Homework Assignment II Solution

1. Definitions (10 pt. 1 pt. each)

 

Please read Chapter 3 and define following terminologies briefly.

 

•   Logical Communication

 

•   Multiplexing and Demultiplexing

 

•   Little Endian

 

•   Handshaking

 

•   Flow Control

 

•   Congestion Control

 

•   Go-Back-N algorithm

 

•   Selective Repeat algorithm

 

•   Fairness

•   Slow Start

 

 

2. TCP Reliable Data Transfer (10 pts)

 Consider two nodes which are connected by an 8Mbps link (assume 1Mbps = 1000Kbps)

and RTT is 0.05 sec. Assume the size of each packet is 8K bits.

 

Answer the following questions for ARQ schemes:

(a) Assume  that  the  link  is  error-free:  what  is  the  maximum  possible  rate  of transmission for Stop-and-wait, GBN, and SR respectively? Why? (6 pts)

 

(b) For GBN, in order to allow sender to continuously send packets without any waiting, what is the minimum window size in terms of the number of packets?

(1 pt)

 

 

(c) Suppose that we transmit 20 packets with sequence number from 1 to 20. The packet  with  sequence  number  16  is  lost  and  all  other  packets  are  received correctly. Assuming there is no ACK lost, for stop-and-wait, GBN, and SR, which packets have been retransmitted? (3 pt)

 

3. TCP Sequence Numbers (10 pts)

 

Host A and B are communicating over a TCP connection, and Host B has already received from A all bytes up through byte 168. Suppose that Host A then sends two segments to Host B back-to-back. The first and second segments contain 20 and 40 bytes of data, respectively. In the first segment, the sequence number is 169, source portnumber is 303, and the destination port number is 80. Host B sends an acknowledgement whenever it receives a segment from Host A.

 

a)  In the second segment sent from Host A to B, what is the sequence number, source port number, and destination port number? (3pts)

 

b)  If the first segment arrives before the second segment, in the acknowledgement of the first arriving segment, what is the acknowledgment number, the source port number, and the destination port number? (3pts)

 

c)  If the second segment arrives before the first segment (out of order arrival), in the acknowledgement of the first arriving segment, what is the acknowledgment number? (1pt)

 

 

 

d) Suppose the two segments sent by A arrive in order at B. The first acknowledgement is lost and the second acknowledgement arrives after the first timeout  interval,  as  shown  in  the  figure  below.  Please  provide  the  sequence number  for  the  third  (retransmitted)  data  segment  (1  pt)  and  provide  the

acknowledgement number for the 2nd and 3rd acknowledgement. (2pts)

 

 

4. Hands-on Practice I: UDP (10 pts)

In this practice, we’ll take a quick look at the UDP transport protocol. As we saw in

Chapter 3, UDP is a connectionless non-thrills protocol. Start capturing packets in Wireshark and then do something that will cause your host to send and receive several UDP packets.  For example, use telnet [domain name].

 

Please answer the following question

 

(a) Select one packet. From this packet, determine how many fields there are in the UDP header. (Do not look in the textbook! Answer these questions directly from what you observe in the packet trace.) Name these fields. (2 pts)

 

(b) What is the maximum number of bytes that can be included in a UDP payload? (1pt)

 

 (c) What is the protocol number for UDP? Give your answer in both hexadecimal and decimal notation. (To answer this question, you’ll need to look into the IP header.) (2pts

 

(d) Search “UDP” in Google and determine the fields over which the UDP checksum is calculated. Capture a VERY SMALL UDP packet. Manually verify the checksum in this packet. Show all work and explain all steps. (5 pts)  You need to paste a screenshot of the UDP packet content as the evidence.

 

 

5. Hands-on Practice II: TCP (10 pts)

 

In this practice, we’ll investigate the behavior of TCP in detail. Before beginning our exploration of TCP, we’ll need to use Wireshark to obtain a packet trace of the TCP. You’ll do so by accessing:

 

http://www-users.cs.umn.edu/~tianhe/csci4211/HTTP-2.htm

 

 

 

First, filter the packets displayed in the Wireshark window by entering “tcp” into the display filter window towards the top of the Wireshark window. What you should see is series of TCP and HTTP messages between your computer and

 

Please answer the following question (along with screenshots as the evidence of your answer)

 

(a) What is the IP address and TCP port number used by the client computer (source) that downloads the bill of rights from www-users.cselabs.umn.edu?  You need to paste an appropriate screenshot as the evidence. (2pts)

 

 

 

Please answer the following question (along with screenshots as the back-up evidence of your answer, if applicable)

 

(b) What is the sequence number of the TCP SYN segment that is used to initiate the TCP connection between the client computer and the webserver? What is it in the segment that identifies the segment as a SYN segment? (2pts)

 

 (c) What is the minimum amount of available buffer space advertised at the received for the entire trace? Does the lack of receiver buffer space ever throttle the sender? (2pts)

 

 

(d)  Are there any retransmitted segments in the trace file? What did you check for in order to answer this question? (2pts) You need to paste a screenshot as the evidence.

 (e) Based on the sequence number, you can calculate how many bytes (TCP payload) have been transferred through this TCP connection. Does the number of total bytes downloaded through this TCP connection equal to the html file size of the bill of rights? Explain why? (2 pts)

 

 

6. Programming assignment: Network performance measurement via socket programming (50 pts)

This  problem  requires  building  server  and  client  applications  in  C  which  exchange packets over UDP and TCP. Using these applications, you will measure the performance of the network path between two hosts.

 

TCP performance measurement:

Build   server   and   client   applications   that   communicate   over   TCP   sockets.   All

measurement must be performed on the client. Read and understand the questions prior to implementation.

 

Run both the client and the server on your home LAN for parts (a) & (b).

 

a.   Using Wireshark, observe the TCP connection establishment process (namely, three way handshake). Repeat 5 times and provide the average time taken for connection establishment. Ignore the time for the last ACK in three way handshaking to reach the server, as this cannot be measured from the client side. Simply put, the establishment time is defined as the duration from when "synchronize" (SYN) segment (first segment) was sent out from the client until when ACK (third segment) was sent out from the client (in response to SYN &ACK segment (second segment) from the server. Also please provide a screenshot of Wireshark showing TCP connection establishment. (5pts)

 

 

b.   Using Wireshark, observe the TCP connection termination process, initiated by the client. Repeat 5 times and provide the average time taken for connection termination. Ignore the time for the last ACK to reach the server, as this cannot be measured from the client side. Simply put, the establishment time is defined as the duration from when FIN segment (First segment) was sent out from the client until when ACK (fourth segment) was sent out from the client in response to FIN segment (third segment) from the server. Also please provide a screenshot of

Wireshark showing TCP connection termination. (5pts)

 

 

Now, run the client and the server on the  CSELabs UNIX machines: The server can be on one of the machines in  Keller Hall 4-250 and the client can be on one of the machines in Lind Hall 40.

 

c.   Show the network path between the server and the client. Note that they need to be at least 2 hops apart. Please provide a screenshot of the traceroute command on Linux as well as a table including all the IP addresses on the path (including the client and the server) (5 pts)

 

 

Implement the message communication protocol (which is described in the last portion of the handout) on the TCP based server and client. The client should download a file from the server using the message protocol. You must submit the code for this part of the assignment.

 

d.   The client should measure and report the time taken to download the file from the server.

Repeat this step for different values of BUF_SZ (256B, 512B, 1024B & 1536B)

and the three input files.

NOTE: The time measurement should only include the time taken for recv() and should not include any other system/function calls (like time for I/O). You may use gettimeofday() to measure the time.

 Fill in the values in the table for each input file. You may consider only messages of type MSG_TYPE_RESP_GET for computation. (10 pts)

 

 

UDP performance measurement:

Build   server   and   client   applications   that   communicate   over   UDP   sockets.   All measurement must be performed on the client. Read and understand the questions prior to implementation. Run the client and the server on the  CSELabs UNIX machines: The server can be on one of the machines in  Keller Hall 4-250 and the client can be on one of the machines in  Lind Hall 40.

 

e.   Use your own program to measure RTT (Round Trip Time) of UDP. Send 10

UDP datagrams from the client to server. The server should reply with a UDP datagram whenever it receives a datagram from the client. RTT can be obtained by computing the difference in time from when a datagram was sent out from client until the time when the corresponding reply was received by the client. Please provide the average, minimum and standard deviation of RTT and a screenshot of 10 UDP exchanges. (5pts)

 

 

Implement the message communication protocol (which is described in the last portion of the handout) on the UDP based server and client. The client should download a file from the server using the message protocol. You must submit the code for this part of the assignment.

 

f.    The client should measure and report the time taken to download the file from the server. Repeat this step for different values of BUF_SZ (256B, 512B, 1024B &

1536B) and the three input files.

NOTE: The time measurement should only include the time taken for recv() and should not include any other system/function calls (like time for I/O). You may use gettimeofday() to measure the time.

Present your results in a tabular form, like you did for part (d) above. Compare the timing results with the results you obtained for the TCP case in part (d). Explain the difference in delay. (10 pts)

 

Message communication protocol:

●   Servers  and  clients  using  this  protocol  can  communicate  with  each  other  by

sending messages. Each message is a C structure of type struct   msg_t.  The servers and clients can only exchange messages in this format. (This means that the buffer passed in the send() and recv() system calls is always a struct    msg_t).

 

#define    BUF_SZ   1024 struct    msg_t   {

               enum    msg_type_t   msg_type;                        /*    message   type   */

               int     cur_seq;                                                                                            /*    current    seq   

number    */

               int    max_seq;                                                                          /*    max   seq   number     

*/

               int     payload_len;                                                                       /*    length     of   

payload    */

               unsigned     char    payload[BUF_SZ];     /*     buffer     for    data    

*/

};

 

●   Some details about struct   msg_t:

a.   msg_type: This indicates the type of the message. Five message types have been predefined.

b.   cur_seq,  max_seq: If a set of related messages have to be transmitted, then cur_seq indicates the sequence number of the current message and max_seq indicates the total number of messages in this sequence

c.   payload: This is a buffer of size BUF_SZ    (1024B) which you can use to

fill in any kind of data.

d.   payload_len:  This  will  indicate  the  size/length  of  valid  data  in  the buffer.

 

●   Message types explained:

a.   MSG_TYPE_GET: Use this message to request a file for download

b.   MSG_TYPE_GET_ERR: Use this message to indicate errors in obtaining the file (if any)

c.   MSG_TYPE_GET_RESP:  Use  this  message  to  send  the  file  across  the network

d.   MSG_TYPE_GET_ACK:     Use     this     message     to     acknowledge     a

MSG_TYPE_GET_RESP message.

e.   MSG_TYPE_FINISH: Use this message to indicate end of session.

 

 

Requirements: 1.   The TCP server executable must be named server_tcp    and    the UDP server executable must be named  server_udp.

2.   The TCP client executable must be named client_tcp  and   the UDP client executable must be named client_udp.

3.   Provide two Makefiles, one in the server directory and one in the client directory, which can build these executables.

4.   Please note that we will try to download files of different sizes while grading - it is your job to ensure that files of any size can be downloaded. You may test your programs against the provided input files (input_small.txt, input_medium.txt, input_large.txt)

5.   The submission must include a README file which clearly contains:

a.   Name of the student, student ID and x500

b.   A brief description of how files are downloaded.

 

 

Server requirements:

The server program will be executed as follows:

$    ./server_tcp    <port

$    ./server_udp    <port

1.   The server must listen for clients on a socket bound to the specified port.

2.   Print the following message on the screen whenever a message is received:

server:    RX   <msg_type    <cur_seq    <max_seq    <payload_len

3.   The server must send a file to the client when the client requests it.

4.   If the requested file is not found in the current working directory, then the server must respond to the client with an appropriate error message

(hint: message type - MSG_TYPE_GET_ERR)

 

Client requirements:

The client program will be executed as follows:

$    ./client_tcp   <server-­‐ip    <port   <filename

$    ./client_udp   <server-­‐ip    <port   <filename

1.   The client must first connect to the specified server on the specified port.

2.   The client must then attempt to download the specified file from the server and save it in the current working directory.

3. File integrity must be preserved when downloading files. This means that downloaded file must exactly match the file on the server. (Hint: Use the diff utility  to  compare  the  two  files).  You  will  lose  significant  points  if  the downloaded file differs from the file on the server.

4.   Print the following message on the screen whenever a message is received:

client:    RX   <msg_type    <cur_seq    <max_seq    <payload_len

 

Sample server:

A sample TCP server application has been provided. The server will allow a client to connect and download a file. You may test your client with this server initially. However, you are required to develop and submit your own servers and clients for this assignment. The sample server works as follows:

1.   Start the TCP server

2.   Start the TCP client & connect to the server.

3.   Send a MSG_TYPE_GET  from  the client to the server, with the name of the file to be download in the payload.

4.   If the server cannot find the file, it will respond with a MSG_TYPE_GET_ERR

message. If the file is found, the server will break the file into chunks and transmit each chunk in a MSG_TYPE_GET_RESP message. Depending on the file size, there will be multiple such messages. The actual file contents will be stored in the payload in each message.

5. The  client  must  send  a  MSG_TYPE_GET_ACK message  for  each MSG_TYPE_GET_RESP it receives - or else the server will not respond with the next message.

 

You can use the same mechanism or a different mechanism to download the files. The

README file should clearly explain the mechanism you use to download the file.

 

 

 

Execution environment:

1.   Please  ensure  that  your  code  compiles  and  executes  on  the   CSELabs  UNIX

machines.   You   will   lose   significant   points   if   your   code   cannot   be compiled/executed on these machines.

2.   While developing/implementing your solution, the server and client can run on the same machine - You can use the IP address as localhost or 127.0.0.1.

3.   When  it  comes  to  actual  measurement,  the  server  and  client  should  run  on different machines, ideally multiple hops away. Our suggestion is to use two CSE

Lab machines as described below:

●   Run the server on one of the machines in Keller Hall 4-250.

●   Run the client on one of the machines in Lind Hall 40.

●   Please use only port numbers between 9000 and 10000. (These ports are

currently allowed by the system staff).

●   You can use ‘ip addr show’ or ‘ipconfig’ to get the IP address of the

machine on which the server is running.

If your code does not execute in this scenario, you will lose significant points on

your submission.

 

Deliverables:

1.   You must upload a single archive file (.zip or .tar or .tar.gz) on moodle.

When extracted, the archive file must be a single folder. The name of the folder should be your student ID. The folder should contain the following files:

●   Readme

●   server source files & Makefile

●   client source files & Makefile

●   message.h

●   MS Word/PDF document

2.   DO   NOT   include   the   test   files   (input_small.txt,   input_medium.txt   &

input_large.txt)

3.   DO NOT include any executable files - we will build the executables using your

Makefiles.

 

For example, here is a sample submission:

1234567/

               Readme

               PA2.doc

               message.h             

               server/

                           server_tcp.c

        server_udp.c

        Makefile

               client/

                           client_tcp.c

                            client_udp.c

                           Makefile

You can create an archive file from the contents of the above directory as follows:

$    tar    cvf    1234567.tar.gz    1234567/

 

Grading:

README, Makefiles, comments, readability: 4 points

Packaging the submission as specified: 3 points

Logging messages on the screen in the correct format: 1 points

File download: 2 points