Goals of Routing Protocols ❒ Find the “optimal route” 10: ❒ Rapid Convergence Inter and intra AS, RIP, OSPF, ❒ Robustness BGP, Router Architecture ❒ Configurable to respond to changes in many variables (changes in bandwidth, delay, Last Modified: queue size, policy, etc.) 3/24/2003 2:39:16 PM ❒ Ease of configuration 4: Network Layer 4a-1 4: Network Layer 4a-2 Real Internet Routing? Recall CIDR We already talked about how routing based on hierarchical allocation of IP address space can allows efficient advertisement ❒ CIDR? of routing information: ❒ Dynamic routing protocols running between Organization 0 every router? 200.23.16.0/23 Organization 1 “Send me anything 200.23.18.0/23 with addresses Organization 2 beginning . Fly-By-Night-ISP 200.23.16.0/20” 200.23.20.0/23 . Internet Organization 7 . 200.23.30.0/23 “Send me anything ISPs-R-Us with addresses beginning 199.31.0.0/16” 4: Network Layer 4a-3 4: Network Layer 4a-4 CIDR? Dynamic Routing? Dynamic Routing Protocols? Our study of dynamic routing protocols thus far = ❒ CIDR by itself is a nice idea but.. idealized graph problem ❍ Hard to maintain ❒ all routers identical ❍ Work around existing IP address space ❒ network “flat” allocations ❍ What about redundant paths? … not true in practice ❒ Dynamic routing protocols? scale: with 50 million destinations: ❍ They maintain/update themselves ❒ can’t store all destinations in routing tables! ❍ Allow for redundant paths ❒ ❍ But could every router in the Internet be a routing table exchange would swamp links! node in the graph? ❒ Neither link state nor distance vector could handle the whole Internet! 4: Network Layer 4a-5 4: Network Layer 4a-6 1 Routing in the Internet Hierarchical Routing ❒ Administrative Autonomy Routers in same AS run gateway routers ❍ Internet = network of networks routing protocol ❒ special routers in AS ❍ Each network controls routing in its own network chosen by ❒ run intra-AS routing ❍ Global routing system to route between Autonomous Systems administrators of that protocol with all other (AS) domain routers in AS ❒ ❒ Two-level routing: ❍ “intra-AS” routing also responsible for ❍ Intra-AS: administrator is responsible for choice protocol routing to destinations outside AS ❍ Inter-AS: unique standard ❍ routers in different AS ❍ can run different intra- run inter-AS routing AS routing protocol protocol with other gateway routers 4: Network Layer 4a-7 4: Network Layer 4a-8 Internet AS Hierarchy Intra-AS and Inter-AS routing C.b Gateways: Intra-AS border (exterior gateway) routers B.a •perform inter-AS A.a routing amongst b A.c c a a themselves C b a B •perform intra-AS routers with other d c routers in their A b AS network layer inter-AS, intra-AS link layer routing in gateway A.c physical layer Inter-AS interior (gateway) routers 4: Network Layer 4a-9 4: Network Layer 4a-10 Intra vs Inter AS Routing Intra-AS and Inter-AS routing protcols Inter-AS ❒ C.b routing For Intra AS routing protocols: many choices; For between B.a Inter AS routing protocols: standard A and B A.a Host ❍ Why does this make sense? b A.c c h2 a a ❒ Intra AS routing protocols focus on performance C b a B optimization; Inter AS routing protocols focus on Host d c Intra-AS routing administrative issues h1 b A within AS B ❍ Why does this make sense? Intra-AS routing ❒ within AS A Choice in Intra-AS ❍ Intra-AS often static routing based on CIDR, can also be ❒ Single datagram is often routed over many hops dynamic (usually RIP or OSPF) via routes established by several intra-AS ❒ Standard Inter-AS BGP is dynamic routing protocols and an inter-AS routing protocol 4: Network Layer 4a-11 4: Network Layer 4a-12 2 Intra-AS Routing RIP ( Routing Information Protocol) ❒ Also known as Interior Gateway Protocols (IGP) ❒ Distance vector algorithm ❒ Most common IGPs: ❒ Included in BSD-UNIX Distribution in 1982 ❒ Single Distance metric: # of hops (max = 15 hops) ❍ RIP: Routing Information Protocol ❍ Can you guess why? ❍ OSPF: Open Shortest Path First ❍ Count to infinity less painful if infinity = 16 ☺ ❍ But limits RIP to networks with a diameter of 15 hops ❍ IGRP: Interior Gateway Routing Protocol (Cisco proprietary) ❒ Distance vectors: exchanged every 30 sec via ❍ Can also be static (via CIDR) but that is not Response Message (also called advertisement) called an IGP ❒ Each advertisement: route to up to 25 destination nets 4: Network Layer 4a-13 4: Network Layer 4a-14 RIP: Link Failure and Recovery RIP Table processing If no advertisement heard after 180 sec --> ❒ neighbor/link declared dead RIP routing tables managed by application-level process called route-d (daemon) ❍ routes via neighbor invalidated ❒ advertisements sent in UDP packets, periodically ❍ new advertisements sent to neighbors repeated ❍ neighbors in turn send out new advertisements (if ❒ Periodically inform kernel of routing table to use tables changed) ❍ link failure info quickly propagates to entire net ❍ poison reverse used to prevent small loops ❍ infinite distance = 16 hops to make make problem with larger loops less painful 4: Network Layer 4a-15 4: Network Layer 4a-16 RIP Table example: netstat -rn OSPF (Open Shortest Path First) ❒ “open”: publicly available Destination Gateway Flags Ref Use Interface ❒ Uses Link State algorithm -------------------- -------------------- ----- ----- ------ --------- ❍ 127.0.0.1 127.0.0.1 UH 0 26492 lo0 LS packet dissemination 192.168.2. 192.168.2.5 U 2 13 fa0 ❍ Topology map at each node 193.55.114. 193.55.114.6 U 3 58503 le0 192.168.3. 192.168.3.5 U 2 25 qaa0 ❍ Route computation using Dijkstra’s algorithm 224.0.0.0 193.55.114.6 U 3 0 le0 default 193.55.114.129 UG 0 143454 ❒ OSPF advertisement carries one entry per neighbor ❒ Three attached class C networks (LANs) router (i.e. cost to each neighbor) ❒ Router only knows routes to attached LANs ❒ Advertisements disseminated to entire AS (via ❒ Default router used to “go up” flooding) ❒ Route multicast address: 224.0.0.0 ❒ Loopback interface (for debugging) 4: Network Layer 4a-17 4: Network Layer 4a-18 3 OSPF “advanced” features (not in RIP) Hierarchical OSPF: Mini Internet ❒ Many have nothing to do with link-state vs distance Within each area, border router vector!! responsible for routing outside the area ❒ Security: all OSPF messages authenticated (to Exactly one area prevent malicious intrusion); TCP connections used is backbone area ❒ Multiple same-cost paths can be used at once (single path need not be chosen as in RIP) ❒ For each link, multiple cost metrics for different TOS (eg, high BW, high delay satellite link cost may set “low” for best effort; high for real time) ❒ Integrated uni- and multicast support: ❍ Multicast OSPF (MOSPF) uses same topology data base as OSPF ❒ Hierarchical OSPF in large domains Backbone area contains all area border routers and possibly others ❍ Full broadcast in each sub domain only 4: Network Layer 4a-19 4: Network Layer 4a-20 Hierarchical OSPF IGRP (Interior Gateway Routing Protocol) ❒ Two-level hierarchy: local area, backbone. ❒ CISCO proprietary; successor of RIP (mid 80s) ❍ Link-state advertisements only in area ❒ Distance Vector, like RIP but with advanced ❍ each nodes has detailed area topology; only know features like OSPF direction (shortest path) to nets in other areas. ❒ several cost metrics (delay, bandwidth, reliability, ❒ Area border routers: “summarize” distances to nets load etc); administer decides which cost metrics in own area, advertise to other Area Border routers. to use ❒ Backbone routers: run OSPF routing limited to ❒ uses TCP to exchange routing updates backbone. ❒ Loop-free routing via Distributed Updating Alg. ❒ Boundary routers: connect to other ASs. (DUAL) based on diffused computation 4: Network Layer 4a-21 4: Network Layer 4a-22 Now on to Inter-AS routing Autonomous systems ❒ The Global Internet consists of Autonomous Systems (AS) interconnected with each other: ❍ Stub AS: small corporation ❍ Multihomed AS: large corporation (no transit traffic) ❍ Transit AS: provider (carries transit traffic) ❒ Major goal of Inter-AS routing protocol is to reduce transit traffic 4: Network Layer 4a-23 4: Network Layer 4a-24 4 Internet inter-AS routing: BGP Internet inter-AS routing: BGP ❒ BGP (Border Gateway Protocol): the de facto Suppose: gateway X send its path to peer gateway W standard ❒ W may or may not select path offered by X ❒ Path Vector protocol: ❍ cost, policy (don’t route via competitors AS!), loop ❍ similar to Distance Vector protocol prevention reasons. ❍ Avoids count-to-infinity problem by identifying ❒ If W selects path advertised by X, then: yourself in a path advertised to you Path (W,Z) = w, Path (X,Z) ❍ each Border Gateway broadcast to neighbors ❒ Note: X can control incoming traffic by controlling its (peers) entire path (I.e, sequence of ASs) to route advertisements to peers: destination ❍ e.g., don’t want to route traffic to Z -> don’t ❍ E.g., Gateway X may send its path to dest. Z: advertise any routes to Z Path (X,Z) = X,Y1,Y2,Y3,…,Z 4: Network Layer 4a-25 4: Network Layer 4a-26 Internet inter-AS routing: BGP Internet Map ❒ BGP messages exchanged using TCP. ❒ Now that we know about autonomous ❒ BGP messages: systems and intra and inter AS routing ❍ OPEN: opens TCP connection to peer and protocols authenticates sender ❒ What does the Internet really look like? ❍ UPDATE: advertises new path (or withdraws old) ❍ That is a actually a hard question to answer ❍ KEEPALIVE keeps connection alive in absence of UPDATES; also ACKs OPEN request ❍ Internet Atlas Project • http://www.caida.org/projects/internetatlas/ ❍ NOTIFICATION: reports errors in previous msg; also used to close connec tion • Techniques, software, and protocols for mapping the Internet, focusing on Internet topology, performance, workload, and routing data 4: Network Layer 4a-27 4: Network Layer 4a-28 CAIDA: NSFNET growth until The Internet around 1990 1995 Backbone nodes elevated 4: Network Layer 4a-29 Low Traffic Volume High 4: Network Layer 4a-30 5 NSF Networking Architecture Network Access Point of Late 1990s ❒ NSFNET Backbone Project successfully ❒ Allows Internet Service Providers (ISPs), transitioned to a new networking government, research, and educational architecture in 1995.