CSCI 278        Computer Networks               Final Exam                   Fall 2004

Instructor: Seung Bae Im

Closed Book, Closed Notes. Do not write on the back of pages!!

Total points are more than 100--Answer all questions!

                                                            Name:_______________________________

1. (4 points each) Mark your answers on this page.

 

(a)    Given the following IP address in binary and the subnet mask of 255.255.255.224, find out the subnet’s Id(subnet’s address) that this station belongs to:

 

 10000100   11110001   10011110   00110000 à Subnet Id = ____________________________

 

(b)    For a subnetting scheme which does not allow all 0’s subnet and all 1’s subnet addresses, choose all the possible values for subnet masks.

____  128

____  192

____  224

____  240

____  248

____  252

____  254

____  255

 

(c) What is the advantage of 4B/5B code over Manchester code?

 

 

 

 

 

(d) Describe “Auto-negotiation” in Ethernet.

 

 

 

 

 

 

 

(e) A host with IP address 132.241.158.70 with the subnet mask of 255.255.255.224 sends a message to a host with IP address 132.241.158.102. Does the message travel through any router?

     ____ Yes

     ____ No

 

(f) We want to combine 8 Class C addresses that starts with 210.1.16.0. Which one should be the “supernet mask”?

     ____ 255.255.192.0

     ____ 255.255.224.0

     ____ 255.255.252.0

     ____ 255.255.240.0

     ____ 255.255.255.0

     ____ None of the above

 

(g) A subnet mask of 255.255.240.0 allows how many hostids per subnet?

     ____ 29 – 2

     ____ 210 – 2

     ____ 211 – 2

____ 212 – 2

____ 213 – 2

     ____ None of the above

 

(h) Which organization controls DNS, IP addresses, Port numbers, etc. ?

     ____ Internet Society(ISOC)

     ____ IETF(Internet Engineering Task Force)

     ____ IEEE

     ____ ICANN

     ____ IEC   

 

2.  4 point each unless otherwise noted.

 

(a)    Describe why we need port numbers in TCP(or UDP) protocols.

 

 

 

 

(b)   What is “supernetting” and why do we need it”?

 

 

 

 

 

(c)  What is the “Closed loop” problem for networks using bridges and switches? What is the solution?

 

 

 

 

 

 

 

(d)    What is “winsock”?

 

 

 

 

(e)    Describe why we have a problem with TCP’s connection termination even with the 3 way handshaking.

 

 

 

 

 

 

(f)     State main reasons(at least three) why Ethernet dominates through the generations of LAN protocols.

 

 

 

 

 

(g)    (12 points) We say that “client-server computing” is achieved using TCP (or UDP) socket-to-socket communication. Describe the following terms:

1)      well-known port numbers

 

 

 

2)      ephemeral port numbers

 

 

 

 

3)      a socket

 

 

 

 

4)      What makes each connection uniquely identified?

 

 

 

 

(h) (14 points) Complete the following table for the comparison of Connection-oriented service  versus Conncectionless service:

 

ISSUES   

Connection-oriented 

Connectionless

Initial setup—

Required/not required

 

 

Connection termination—

Required/Not required

 

 

Packet sequence—

In sequence guaranteed/Not guaranteed

 

 

Error control—

Done/Not done

 

 

Flow control—

Done/Not done

 

 

Layer overhead—

High/Low

 

 

Example protocols

 

 

 

 

 

 

3. (2 points for each fill in blanks & 4 points each for other questions)

 

(a)    Fill in the blanks: DNS name resolution works as follows:

A name resolver on a computer sends a DNS request(=www.google.com) to            

                  (#1)______________________

The (#1), looks in its cache to find an entry since (#1) is a ______________________

If (#1) does not find an answer and assuming (#1) was rebooted just few seconds ago, it sends

 a query to (#2)_____________________ server

The (#2) will respond with the IP addresses of (#3)_______________ servers

The (#1) then sends a query to (#3), then (#3) will respond with IP address of

                  (#4)_____________________ server

The (#1) then sends a query to (#4), then (#4) responds with the IP address of

                              ______________________

The (#1) then caches it and sends DNS reply to the resolver

Now, if another computer sends a DNS request(=www.google.com) to (#1), then (#1) searches its cache and finds the entry for www.google.com and replies but the reply is marked as       ___________________________

 

(b)    In the routing protocol OSPF, what is the function of a “Area-Border router”?

 

 

 

 

(c)    In “Link state routing” algorithm, a router sends its ____________________ to

_______________________________ . In “Distance Vector routing” algorithm, a router sends its

____________________________ to _____________________________.

 

(d) What is MTU(Maximum Transmission Unit)?

 

 

 

 

 (e) What is a “Multi-homed device”?

 

 

 

 (f) The current version of IP, called as ____________, has the problem of the address space running out. The current solutions include ____________________, _____________________ but ultimately the problem can be solved when the next generation, called as ___________, is adopted.

 (g) When an IP datagram is sent, it is sent in one of the following three modes:

            (1) Unicast, (2) _________________, or (3) _________________

 

 (h) The routing protocol, RIP(Routing Information Protocol), is simple but it has the following three problems:

            (1) ________________________________________________________

            (2) ________________________________________________________

            (3) ________________________________________________________