
Count-To-Infinity
5051-
1-51
-15
60
501
x z14
50
y60
x
y
z
- 4 5
4 - 1
5 1 -
Before change:
5051-
1-8
-150
5051-
1-6
-150
5051-
1-51
-17
y to x via z
z to x via y
y to x via z
51
51 50
50
CS 4284 Spring 2013
Distance Vector: Link Cost ChangesLink cost changes:• good news travels fast • bad news travels slow - “count to infinity”
problem!• 44 iterations before algorithm stabilizesPoisoned reverse: • 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?
x z14
50
y60
CS 4284 Spring 2013
Poisoned Reverse
- 51 50
51 - 1
5 1 -
60
501
- 51 50
6 - 1
50 1 -
x z14
50
y60
- 51 50
51 - 1
5 1 -
After change:
5051-
1-51
-150
y to x via z
z to x via x
- 1
Don’t tell z dist to x
x
y
z
5051-
1-51
-150
CS 4284 Spring 2013
Comparison of LS and DV algorithmsMessage complexity• LS: with n nodes, E links,
O(nE) msgs sent • DV: exchange between
neighbors only– convergence time varies
Speed of Convergence• LS: O(n2) algorithm requires
O(nE) msgs– may have oscillations
• DV: convergence time varies– may be routing loops– count-to-infinity problem
Robustness: what happens if router malfunctions?
LS: – node can advertise
incorrect link cost– each node computes only
its own tableDV:
– DV node can advertise incorrect path cost
– each node’s table used by others
• error propagates thru network
CS 4284 Spring 2013
Hierarchical Routing
scale: with 200 million destinations:
• can’t store all dest’s 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 true in practice
CS 4284 Spring 2013
Hierarchical Routing
• aggregate routers into regions, “autonomous systems” (AS)
• routers in same AS run same routing protocol– “intra-AS” routing
protocol– routers in different
AS may run different intra-AS routing protocols
Gateway router• Direct link to router
in another AS
• Example: – [AS1312]
CS 4284 Spring 2013
3b
1d
3a1c
2aAS3
AS1AS21a
2c2b
1b
Intra-ASRouting algorithm
Inter-ASRouting algorithm
Forwardingtable
3c
Interconnected ASes
• Forwarding table is configured by both intra- and inter-AS routing algorithm– Intra-AS sets entries for
internal dests– Inter-AS & Intra-AS sets
entries for external dests
CS 4284 Spring 2013
3b
1d
3a1c
2aAS3
AS1AS21a
2c2b
1b
3c
Inter-AS tasks• Suppose router in AS1
receives datagram for which dest is outside of AS1– Router should forward
packet towards one of the gateway routers, but which one?
AS1 needs:1. to learn which dests are
reachable through AS2 and which through AS3
2. to propagate this reachability info to all routers in AS1
Job of inter-AS routing!
CS 4284 Spring 2013
3b
1d
3a1c
2aAS3
AS1AS21a
2c2b
1b
3c
Inter-AS tasks (cont’d)• Suppose router in AS1
receives datagram for which dest is outside of AS1– Router should forward
packet towards one of the gateway routers, but which one?
Scenario 1Destination x is reachable
through single ASInter-AS must propagate this
information inside AS
AS3 advertises route to x
AS2 does not advertise route to x
Spread in AS1:use gateway 1c
to get to x
Least cost path to 1c is interface l, add (x,
l) to routing table
CS 4284 Spring 2013
3b
1d
3a1c
2aAS3
AS1AS21a
2c2b
1b
3c
Hot Potato Routing• Suppose router in AS1
receives datagram for which dest is outside of AS1– Router should forward
packet towards one of the gateway routers, but which one?
Scenario 2:Destination is reachable through
multiple AS:Hot potato routing: send packet
towards closest of two routers
CS 4284 Spring 2013
3b
1d
3a1c
2aAS3
AS1AS21a
2c2b
1b
3c
Hot Potato Routing• Suppose router in AS1
receives datagram for which dest is outside of AS1– Router should forward
packet towards one of the gateway routers, but which one?
Scenario 2:Destination is reachable through
multiple AS:Hot potato routing: send packet
towards closest of two routers
AS3 advertises route to x
AS2 advertises route to x
Spread in AS1:1c and 1b lead to
x
1b is closer than 1c, interface k leads to 1b, so add (x, k) to
routing table
CS 4284 Spring 2013
Summary
• Routing algorithms:– Link State– Distance Vector
• Hierarchical Routing– AS: autonomous systems
• Next: application to Internet/IP