Advanced Computer Networks. Lecture 4 Internetworking and Routing

Transcription

1 Advanced Computer Networks Lecture 4 Internetworking and Routing 1

2 Techniques Naïve: Flooding Outline Distance vector: Distributed Bellman Ford Algorithm Link state: Dijkstra s Shortest Path First-based Algorithm Routing in the Internet Hierarchy and Autonomous Systems Interior Routing Protocols: RIP, OSPF Exterior Routing Protocol: BGP Multicast Routing Routing is a very complex subject, and has many aspects. Here, we will concentrate on the basics. 2

3 A The Problem B R 2 R 1 R 4 How does R 1 choose a next-hop on the path towards host B? R 3 3

4 Routing Metrics Metrics Delay to send an average size packet (Make high speed links attractive, but closeness counts) Bandwidth Link utilization Stability: Is a link (or path) up or down? Today: about 1/3 of Internet routes are asymmetric 4

5 A Example network Objective: Determine the route from A to B that minimizes the path cost. Examples of link cost: Distance, data rate, price, congestion/delay, R R 2 R 4 R R 3 4 R 5 2 R 7 3 R 8 2 B 5

6 Example network In this simple case, solution is clear from inspection A R R 2 R 4 R R 3 4 R 5 2 R 7 3 R 8 2 B 6

7 So what about this network...!? The public Internet in 1999 Learn more at 7

8 Technique 1: Naïve Approach Flood! -- Routers forward packets to all ports except the ingress port. R 1 Advantages: Simple. Every destination in the network is reachable. Disadvantages: Some routers receive a packet multiple times. Packets can go round in loops forever. Inefficient. 8

9 Spanning Trees Objective: Find the lowest cost route from each of (R 1,, R 7 ) to R 8. R R 2 R R R R R 6 R 8 2 9

10 R 1 R 3 A Spanning Tree R 2 R 5 R R R 6 R 8 2 The solution is a spanning tree with R 8 as the root of the tree. Tree: There are no loops. Spanning: All nodes included. We ll see two algorithms that build spanning trees automatically: The distributed Bellman-Ford algorithm Dijkstra s shortest path first algorithm 10

11 Technique 2: Distance Vector The Distributed Bellman-Ford Algorithm 1. Let X = ( C, C,..., C ) where: C = cost from R to R. 2. Set X n i i = (,,,..., ). 3. Every T seconds, router i sends C, to for all i, to its neighbors. This is the Distance vector. 4. If router i is told of a lower cost path to R n+ 1, it updates C. Hence, n+ 1 i f( X ) where f (.) determines the next step improvement. 5. If X X then goto step (3). 6. Stop. n i X = n 11

12 Bellman-Ford Algorithm R 1 R 2 Example R R R 3 4 R 8 R 4 4 R 6 R 1 R 2 Inf Inf R 3 4, R 8 R 4 Inf R 5 2, R 8 R 6 2, R 8 R 7 3, R R R R 2 4 R R 7 R R 3 R

13 Bellman-Ford Algorithm R 1 6, R 3 R 2 4, R 5 R 3 4, R 8 R 4 6, R 7 R 5 2, R 8 R 6 2, R 8 R 7 3, R 8 R 1 5, R 2 R 2 4, R 5 R 3 4, R 8 R 4 5, R 2 R 5 2, R 8 R 6 2, R 8 R 7 3, R R R 1 2 R 2 2 R 4 R R 2 R R 3 R 8 Solution R R R R R 7 R R

14 Bellman-Ford Algorithm Questions: 1. How long can the algorithm take to run? 2. How do we know that the algorithm always converges? 3. What happens when link costs change, or when routers/links fail? 14

15 A Problem with Bellman-Ford Bad news travels slowly R R 2 R 3 R 4 Consider the calculation of distances to R 4 : Time R 1 R 2 R 3 3,R 2 2,R 3 3,R 2 2,R 3 3,R 2 4,R 3 5,R 2 4,R 3 Counting to infinity 1, R 4 3,R 2 3,R 2 5,R 2 R 3 R 4 fails 15

16 Counting to Infinity Problem Solutions 1. Set infinity = some small integer (e.g. 16). Stop when count = Split Horizon: Because R 2 received lowest cost path from R 3, it does not advertise cost to R 3 3. Split-horizon with poison reverse: R 2 advertises infinity to R 3 4. There are many problems with (and fixes for) the Bellman-Ford algorithm. 16

17 Example: Link failure with split horizon R 1 3 A R 2 R 3 3 R 4 R 5 4 B 2 17

18 A 0, A A 0, A A 0, A A 0, A R 1 3, A R 1 3, A R 1 3, A R 1 3, A R 2 4, A R 2 4, A R 2 4, A R 2 4, A 3 A 4 4 R 3 4, A R 4 6, R 2 R 5 7, R 3 R 3 inf, A R 4 6, R 2 R 5 7, R 3 R 3 inf R 4 6, R 2 R 5 inf, R 3 inf R 4 6, R 2 R 5 inf, R 1 B 9, R 5 B 9, R 5 B 9, R 5 B inf R 2 R A R 1 0, A 3, A A R 1 0, A 3, A A R 1 0, A 3, A R 4 4 B 2 R 5 R 2 R 3 R 5 4, A inf 6, R 2 inf R 2 R 3 4, A inf, A R 4 6, R 2 R 5 12, B R 2 4, A R 3 15, R 5 R 4 6, R 2 R 5 12, B B 10,R 4 B 10,R 4 B 10,R 4 18

19 Technique 3: Link State Dijkstra s Shortest Path First Algorithm Routers send out update messages whenever the state of an incident link changes. Called Link State Updates Based on all link state updates received each router calculates lowest cost path to all others, starting from itself. Use Dijkstra s single-source shortest path algorithm Assume all updates are consistent At each step of the algorithm, router adds the next shortest (i.e. lowest-cost) path to the tree. Finds spanning tree rooted at the router. 19

20 Reliable Flooding of LSP The Link State Packet: The ID of the router that created the LSP List of directly connected neighbors, and cost Sequence number TTL Reliable Flooding Resend LSP over all links other than incident link, if the sequence number is newer. Otherwise drop it. Link State Detection: Link layer failure Loss of hello packets 20

21 Dijkstra s Shortest Path First Algorithm Example Step 1: Shortest path set, S = { R8}. Candidate set, C= { R3, R5, R7, R6} Step 2: S = { R, R}, 8 5 C= { R, R, R, R } R 5 R 8 Step 3: C S = { R, R, R}, = { R, R, R, R } R 5 R 6 R 8 Step 4: S = { R, R, R, R }, C= { R, R, R }. R 5 R 7 R 6 R 8 21

22 Dijkstra s SPF Algorithm Step 8 : S = { R8, R5, R6, R7, R2, R1, R4}, C = {}. R R 2 R 4 R 6 2 R 3 4 R 7 R R 8 22

23 Distance Vector vs Link State Messages Size: small with LS; potentially large with DV Exchange: LS flood!; DV only to neighbors Space requirements LS maintains entire topology DV maintains only neighbor state Robustness: LS can broadcast incorrect/corrupted LSP Can be made robust since sources are aware of alternate paths DV can advertise incorrect paths to all destinations Incorrect calculation can spread to entire network Examples (coming up later): LS: OSPF DV: RIP, RIP2 23

24 Techniques Outline Flooding Distributed Bellman Ford Algorithm Dijkstra s Shortest Path First Algorithm Routing in the Internet Hierarchy and Autonomous Systems Interior Routing Protocols: RIP, OSPF Exterior Routing Protocol: BGP Multicast Routing 24

25 Routing in the Internet The Internet uses hierarchical routing The Internet is split into Autonomous Systems (AS s) Examples of AS s: Stanford (32), HP (71), MCI Worldcom (17373) Try: whois h whois.arin.net ASN MCI Worldcom Within an AS, the administrator chooses an Interior Gateway Protocol (IGP) Examples of IGPs: RIP (rfc 1058), OSPF (rfc 1247). Between AS s, the Internet uses an Exterior Gateway Protocol AS s today use the Border Gateway Protocol, BGP-4 (rfc 1771) 25

26 Routing in the Internet AS A AS B AS C BGP BGP Interior Gateway Protocol Interior Gateway Protocol Interior Gateway Protocol Stub AS Transit AS e.g. backbone service provider Stub AS 26

27 Routing within a Stub AS There is only one exit point, so routers within the AS can use default routing. Each router knows all Network IDs within AS. Packets destined to another AS are sent to the default router. Default router is the border gateway to the next AS. Routing tables in Stub AS s tend to be small. 27

28 Interior Routing Protocols RIP Uses distance vector (distributed Bellman-Ford algorithm). Updates sent every 30 seconds. No authentication. Originally in BSD UNIX. Widely used for many years; not used much anymore. OSPF Link-state updates sent (using flooding) as and when required. Every router runs Dijkstra s algorithm. Authenticated updates. Autonomous system may be partitioned into areas. Widely used. 28

29 Exterior Routing Protocols Problems: Topology: The Internet is a complex mesh of different AS s with very little structure. Autonomy of AS s: Each AS defines link costs in different ways, so not possible to find lowest cost paths. Trust: Some AS s can t trust others to advertise good routes (e.g. two competing backbone providers), or to protect the privacy of their traffic (e.g. two warring nations). Policies: Different AS s have different objectives (e.g. route over fewest hops; use one provider rather than another). 29

30 Border Gateway Protocol (BGP-4) BGP is not a link-state or distance-vector routing protocol. Instead, BGP uses Path vector BGP advertises complete paths (a list of AS s). Also called AS_PATH (this is the path vector) Example of path advertisement: The network /16 can be reached via the path {AS1, AS5, AS13}. Paths with loops are detected locally and ignored. Local policies pick the preferred path among options. When a link/router fails, the path is withdrawn. 30

31 Customers and Providers provider provider customer IP traffic customer Customer pays provider for access to the Internet Customer may not always need BGP 31

32 Customer-Provider Hierarchy provider customer IP traffic 32

33 The Peering Relationship peer provider traffic allowed peer customer traffic NOT allowed Peers provide transit between their respective customers Peers do not provide transit between peers Peers (often) do not exchange $$$ 33

34 BGP Messages Open : Establish a BGP session. Keep Alive : Handshake at regular intervals. Notification : Shuts down a peering session. Update : Announcing new routes or withdrawing previously announced routes. BGP announcement = prefix + path attributes Attributes include: Next hop, AS Path, local preference, Multi-exit discriminator, Used to select among multiple options for paths 34

35 BGP Route Selection Summary Highest Local Preference Enforce relationships E.g. prefer customer routes over peer routes Shortest ASPATH Lowest MED i-bgp < e-bgp Lowest IGP cost to BGP egress traffic engineering Lowest router ID Throw up hands and break ties 35

36 ASPATH Attribute /16 AS Path = AS 1129 Global Access /16 AS Path = AS 1755 Ebone /16 AS Path = AS 1239 Sprint /16 AS Path = Pick shorter AS path AS RIPE NCC RIS project /16 AS Path = 6341 AS 6341 AT&T Research /16 Prefix Originated AS 7018 AT&T /16 AS Path = /16 AS Path = AS 3549 Global Crossing 36

37 So Many Choices peer provider peer customer AS 4 Frank s Internet Barn AS 3 AS 2 Which route should Frank pick to /16? AS /16 37

38 peer provider peer customer Frank s Choices Route learned from customer preferred over route learned from peer, preferred over route learned from provider AS 4 local pref = 80 local pref = 90 AS 3 local pref = 100 Set appropriate local pref to reflect preferences: Higher Local preference values are preferred AS 2 AS /16 38

39 Traceroute with ASNs TTL LFT trace to :80/tcp 1 [AS7011] [ELI-NETWORK-ELIX] eli-gw.home.mainnerve.net ( ) 20.2ms 2 [AS5650] [ELI-NETBLK98] ms 3 [AS5650] [ELI-NETBLK99] s gw01.phnx.eli.net ( ) 20.3ms 4 [AS5650] [ELI-2-NETBLK99] srp2-0.cr01.phnx.eli.net ( ) 20.3ms 5 [AS5650] [ELI-NETBLK5] p6-0.cr01.lsan.eli.net ( ) 40.3ms 6 [AS5650] [ELI-NETBLK5] p9-0.cr02.sntd.eli.net ( ) 40.3ms 7 [AS5650] [ELI-2-NETBLK99] srp3-0.cr01.sntd.eli.net ( ) 40.3ms 8 [AS5650] [ELI-NETBLK5] so er01.plal.eli.net ( ) 40.3ms 9 [AS5650] [SAVVIS] bpr2-ge paloaltopaix.savvis.net ( ) 40.2ms 10 [ASN?] [SAVVIS] dcr2-so sanfranciscosfo.savvis.net ( ) 40.3ms 11 [ASN?] [SAVVIS] dcr1-loopback.washington.savvis.net ( ) 100.4ms 12 [ASN?] [SAVVIS] bhr1-pos-10-0.sterlingdc2.savvis.net ( ) 100.5ms 13 [ASN?] [SAVVIS] csr1-ve240.sterlingdc2.savvis.net ( ) 100.5ms [neglected] no reply packets received from TTL [ASN?] [SAVVIS] [target] : ms 39

40 Who owns an address block? prompt> whois OrgName: Savvis OrgID: SAVVI-2 Address: 3300 Regency Parkway City: Cary StateProv: NC PostalCode: Country: US ReferralServer: rwhois://rwhois.exodus.net:4 321/ NetRange: CIDR: /14 NetName: SAVVIS NetHandle: NET Parent: NET NetType: Direct Allocation NameServer: DNS01.SAVVIS.NET NameServer: DNS02.SAVVIS.NET NameServer: DNS03.SAVVIS.NET NameServer: DNS04.SAVVIS.NET Comment: RegDate: Updated: # ARIN WHOIS database, last updated :10 # Enter? for additional hints on searching ARIN's WHOIS database. 40

41 Prompt> whois SU-NET OrgName: Stanford University OrgID: STANFO Address: Pine Hall 115 City: Stanford StateProv: CA PostalCode: Country: US NetRange: CIDR: /16 NetName: SU-NET NetHandle: NET Parent: NET NetType: Direct Assignment NameServer: ARGUS.STANFORD.EDU NameServer: AVALLONE.STANFORD.EDU NameServer: ATALANTE.STANFORD.EDU Comment: RegDate: Updated: TechHandle: JK535-ARIN TechName: Kohn, Jay TechPhone: Tech Organizations North American AS Numbers and Addresses DNS Top level domains and delegates IP Address blocks # ARIN WHOIS database, last updated :10 To receive AS from a particular route arbiter: Whois h whois.ra.net

42 Multicast Routing Applications that benefit from multicast. Trees, addressing and forwarding. Host/Router Interaction Multicast routing Distance Vector-based (DVMRP, PIM-DM) Link-state based (MOSPF) Rendezvous-based (PIM-SM, CBT) Some interesting questions 42

43 Multicast Trees The basic idea G Server G Server G G G G G G G G Multiple unicasts Single multicast 43

44 Applications that need multicast One way, single sender: one-to-many TV Non-interactive learning Database update Information dispersal (e.g. Pointcast) Software updates/patches Two way, interactive, multiple sender: many-to-many Teleconference Interactive learning 44

45 Multicast Routing A multicast tree is a spanning tree with the sender at the root, spanning all the members of the group. 45

46 Multicast Trees e.g. a teleconference S 1 Sender/Speaker Multicast Group (S 1,G) S 1 Class D R 46

47 IP Multicast Addresses Class D IP addresses Group ID How to allocated these addresses? Well-known multicast addresses, assigned by IANA Transient multicast addresses, assigned and reclaimed dynamically, e.g., by sdr program 47

48 IP Multicast Service Model (rfc1112) Each group identified by a single IP address Groups may be of any size Members of groups may be located anywhere in the Internet Members of groups can join and leave at will Senders need not be members Group membership not known explicitly Analogy: Each multicast address is like a radio frequency, on which anyone can transmit, and to which anyone can tune-in. 48

49 Multicast Trees and Addressing All members of the group share the same Class D Group Address. An end station may be the member of multiple groups. An end-station joins a multicast group by (periodically) telling its nearest router that it wishes to join (uses IGMP Internet Group Management Protocol). Routers maintain soft-state indicating which end-stations have subscribed to which groups. 49

50 Multicast Trees Multiple source trees S 2 Class D R S 2 Sender/Speaker Multicast Group (S 2,G) 50

51 Multicast Forwarding is Sender-specific Group Address Src Address Src Interface Dst Interface G S 1 1 2,3 S 2 2 1,3 S 1 G 1 R S 2 G 51

52 Outline Applications that need multicast. Trees, addressing and forwarding. Host/Router Interaction Multicast routing Distance Vector-based: DVMRP, PIM-DM Link-state based: MOSPF Rendezvous-based: PIM-SM, CBT Some interesting problems 52

53 IP Multicast Architecture Service model Host-to-router protocol (IGMP) Hosts Routers Multicast routing protocols (various) 53

54 Internet Group Management Protocol End system to router protocol is IGMP Each host keeps track of which mcast groups are subscribed to Socket API informs IGMP process of all joins Objective is to keep router up-to-date with group membership of entire LAN Routers need not know who all the members are, only that members exist 54

55 How IGMP Works Routers: Q Hosts: On each link, one router is elected the querier Querier periodically sends a Membership Query message to the all-systems group ( ), with TTL = 1 On receipt, hosts start random timers (between 0 and 10 seconds) for each multicast group to which they belong 55

56 How IGMP Works (cont.) Routers: Q Hosts: G G G G When a host s timer for group G expires, it sends a Membership Report to group G, with TTL = 1 Other members of G hear the report and stop their timers Routers hear all reports, and time out non-responding groups 56

57 How IGMP Works (cont.) Note that, in normal case, only one report message per group present is sent in response to a query Power of randomization + suppression Query interval is typically seconds When a host first joins a group, it sends one or two immediate reports, instead of waiting for a query 57

58 Outline Applications that need multicast. Trees, addressing and forwarding. Host/Router Interaction Multicast routing Distance Vector-based: DVMRP, PIM-DM Link-state based: MOSPF Rendezvous-based: PIM-SM, CBT Some interesting problems 58

59 Distance-vector Multicast RPB: Reverse-Path Broadcast Uses existing unicast shortest path routing table. Computed using Distance vector If packet arrived through interface that is the shortest path to the packet s SA, then forward packet to all interfaces. Else drop packet. 59

60 Distance-vector Multicast RPB: Reverse-Path Broadcast Unicast DV Routing Table Address Port S 1 1 S 1 Sender/Speaker Multicast Group (S 1,G) 1 Shortest Path to Source 3 2 LAN Q: Is it shortest path from source? 60

61 Distance-vector Multicast RPB: Reverse-Path Broadcast S 1 Sender/Speaker Multicast Group (S 1,G) Designated Parent Router: One parent router picked per LAN (one closest to source). LAN 61

62 Distance-vector Multicast RPM: Reverse-Path Multicast RPM = RPB + Prune RPB used when a source starts to send to a new group address. Routers that are not interested in a group send prune messages up the tree towards source. Prunes sent implicitly by not indicating interest in a group. DVMRP works this way. 62

63 Protocol Independent Multicast PIM-DM (Dense Mode) uses RPM. PIM-SM (Sparse Mode) designed to be more efficient that DVMRP Key idea: use a rendezvous point (RP) Routers explicitly join multicast tree by sending unicast Join and Prune messages. Routers join a multicast tree via an RP for each group. Several RPs per domain (picked in a complex way). Provides either: Shared tree for all senders (default) Source-specific tree 63

64 Unicast to R3: (S,G) Source knows to send all G packets to RP. PIM-SM RP unwraps mcast pkt and forwards to local tree. Source tunnels mcast-in-ucast packets to RP. S RP R2 R1 Unicast to RP: (*, G) RP knows R1 has joined R2 learns to send (*,G) packets to R1 IGMP Join Sender/Source 64

65 PIM-SM S RP R2 R1 Optional Sourcespecific join to bypass RP router: Routers along the way learn new path for (S,G). Sender/Source 65

66 Outline Applications that need multicast. Trees, addressing and forwarding. Host/Router Interaction Multicast routing Distance Vector-based: DVMRP, PIM-DM Link-state based: MOSPF Rendezvous-based: PIM-SM, CBT Some interesting problems 66

67 Multicast: Interesting Questions How to make multicast reliable? How to implement flow-control? How to support/provide different rates for different end users? How to secure a multicast conversation? Will multicast become widespread? Several protocols for multicast routing in IP But IP multicast is not enabled in routers! No one uses IP multicast, really End-system based, overlay-based approaches more popular 67

68 Overlay Multicast 68

69 Failure of IP Multicast Not widely deployed even after 15 years! Use carefully e.g., on LAN or campus, rarely over WAN Various failings Scalability of routing protocols Hard to manage Hard to implement TCP equivalent Hard to get applications to use IP Multicast without existing wide deployment Hard to get router vendors to support functionality and hard to get ISPs to configure routers to enable 69

70 Supporting Multicast on the Internet Application IP?? At which layer should multicast be implemented? Network Internet architecture 70

71 IP Multicast Berkeley MIT UCSD routers end systems multicast flow CMU Highly efficient Good delay 71

72 End System Multicast Berkeley MIT MIT1 UCSD MIT2 CMU1 CMU Berkeley UCSD Overlay Tree MIT1 MIT2 CMU2 CMU1 CMU2 72

73 Benefits Over IP Multicast Quick deployment All multicast state in end systems Computation at forwarding points simplifies support for higher level functionality UCSD Berkeley MIT MIT1 MIT2 CMU1 CMU CMU2 73

74 Concerns with End System Multicast Self-organize recipients into multicast delivery overlay tree Must be closely matched to real network topology to be efficient Performance concerns compared to IP Multicast Increase in delay Bandwidth waste (packet duplication) Penalty can be kept small in practice Berkeley MIT1 Berkeley MIT1 UCSD MIT2 UCSD MIT2 CMU1 CMU1 IP Multicast CMU2 End System Multicast CMU2 74

75 Important Concepts Multicast provides support for efficient data delivery to multiple recipients Requirements for IP Multicast routing Keeping track of interested parties Building distribution tree Broadcast/suppression technique Difficult to deploy new IP-layer functionality End system-based techniques can provide similar efficiency Easier to deploy 75