Midterm Exam Name: _____________________ Autumn 1999 SSN: ______________________ Computer Networks TDC463 * You have the entire class time to finish this exam. * NO calculators or other electronic devices allowed. * Show all work when problems involve any calculations. * Be neat, if I cannot read it, it's wrong. * Read questions and instructions carefully. * DON'T PANIC! Total Score Possible: 74 For questions 1 through 10, circle the best answer(s) listed. Point values are not an indicator of how many answers you need to circle. 1. It is not possible to mix devices at different data rates on a shared (repeater) hub. [1 point] True False Answer: True 2. In most countries, the telecommunications network is run by: [1 point] a. AT&T b. A national PTT c. A subcommittee of the IEEE d. The CCITT (renamed in 1993 to ITU-T) e. The ISO Answer: b 3. Which routing scheme collects data from its near and far neighbors using LSPs then applies the Dijkstra algorithm to build a map of shortest paths to each network? [1 point] a. RIP b. Bellman-Ford c. Spanning Tree d. Link State e. BGP Answer: d 4. Ethernet networks can use either Transparent or Source Route Bridging. [1 point] True False Answer: False 5. RIPv2 differs from RIPv1 in the following ways: [1 points] a. Supports IPv6 b. Supports subnet masks c. Supports multiple metrics d. Support authentication e. Supports multiple layer 3 protocols Answer: b and d Scoring: 0.5 for each one correct, -0.5 for each one incorrect 6. In Ethernet, late collisions [2 points] a. are normal events, used to arbitrate access to the medium. b. occur within the last few bit times of a minimum sized Ethernet frame. c. are theoretical events only, they never occur in practice. d. are detrimental to network performance and should be corrected immediately. Answer: d Scoring: If right answer plus any other, substract -1 for each one wrong 7. The following are characteristic of the Spanning Tree algorithm: [2 points] a. At least two master bridges (roots) are designed for redundancy. b. Its primary purpose is to create a loop-free environment. c. Bridges communicate with end stations (hosts) to determine optimal paths. d. BDPU messages are exchanged between bridges. Answer: b and d Scoring: b = 1.5, d = 0.5, wrong answers -0.5 8. The three primary functions of the network layer are: [2 points] a. Error control b. Routing c. Flow control d. Addressing e. Fragmentation Answer: b, d and e Scoring: -1 for each one wrong 9. ________ performs error control and flow control at the data link layer. [1 point] a. MAC b. LLC c. Category 5 cabling d. ARP e. IP Answer: b 10. Which of the following are common names for a packet switch: [1 point] a. Gateway b. Layer 3 switch c. Router d. Intermediate system Answer: a, b, c and d Scoring: 0.25 for each For questions 11 through 20, fill in the blank with the most appropriate information. 11. The maximum number of routing hops to reach a network using RIPv1 is ______ hops. [1 point] Answer: 16 12. The IP address 10011011.11100010.00110010.01101000 in binary is ___________________ in dotted decimal notation. [2 points] Answer: 155.226.50.104 13. The subnet mask, in dotted decimal notation, for any IP address with a '/18' mask is ____________________. [2 points] Answer: 255.255.192.0 14. A token frame in a Token Ring network is _________ bits long. [1 point] Answer: 24 15. The minimum size Ethernet frame is _______ bits long. [1 point] Answer: 512 16. Mobile IP requires a _________ agent and a _________ agent. [1 point] Answer: Home and Foreign Scoring: 0.5 for each 17. An IPv6 address is ________ times larger than an IPv4 address. [2 points] Answer: 4 18. Assume the following: (1) A network has an n-layer protocol hierarchy. (2) Applications generate messages of length M bytes. (3) At each of the layers, an h-byte header is added. Then the fraction of network bandwidth consumed by headers is ____________/____________. [2 points] Answer: n*h/(m+(n*h)) 19. TCP/IP is an acronym for ____________________________________. [1 point] Answer: Transport Control Protocol/Intenret Protocol Scoring: 0.5 if half correct 20. CSMA/CD is an acronym for ___________________________________. [1 point] Answer: Carrier Sense Multiple Access/Collision Detection Socring: 0.5 if half correct 21a. Match the terms with the best corresponding concepts or definitions. [7 points] Scoring: 0.5 for each correct __c__ datagram a. Translation of a host MAC address to IP address __g__ presentation b. An algorithm to determine which bridges do not forward frames __n__ IETF c. TCP/IP's term for an internet packet __k__ back-off d. TCP/IP's error reporting mechanism __f__ BOOTP/DHCP e. A measure of capacity on a physical transmission system __j__ transport f. Dynamically assigned IP addresses to hosts __i__ ISO g. A OSI layer that describes how to represent data __a__ ARP h. A measure of data throughput in a computer network __m__ TTL i. The organization which defined the OSI Reference Model __e__ bandwidth j. An OSI layer which defines end-to-end flow control __h__ bps k. Follows the jam signal in Ethernet networks __b__ Spanning Tree l. The opposite of NAK __l__ ACK m. An IP header field used to control how many hops a packet can make __d__ ICMP n. The group that creates and maintains Internet RFCs 21b. Complete the following table of special IP addresses as best you can. [ 4 points] Scoring: 0.5 for each correct +----------------+----------------+------------------------+---------------+ |Address Prefix | Address Suffix | Name/Purpose and/or | Example | |(net/subnet id) | (host id) | Description | | +----------------+----------------+------------------------+---------------+ 01111111 variable Loopback 127.0.0.1 ---------------------------------------------------------------------------- All binary 1's All binary 1's Limited broadcast 255.255.255.255 ---------------------------------------------------------------------------- net/subnet id All binary 1's Directed broadcast 16.2.4.255/24 ---------------------------------------------------------------------------- first 5 bits of IP address are variable Class E - reserved 251.6.5.2 11110 ---------------------------------------------------------------------------- All binary 0's All binary 0's this net, this host 0.0.0.0 ---------------------------------------------------------------------------- first 4 bits of IP address are variable Class D - multicast 224.1.19.3 1110 ---------------------------------------------------------------------------- Network and subnet id All binary 0's Specifies network 140.192.1.0/25 ---------------------------------------------------------------------------- For questions 22 through 28, briefly answer the question in 1 to 5 sentences. 22. What is the difference between a static route and a default route? [2 points] Answer: A static route is manaully, hard coded configured route to a network. A default route is a route of last resort, when there is no match for a route in the routing table. Scoring: 1 point for each definition 23. Given an IP address of 130.19.52.12 and subnet mask of 255.255.255.192, determine the following information: [5 points] a. Directed broadcast address in dotted decimal notation. b. The maximum number of hosts that can have an IP address on this network. c. The 'this network' address (network/subnet id) using dotted decimal notation. Answer (a): 130.19.52.63 (b): 62 (64-2) (c): 130.19.52.0/26 Scoring: 1 point for a, 2 points for b, 2 points for c 24. What advantage does a circuit switched network (i.e. PSTN) have over a packet switched network (i.e. the Internet)? [2 points] Answer: The bandwidth is fixed. The delay is fixed. The error rate is fixed. Constant performance. Scoring: Partial credit based on answer 25. Besides the source address, destination address and the data, what are two other fields you might find in an IP datagram header? Be specific and describe each in two sentences or less. [2 points] Answer: Version - 4 bit field specifies the IP version TTL - time to live, used to control maximum router hops Socring: 1 point for each 26. A TCP/IP packet goes through the Internet to a destination, but is delayed en route. A sender's acknowledgement timer expires so a retransmission is sent. The retransmitted packet arrives and then the original packet arrives shortly thereafter. What should be done with the original packet? [1 point] Answer: You should discard the late packet. 27. Suppose that in the network layer, routers and links were subjected to "stressful" conditions that might cause them to fail fairly often. Does this argue for a connection-oriented or a connectionless environment? [1 point] Answer: Connectionless. Connectionless would be able to route around the problem, without the problem of having to setup a connection each time after a failure. 28. Why is it necessary to pad an Ethernet frame to 64 bytes? [3 points] Answer: To allow stations in the collision domain time to see the frame and to transmit long enough (2 times the network diameter) in order to detect a collision. 29 and 30. Pick TWO of the following questions from I. through X. to answer. Answer with as much level of detail as possible, but not more than 1 page worth of text. You will be graded on the number of unique points you uncover and describe (or in other words, how well you know the material). [10 points each] Answers and scoring for the following questions vary depending on the understanding of the problem. I. Describe the major architectural differences, using examples from the OSI Reference Model, between the public switched telephone network (PSTN) and a TCP/IP network (Internet). II. What are the differences between a unicast, multicast and broadcast address? How is each used? Give some examples with layer 2 and layer 3 protocols. III. Describe the process of transmitting a frame from one token ring network, to another token ring network, which is separated by a source routing bridge. Assume the sender cannot send the frame immediately. Draw a diagram to help illustrate the concepts. IV. Describe the process of transmitting a frame from one Ethernet network to another Ethernet network, which is separated by a transparent bridge. Assume the sender cannot send the frame immediately. Draw a diagram to help illustrate the concepts. V. What is hiearchical addressing and why is it used? Give examples of protocols and techniques used in the PSTN and the Internet to demonstrate. VI. Draw two Ethernet networks interconnected by a router. Draw at least one host on each Ethernet network. Clearly show the protocol stack (in OSI terms) in use for all devices. Describe the services provided in each layer that are used in each device. VII. Describe at least two techniques that help prevent the 'count-to-infinity' problem in distance vector routing protocols. VIII. Draw a network of 4 data links connected serially by 3 routers. Draw a host on each far end of the data link networks. Assign IP addresses and DNS names where appropriate. Show how a traceroute from one host to another would function. Detail each packet transmission required. IX. Describe the 5 phases of inserting into a Token Ring network. X. Find and describe everything that's wrong with Figure 1.