Homework answers / question archive / This README file has 3 sections: 1

This README file has 3 sections: 1

Computer Science

Share With

---

This README file has 3 sections:

1. Building, testing and running file transfer applications 
2. An overview of input files
3. An overview of example protocol log output files

---------------------------------------------------------------------------
1. BUILDING, TESTING AND RUNNING FILE TRANSFER APPLICATIONS

The RFT application comprises client and server applications:
    rft_client
    rft_server

Just like the preliminary exercise, you will have to run the server in
one terminal window and run the client in another terminal window either
on the same host (in which case the server address will be 127.0.0.1) or
a different host (in which case the server address will be for an external
IP interface).

There are three ways to test and run the project.

1.1. Run the test programs

The simplest way to build and run the test programs is to enter the
following command:

    ./runtests.sh
    
this will compile test programs and run them. It uses the same test
framework as was used for assignment 1 and output is similar.

Enter:

    ./runtests.sh -h
    
for a usage message and how to test separate parts of the assignment.

Some tests will block if there are errors because they run two 
processes for the client and server and, if the client does not
successfully send data the server will block waiting.

1.2. Use run shell scripts to build and run the RFT applications

The following two shell scripts are provided as convenient ways to the
run the applications:

    run-server.sh to compile and run the rft_server
        (which also compiles the rft_client)
    
    run-client.sh to run the rft-client
    
By default, the run-server.sh script will compile the applications and
start the server on port 20333, and the run-client.sh script will start
the client and attempt to send the file in660.txt to the server for
output on the server side in the file ./out/out.txt. The normal transfer
protocol is used by default.

That is, if you do:

    ./run-server.sh 
    
in one terminal window, and:

    ./run-client.sh
    
in another terminal window on the same host, then provided the
rft_client_util library has been implemented, the client will transfer
the file in660.txt to the server using the send_file_normal function.

Both scripts have options to allow you to specify a different port
and/or server address and to transfer different files using either the
normal transfer protocol without retransmission or the protocol with
timeout and retransmission.

To see the server script help and usage message, enter the command:

    ./run-server.sh -h
    
To see the client script help and usage message, enter the command:

    ./run-client.sh -h
    
When a transfer is completed you can use the diff command to check that
the client-side input file is the same as the server-side output file.

For example, if the client-side input file is in660.txt and the
server-side output file is out/out.txt, the command:

    diff in660.txt ./out/out.txt
    
will compare the two files. See the man page for diff to interpret its
output.

TIP: if you want to test that you have correctly implemented reaching
the limit of consecutive retries, just run the client with input file
in660.txt with a loss probability of 0.5 or more without running the
server. That is, enter the command:

    ./run-client.sh 660 0.5
    
which should generate enough message loss to reach the limit of
consecutive retries (provided there is no server running and therefore
no ACKs at all).

When testing the transfer with timeout and retransmission use the
in660.txt file to ensure there are sufficient segments to transfer for
the percentage of lost segments to be close to the loss probability you
specify. Run a number of transfers with different probabilities. For
example, try loss probabilities less than 0.3 (tending to no segments
corrupted), 0.3 (for approx. 1/3 of segments corrupted), 0.5 (for
approx. 1/2 of segments corrupted) and close to 1.0 (tending to all
segments corrupted). The actual proportion of lost/corrupted segments is
random. Therefore, if you specify a loss probability of 0.5, you are
not likely to get exactly 50% of lost segments. The actual loss will
be in a range around 50%.

The higher the loss probability you specify the more likely it is that
the count of consecutive retries will exceed the maximum consecutive
retries allowed.

1.3. Run the rft_client and rft_server applications directly

You do not have to use the shell scripts. You can run the rft
applications directly.

To do this, enter the command:

    make 
    
to build the applications.

Then, assuming compilation succeeds, enter the command:

    ./rft_server 
    
to see the options for running the server application.

And:

    ./rft_client
    
to see the options for running the client application

---------------------------------------------------------------------------
2. AN OVERVIEW OF INPUT FILES

The following input files are provided for you:

    in0.txt
    in1.txt
    in17.txt
    in34.txt
    in35.txt
    in36.txt
    in350.txt
    in660.txt
    
The number that follows "in" in the filename is the size of the file in
bytes.

The sizes have been chosen deliberately to test the following file
transfer conditions:

    - zero sized file
    - boundary condition file sizes, e.g. one more than zero and sizes  
      one off the significant file data in a data segment payload
    - multiples of significant file data size
    - a larger file to test random data segment loss
    
When testing your solutions it is a good idea to try each of the input
files to check that the solution works correctly for the different file
sizes.

---------------------------------------------------------------------------
3. AN OVERVIEW OF EXAMPLE PROTOCOL LOG OUTPUT FILES

The following files provide examples of the log output from the
rft_client and rft_server:

    README-client-normal-tfr-out.txt:
        output from the client for successful execution of a normal 
        transfer
    README-server-normal-tfr-out.txt:
        output from the server for successful execution of a normal 
        transfer
        
    README-client-normal-tfr-err.txt:
        output from the client when there is an error in a normal 
        transfer

    README-client-timeout-tfr-out.txt
        output from the client for successful execution of a transfer 
        with timeout
    README-server-timeout-tfr-out.txt:
        output from the server for successful execution of a normal 
        transfer

    README-client-timeout-tfr-err.txt
        output from the client when there is an error in a transfer 
        with timeout

The above files are intended to give you an idea of the sort of output
to expect during protocol execution, provided you have used logging and
exit with error correctly.

---------------------------------------------------------------------------