Study with the several resources on Docsity
Earn points by helping other students or get them with a premium plan
Prepare for your exams
Study with the several resources on Docsity
Earn points to download
Earn points by helping other students or get them with a premium plan
A set of problems related to basic concepts and terms in networking, including autonomous systems, congestion control, cryptographic hash, digital signature, and more. The document also includes questions related to peer-to-peer systems and cryptography. The problems are not mandatory, but they are similar to those that may appear on the final exam. useful for students who want to test their knowledge and prepare for the final exam.
Typology: Exercises
1 / 19
This page cannot be seen from the preview
Don't miss anything!
You are not required to do these problems, and we are not asking you to hand them in. However, if you can’t do these problems, then you will have trouble on the final. In fact, many of these questions came from last year’s final. As the saying goes, “Know This”.
Consider the following concepts and terms: Autonomous Systems, Bellman-Ford, Congestion Control, Count-to-Infinity, Cryptographic Hash, Digital Signature, Dijkstra’s Algorithm, Duplicate Acknowledgements, End-to-End Principle, Fate Sharing, Flow Control, Hidden Terminal, Peer-to- Peer, Poisoned Reverse, Policy Oscillations, Priority Packet Scheduling, Slow Start, Public Key Cryptography, Symmetric Key Cryptography, Three-Way Handshake. Match these terms to the descriptions, using each term exactly once. (a) A design style that involves many equivalent nodes, rather than a few specialized servers. Peer- to-Peer (b) Something that can happen in BGP that results in unstable routing tables. Policy Oscilations (c) The problem in wireless networking of a sender not being able to detect if its transmissions will collide with those of another node because the sender cannot itself hear the other node’s transmissions, even though the receiver can. Hidden Terminal (d) A technique in distance-vector routing that prevents some cases of looping by not allowing a node A to advertise a route to node B if A would forward packets to node B as the first hop of that route. Poisoned Reverse (e) If the technique in item (d) above is not applied, this phenomenon can occur, which then requires many iterations of the routing protocol before a routing loop is removed. Count-to-Infinity (f) The mechanism used by TCP to open up the congestion window quickly, and is used only if the connection is not operating in Congestion Avoidance. Slow Start (g) A class of encryption algorithms that require use of a shared secret key. Symmetric Key Cryptography (h) A class of encryption algorithms that allow a host to make an encryption key widely known while remaining the only entity that can decrypt messages encrypted with this key. Public Key Cryptography (i) The mechanism used by TCP to reliably establish a connection. Three-Way Handshake (j) The entities that BGP provides routes between; BGP describes its paths in terms of a series of these. Autonomous Systems (k) A cryptographic primitive that ensures that the particular piece of data did come from a particular source (or equivalently, prevents a source from repudiating that it sent that message). Digital Signature
i. Five Basic Design Decisions: As we discussed in class, the Internet architecture was shaped by five basic design decisions. Please list the two in the following list that are NOT among these five decisions: b, f (a) Layering (b) Longest prefix match (c) Best-effort service (d) The end-to-end principle and fate sharing (e) A single universal internetworking layer (f) Sliding window flow control (g) Packet switching ii. ARP: A typical ARP exchange goes as follows: (a) Initiating host sends: ARP request (b) Responding host sends: ARP response Which of these messages are broadcast? a iii. Netmask: Which of the following methods are ways a host can learn the netmask for the subnet? a, d (a) Configuration (b) ICMP (c) ARP (d) DHCP (e) NAT iv. ICMP: Which of the following are valid ICMP messages? a, c, e, g (a) Need Fragmentation (b) OMG (c) Source Quench (d) Invalid Address Format (e) Host Unreachable (g) TTL Expired Which ICMP message (in the list above) is used in Traceroute? g Which ICMP message (in the list above) is used to discover path MTU? a
v. Peer-to-Peer: P2P systems typically do some combination of three tasks, searching (e.g., keyword search), lookup (mapping name to location), and download. Which approach is often used for download? c (a) Some form of flooding (b) Distributed Hash Tables (c) Chunking Which approach is typically used for search? a (a) Some form of flooding (b) Distributed Hash Tables (c) Chunking Which approach is typically used for lookup? b (a) Some form of flooding (b) Distributed Hash Tables (c) Chunking Which factor is most responsible for making chunking advantageous? b (a) Number of participating peers (b) Asymmetry of bandwidth (downloading at higher rate than uploading) (c) Lack of centralized control (d) Self-scaling vi. Cryptography: Which one of the following is an easy way for host A to use public key cryptography to authenticate host B? d (a) Encrypt B’s public key with A’s private key (b) Ask B to encrypt A’s private key with B’s public key (c) Ask B to encrypt a nonce using B’s public key (d) Ask B to decrypt a nonce that has been encrypted with B’s public key vii. Security: What are the 3 security goals (as described by the esteemed Professor Paxson)? b (a) Cryptography, Isolation, Authentication (b) Confidentiality, Integrity, Availability (c) Cash, Infamy, Awe (d) Consistency, Isolation, Availability (e) Confidentiality, Integrity, Authentication (f) Credibility, Intrusion, Anonymity
i. TCP Throughput: Consider two TCP connections whose throughput obeys the TCP throughput equation. The first TCP connection has the following parameters: MSS = 1000 bytes, RTT = .2msec, drop rate = .5% The second TCP connection has the following parameters: MSS = 500 bytes, RTT = .8msec, drop rate = 2%. What is the ratio of throughputs (the throughput of the first TCP connection divided by the throughput of the second TCP connection)? 16 ii. Fair Shares: Consider a shared link L with five connections. Each connection is limited, by its own access link (which it uses to reach the shared link L), to the following bandwidths: Flow 1: 1Gbps Flow 2: 2Gbps Flow 3: 3Gbps Flow 4: 4Gbps Flow 5: 5Gbps If the shared link L has capacity C=10Gbps, what are the fair shares? 1, 2, 7/3, 7/3, 7/ For which values of C (the capacity of the shared link) do flows 2 and 3 have the same fair shares? C ≤ 9 [express your answer as bounding inequalities on C] For which values of C (the capacity of the shared link) do flows 4 and 5 have the same fair shares? C ≤ 14 [express your answer as bounding inequalities on C] For which value(s) of C will flow 3 get 2.5Gbps? 10. If a particular flow gets less than its access bandwidth, does any flow receive more bandwidth than that flow? No iii. Transfer times: You are trying to transfer the contents of a 1.25terabyte disk drive between here and New York, and have at your disposal two methods: (a) sending the data over a 100mbps link or (b) sending the drive by Federal Express (with a guarantee that it will arrive in 24 hours). Assume the network charges 10-^10 cents per bit transmitted, and Federal Express charges $30 for the package. Which is faster? Federal Express Which is cheaper? Network iv. Headers: You are accessing a web site using your browser, from a host that is connected to an Ethernet within Soda Hall. A packet sniffer on the Soda Hall Ethernet captures a packet from your web session, which has TCP, IP, HTTP and Ethernet headers: starting from the outermost header (the header with bits at the very front of the packet), what is the order of the headers you need to traverse before reaching the payload? Ethernet, IP, TCP, HTTP
v. Public Key Encryption: You are releasing your Ph.D. thesis to the world, and want to make sure that everyone knows that it came from you, rather than from some imposter posing as you. Would you encrypt the file with your private key or with your public key? Private vi. IP Multicast: Consider the (a) DVMRP and (b) CBT multicast routing designs. When the first packet is sent to a multicast group G, in which design is that packet broadcast? DVMRP vii. Ethernet: Consider three hosts H1, H2, and H3 on an Ethernet; in the first seven time slots the following transmissions take place: Slot 1: H1, H2 transmit Slot 2: H1 transmits Slot 3: H2, H3 transmit Slot 4: idle Slot 5: H1, H2, H3 transmit Slot 6: H1, H3 transmit Slot 7: H2 transmits What is the backoff counter of the three hosts? (backoff counter starts at zero, is incremented to one after first backoff, etc.) H1: 2 H2: 0 H3: 3
Consider the layer 2 network above connecting the switches numbered 1 through 8 (for the purposes of the spanning tree protocol, these numbers are the switch IDs). Switches 4, 8, 5, and 7 have local Ethernets that each have two hosts on them (Hosts A through H). Recall that, when constructing the spanning tree, (i) the root is the bridge with the lowest ID, and (ii) when there is more than one shortest path to the root, the path whose first hop goes through the bridge with the lower ID is chosen. Also, each link is considered to have a length of one when computing path lengths. i. Compute the spanning tree for this network. List the links in the spanning tree (denote, for example, a link between nodes 1 and 4 as 1-4, etc.). 4 - 8, 4- 1 , 1-6, 1-3, 3-2, 3- 5 , 2- 7 ii. Assume that the switches are “self-learning” and consider the following set of transmissions: (a) host A to host H (b) host H to host E (c) host D to host F (d) host E to host A (e) host C to host D (f) host E to host H Which of these transmissions are NOT broadcast to all nodes? (d) (e) (f) ___ Which of these transmissions are dropped by their first hop switch?_ (e) _____ iii. In transmission (f), what path (in terms of the node IDs traversed) does the packet take from host E to host H? 8 - 4 - 1 - 3 - 2 - 7 _ 4 8 6 2 7 (^1 3 ) Host A Host B Host C^ Host D Host E Host F (^) Host G Host H
We consider three collision resolution schemes:
Consider the network in the diagram below, where a newly arrived laptop has just been plugged into the Ethernet. The laptop is trying to reach a file server that is located on the same Ethernet (as is the DHCP server), but the DNS server is on a remote network. Below are the initial messages sent and received by the laptop (no other messages are shown, and assume these packets are sniffed on the Ethernet to which the laptop is attached). The MAC addresses are shortened, for convenience, and the laptop begins without an IP address, but is later assigned one (4.9.8.22) by DHCP. The following messages are sent, in some order. Below they are listed in alphabetical order, but your job is to list them in order.
i. In the table on the following page, list the messages in the correct order and fill in the source and destination MAC and IP addresses (when applicable).
When you connect to an FTP server, you are actually making two connections. First, the so-called control connection is established, over which FTP commands and their replies are transferred. Then, in order to transfer a file or a directory listing, the client sends a particular command over the control connection to establish the data connection. The data connection can be established two different ways, using active mode or passive mode. (a) Passive Mode : the client sends the PASV command to the server, and the server responds with an address. The client then issues a command to transfer a file or to get a directory listing, and establishes a secondary connection to the address returned by the server. (b) Active Mode : the client opens a socket on the local machine and tells its address to the server using the PORT command. Once the client issues a command to transfer a file or listing, the server will connect to the address provided by the client. In both cases, the actual file or listing is then transferred over the data connection. Consider a client behind a NAT box trying to use FTP using one of these two modes. Assume the NAT has not been modified to understand the semantics of FTP. Does the client’s FTP attempt succeed using Passive Mode? (yes/no) Yes. Does the client’s FTP attempt succeed using Active Mode? (yes/no) No.
The nodes in the graph below construct a spanning tree using the standard spanning tree protocol. The bridges (or switches), Bi, have MAC address i. After the spanning tree protocol is run, which links are in the spanning tree? L1, L2, L5, L6, L9, L i. Assume node B1 fails, and the bridges reform the spanning tree. Which links are now in the spanning tree? L4, L7, L9, L10, L iii. Assume that soon after B1 fails, B4 and B7 also fail. How many spanning trees are formed, and what are their roots? 2 spanning trees, with roots B2 and B Reminder: (i) the root is the bridge with the lowest ID, and (ii) when there is more than one shortest path to the root, the path whose first hop goes through the bridge with the lower ID is chosen. Also, each link is considered to have a length of one when computing path lengths.
Consider the network below, with each link having the same cost (in the metric used in unicast routing). The members of the multicast group are denoted by m1, m2, m3, and routers in the network are denoted by A, B, C, D, E, F. (i) Assume that CBT multicast routing is used, with the center (core) located as marked in the network. Describe network paths by the series of routers the packets pass through. What path do packets sent by M1 to the group take to reach M2? M1-‐A-‐B-‐C-‐E-‐M What path do packets sent by M3 to the group take to reach M2? M3-‐D-‐C-‐E-‐M 2 (ii) Assume that single-source multicast trees are used, and ignore the node labeled as core. What path do packets sent by M1 to the group take to reach M2? M1-‐A-‐F-‐E-‐M What path do packets sent by M3 to the group take to reach M2? M3-‐D-‐E-‐M 2
