Computer Networks. Instructor: Niklas Carlsson

Computer Networks Instructor: Niklas Carlsson Email: niklas.carlsson@liu.se Notes derived Computer Networking: A Top Down Approach, b Jim Kurose and Keith Ross, Addison-Wesle. The slides are adapted and modified based on slides the book s companion Web site, as well as modified slides b Anirban Mahanti and Care Williamson.

Routing Algorithms Link State Distance Vector Hierarchical Routing

Interpla between routing and forwarding routing algorithm local forwarding table header value output link 000 00 0 00 Effectivel building the forwarding tables (and determining the routing path) value in arriving packet s header 0

Let s start with the case where we operate the full (sub) network c a b AS a AS c d b a c AS b 4

Graph abstraction Graph: G = (N,E) u v w N = set of routers = { u, v, w,,, } E = set of links ={ (u,v), (u,), (v,), (v,w), (,w), (,), (w,), (w,), (,) } Remark: Graph abstraction is useful in other network contets too Eample: PP, where N is set of peers and E is set of TCP connections

Graph abstraction: costs u v w c(, ) = cost of link (, ) - e.g., c(w,) = cost could alwas be, or inversel related to bandwidth, or inversel related to congestion Cost of path (,,,, p ) = c(, ) + c(, ) + + c( p-, p ) Question: What s the least-cost path between u and? Routing algorithm: find good paths source to destination router. 6

Routing Algorithm Classification. Global vs decentralied? Global: all routers have complete topolog, link cost info link state algorithms Decentralied: router knows about phsicallconnected neighbors Iterative, distributed computations distance vector algorithms. Static vs dnamic? Static: routes change slowl over time Dnamic: routes change more quickl periodic update in response to link cost changes. Load sensitivit? Man Internet routing algorithms are load insensitive 7

A Link-State Routing Algorithm Dijkstra s algorithm net topolog, link costs known to all nodes accomplished via link state broadcast all nodes have same info computes least cost paths one node ( source ) to all other nodes gives forwarding table for that node iterative: after k iterations, know least cost path to k dest. s Notation: c(,): link cost node to ; = if not direct neighbors D(v): current value of cost of path source to dest. v p(v): predecessor node along path source to v N': set of nodes whose least cost path definitivel known 9

Dijsktra s Algorithm Initialiation: N' = {u} for all nodes v 4 if v adjacent to u 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 0 add w to N' update D(v) for all v adjacent to w and not in N' : D(v) = min( D(v), D(w) + c(w,v) ) /* new v is either old v or known 4 shortest path w plus cost w to v */ until all nodes in N' 0

Dijkstra s algorithm: eample Step 0 4 N' u D(v),p(v),u D(w),p(w),u D(),p(),u D(),p() D(),p() u v w

Dijkstra s algorithm: eample Step 0 4 N' u u D(v),p(v),u D(w),p(w),u D(),p(),u D(),p() D(),p() u v w

Dijkstra s algorithm: eample Step 0 4 N' u u D(v),p(v),u,u D(w),p(w),u 4, D(),p(),u D(),p(), D(),p() u v w

Dijkstra s algorithm: eample Step 0 4 N' u u u D(v),p(v),u,u D(w),p(w),u 4, D(),p(),u D(),p(), D(),p() u v w 4

Dijkstra s algorithm: eample Step 0 4 N' u u u D(v),p(v),u,u,u D(w),p(w),u 4,, D(),p(),u D(),p(), D(),p() 4, u v w

Dijkstra s algorithm: eample Step 0 4 N' u u u uv D(v),p(v),u,u,u D(w),p(w),u 4,, D(),p(),u D(),p(), D(),p() 4, u v w 6

7 Dijkstra s algorithm: eample Step 0 4 N' u u u uv D(v),p(v),u,u,u D(w),p(w),u 4,,, D(),p(),u D(),p(), D(),p() 4, 4, u w v

8 Dijkstra s algorithm: eample Step 0 4 N' u u u uv uvw D(v),p(v),u,u,u D(w),p(w),u 4,,, D(),p(),u D(),p(), D(),p() 4, 4, u w v

9 Dijkstra s algorithm: eample Step 0 4 N' u u u uv uvw D(v),p(v),u,u,u D(w),p(w),u 4,,, D(),p(),u D(),p(), D(),p() 4, 4, 4, u w v

0 Dijkstra s algorithm: eample Step 0 4 N' u u u uv uvw uvw D(v),p(v),u,u,u D(w),p(w),u 4,,, D(),p(),u D(),p(), D(),p() 4, 4, 4, u w v

Dijkstra s algorithm, discussion Algorithm compleit: n nodes each iteration: need to check all nodes, w, not in N n(n+)/ comparisons: O(n ) more efficient implementations possible: O(nlogn) Oscillations possible: e.g., link cost = amount of carried traffic D A +e 0 0 0 e C e initiall B D A +e 0 0 +e C 0 B recompute routing D A 0 +e 0 0 +e C B recompute D A +e 0 0 +e C 0 recompute B

Distance Vector Algorithm () Bellman-Ford Equation (dnamic programming) Define d () := cost of least-cost path to

Distance Vector Algorithm () Bellman-Ford Equation (dnamic programming) Define d () := cost of least-cost path to Then d () = min {c(,v) + d v () } where min is taken over all neighbors of 4

Bellman-Ford eample () u v w Clearl, d v () =, d () =, d w () = B-F equation sas: d u () = min { c(u,v) + d v (), c(u,) + d (), c(u,w) + d w () } = min { +, +, + } = 4 Node that achieves minimum is net hop in shortest path forwarding table

Distance Vector Algorithm () D () = estimate of least cost to Distance vector: D = [D (): є N ] Node knows each neighbor v: c(,v) Node maintains D () Node also maintains its neighbors distance vectors (sent to b neighbors) For each neighbor v, maintains D v = [D v (): є N ] 6

Distance vector algorithm (4) Basic idea: Each node periodicall sends its own distance vector estimate to neighbors When node receives new DV estimate neighbor, it updates its own DV using B-F equation: D () min v {c(,v) + D v ()} for each node N Under some conditions, the estimate D () converge the actual least cost d () 7

Distance Vector Algorithm () Iterative, asnchronous: each local iteration caused b: local link cost change DV update message neighbor Distributed: each node notifies neighbors when its DV changes neighbors then notif their neighbors if necessar Each node: wait for (change in local link cost or msg neighbor) recompute estimates if DV to an dest has changed, notif neighbors 9

node table 0 7 node table 0 node table 7 0 time 7 0

node table 0 7 node table 0 node table 7 0 0 7 0 7 time 7

node table 0 7 node table 0 node table 7 0 0 7 0 0 7 7 0 0 7 0 time 7

D () = min{c(,) + D (), c(,) + D ()} = min{+0, 7+} = node table 0 7 0 7 0 node table 0 node table 7 0 0 7 7 0 0 7 0 time D () = min{c(,) + D (), c(,) + D ()} = min{+, 7+0} = 7 4

D () = min{c(,) + D (), c(,) + D ()} = min{+0, 7+} = node table 0 7 0 0 7 0 node table 0 node table 7 0 0 7 7 0 0 7 0 time D () = min{c(,) + D (), c(,) + D ()} = min{+, 7+0} = 7

D () = min{c(,) + D (), c(,) + D ()} = min{+0, 7+} = node table 0 7 0 0 7 0 node table 0 node table 7 0 0 7 0 7 0 0 7 0 0 time D () = min{c(,) + D (), c(,) + D ()} = min{+, 7+0} = 7 6

node table 0 7 node table 0 node table 7 0 0 0 7 0 0 7 0 7 0 0 7 0 0 time 7 7

node table 0 7 node table 0 node table 7 0 0 0 7 0 0 7 0 7 0 0 7 0 0 0 0 time 7 8

node table 0 7 node table 0 node table 7 0 0 0 7 0 0 7 0 7 0 0 7 0 0 0 0 0 0 0 0 time 7 9

node table 0 7 node table 0 node table 7 0 0 0 7 0 0 7 0 7 0 0 7 0 0 0 0 0 0 0 0 time 7 40

node table 0 7 node table 0 node table 7 0 0 0 7 0 0 7 0 7 0 0 7 0 0 0 0 0 0 0 0 0 0 0 time 7 4

Mabe another eample 4

Distance Vector: link cost changes Link cost changes: node detects local link cost change updates routing info, recalculates distance vector 4 0 if DV changes, notif neighbors good news travels fast At time t 0, detects the link-cost change, updates its DV, and informs its neighbors. At time t, receives the update and updates its table. It computes a new least and sends its neighbors its DV. At time t, receives s update and updates its distance table. s least costs do not change and hence does not send an message to. 4

Distance Vector: link cost changes Link cost changes: good news travels fast bad news travels slow - count to infinit problem! 44 iterations before algorithm stabilies: see tet 60 4 0 4

Distance Vector: link cost changes Link cost changes: good news travels fast bad news travels slow - count to infinit problem! 44 iterations before algorithm stabilies: see tet 60 4 0 Poissoned 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 completel solve count to infinit problem? 46

Comparison of LS and DV algorithms Message compleit LS: with n nodes, E links, O(nE) msgs sent DV: echange between neighbors onl Speed of Convergence LS: O(n ) algorithm requires O(nE) msgs ma have oscillations DV: convergence time varies ma have routing loops count-to-infinit problem Robustness: what happens if router malfunctions? LS: DV: node can advertise incorrect link cost each node computes onl its own table DV node can advertise incorrect path cost each node s table used b others error propagate thru network 47

Hierarchical Routing and Routing on the Internet 49

Hierarchical Routing: Motivation Our routing stud thus far - idealiation all routers identical, network flat scale: with billions of destinations: can t store all dest s in routing tables! routing table echange would swamp links! administrative autonom internet = network of networks each network admin ma want to control routing in its own network 0

Hierarchical Routing aggregate routers into regions, autonomous sstems (AS) routers in same AS run same routing protocol intra-as routing protocol routers in different AS can run different intra- AS routing protocol Gatewa router Direct link to router in another AS Establishes a peering relationship Peers run an inter-as routing protocol

Interconnected ASes c a b AS a c d b Intra-AS Routing algorithm AS Forwarding table Inter-AS Routing algorithm a c AS b Forwarding table is configured b both intra- and inter-as routing algorithm Intra-AS sets entries for internal dests Inter-AS & Intra-As sets entries for eternal dests

Inter-AS tasks Obtain reachabilit information neighboring AS(s) Propagate this info to all routers within the AS All Internet gatewa routers run a protocol called BGPv4 (we will talk about this soon) c a b AS a c d b AS a c AS b

Intra-AS Routing Also known as Interior Gatewa Protocols (IGP) Most common Intra-AS routing protocols: RIP: Routing Information Protocol (DV protocol) OSPF: Open Shortest Path First (Link-State) Including hierarchical OSPF for large domains EIGRP: Enhanced Interior Gatewa Routing Protocol Cisco protocol: originall proprietar, made open in 0 4

Hierarchical OSPF

Internet inter-as routing: BGP BGP (Border Gatewa Protocol): the de facto standard BGP provides each AS a means to:. Obtain subnet reachabilit information neighboring ASs.. Propagate the reachabilit information to all routers internal to the AS.. Determine good routes to subnets based on reachabilit information and polic. Allows a subnet to advertise its eistence to rest of the Internet: I am here 6

BGP basics Pairs of routers (BGP peers) echange routing info over semipermanent TCP connections: BGP sessions Note that BGP sessions do not correspond to phsical links. When AS advertises a prefi to AS, AS is promising it will forward an datagrams destined to that prefi towards the prefi. AS can aggregate prefies in its advertisement c a b AS a AS c d b a c AS b ebgp session ibgp session 7

Path attributes & BGP routes When advertising a prefi, advert includes BGP attributes. prefi + attributes = route Two important attributes: AS-PATH: contains the ASs through which the advert for the prefi passed: AS 67 AS 7 NEXT-HOP: Indicates the specific internal-as router to net-hop AS. (There ma be multiple links current AS to net-hop-as.) When gatewa router receives route advert, uses import polic to accept/decline. 8

BGP route selection Router ma learn about more than route to some prefi. Router must select route. Elimination rules:. Local preference value attribute: polic decision. Shortest AS-PATH. Closest NEXT-HOP router: hot potato routing 4. Additional criteria 9

Qui based on China Telecom Incident Slides (0-46) used for qui found here: http://www.ida.liu.se/~nikca/papers/wiscs4.slides.pdf Attack described and characteried in R. Hiran, N. Carlsson, and P. Gill, Characteriing Large-scale Routing Anomalies: A Case Stud of the China Telecom Incident, Proc. PAM, Hong Kong, China, Mar. 0. 60

Multicast/Broadcast duplicate R duplicate creation/transmission R R R duplicate R R4 R R4 (a) (b) Source-duplication versus in-network duplication. (a) source duplication, (b) in-network duplication 6

Network Laer Routing: summar What we ve covered: network laer services routing principles: link state and distance vector hierarchical routing Internet routing protocols RIP, OSPF, BGP what s inside a router? Net stop: the Data link laer! 6