CS422: Fall2005 Final Exam

Name:___________________________________________

Part 1. Answer True or False (1 point each)

_F_ Distance Vector Routing converges faster than Link State Routing.
_F_ An ICMP message with a wrong destination will generate another ICMP message.
_T_ A Transport protocol may be unreliable
_T_ Four packets are used to shut down a connection in TCP.
_T_ NAT breaks protocol layering.
_F_ DHCP stands for Dynamic Host Control Protocol
_F_ In DNS recursive resolution if a server do not know the answer it will return the address of another DNS server
_T_ Binary files attached to an e-mail are translated to text before being sent using SMTP.
_T_ RTP - Real Time Protocol -  does not offer a quality of service guarentees.
_T_ A packet fragment can be fragmented.

Part 2. Short Questions: (4 points each)

1. Explain why Hosts (and not only routers) have routing tables.

A network may have more than one router, so the host may have more than one router to choose from.

The host may have more than one interfaces.

2. Explain why the subnet mask is necessary in a host.

The subnet mask is necessary so that a host can know if it can send a packet to another host directly or if it needs to be routed.

3. Explain when ICMP Source Quench is produced.

It is produced by a router that is congested when packets are dropped.

4. Explain in what situations traceroute will give an incorrect answer.

If the packets follow two different routes or if the routes change in the middle of a probe, traceroute will give incorrect answers.

5. Explain why TCP is not used for VoIP

Because it may add extra delay due to retransmissions.

6. Write the equations used to estimate the average round trip time and variance in TCP Adaptive Retransmission.

RTTnew = .125 * R(i.e., new sample) + .875 * RTTold

RTTVARnew = .25 * abs(R-RTTnew) + .75 * RTTVARold

7. Explain the steps follow during Slow Start and Congestion Avoidance.

Slow start

- set congestion window to the size of one segment.

- The sender sends that many bytes and wait for an ack.

- If act goes throw, double the size of congestion window until ½ of original window is reached.

Congestion avoidance

- Increase congestion window by one segment for every successful ack until complete window is reached.

8. Explain the TCP FAst Retransmission mechanism.

When there are two consecutive acknowledgements for the same sequence number that is a sign that a packet was lost. TCP immediately retransmits without waiting for the retransmission timer to expire.

9. Write down the TCP Header format and its fields. Show also the size of each of the fields.

Refer to the lecture slide.

10. Assume that a TCP connection with parameters <src-ip-addr, src-port,dest-ip-addr, dest-port> is initiated by a host behind a NAT box with IP address IPnat. Explain how NAT works using these parameters.

• Machine uses NAT box as a router.

• Router changes header to <IPnat,PORTrand,IPdest,PORTdest>

• NAT adds mapping <PORTrand,IPsrc,PORTsrc> to table

• When a packet <IPdest, PORTdest, IPnat, PORTrand> comes back, it will change it to <IPdest, PORTdest, IPsrc, PORTsrc>

11. Ac omputer in an ethernet network sends a TCP packet using VPN on TCP. Give the list of protocol headers in order  used by the TCP packet in the ethernet network. Mention tonly the protocol names. If a header protocol is used twice, mention the protocol header twice.

Ethernet

IP

TCP

IP

TCP

12. What the DNS types are and explain why the DNS type is important in the DNS query.

• Type A – address

• Type Mx – mail exchange

• Type CNAME – computer name

- The same name may translate to different IP addresses based on the type.

13. List the 4 messages used in DHCP.

DHCP discover

DHCP offer

DHCP request

DHCP acknowledge 

14. Explain why routers do not do reassembly?

Because fragments of the same packet may follow different routes. 

15. Explain what application uses the TCP urgent pointer and how the application uses it.

Telnet uses the TCP urgent pointer to pass ctrl-c that will skip the output of the program.

16. List the limitations of IP when it is used to send streamed multimedia.

packet loss

end to end delay

packet jitter

17. Explain what is XDR and what it is used for.

XDR – External data representation

It is a common representation used to send arguments or results in an RPC through

the network to be able to communicate client and servers on machines with different data representations.

18. Explain what is a Proxy Object in RMI.

A proxy object is a local representation of the remote object. Every method invocation on the proxy object

is forwarded to the remote object. The methods of the proxy object encode/decode the argument and send it to the server.

19. Explain how Adaptive Playout Delay works.
It estimates the variance and average of the network delay.

It adjusts the playout accordingly,

Playout Delay = (Net Delay avg + 4Net Delay Var)

It adjusts accordingly at the beginning of each talk spurt.



20.Explain what is J2EE

J2EE – Java 2 Enterprise Edition

It is an application server architecture that describes the client/server interaction.

Part 3. Project Questions: (5 points each)

<>

21.  In the IP telephone explain the mechanism used to know when a new voice packet can be formed.

There is a set of markers in the capture buffer set by the program.

When the capture reaches one of the markers , an event is generated and the callback is executed.

22. In the IP telephone explain when the player starts playing the data received.

It starts playing the data received when the jitter buffer has enough data to account for the variability of the network delay.