Simple protocol developed for the course Distributed Programming I of the first year of Master in Computer Engineering of Politecnico di Torino.
Develop a TCP sequential server (listening to the port specified as the first parameter of the command line, as a decimal integer) that, after having established a TCP connection with a client, accepts file transfer requests from the client and sends the requested files back to the client, following the protocol specified below. The files available for being sent by the server are the ones accessible in the server file system from the working directory of the server. Develop a client that can connect to a TCP server (to the address and port number specified as first and second command-line parameters, respectively). After having established the connection, the client requests the transfer of the files whose names are specified on the command line as third and subsequent parameters, and stores them locally in its working directory. After having transferred and saved locally a file, the client must print a message to the standard output about the performed file transfer, including the file name, followed by the file size (in bytes, as a decimal number) and timestamp of last modification (as a decimal number).
The protocol for file transfer works as follows: to request a file the client sends to the server the three ASCII characters “GET” followed by the ASCII space character and the ASCII characters of the file name, terminated by the ASCII carriage return (CR) and line feed (LF):
- - - - - - - - - - - - - - - - - - - -
| G | E | T | | filename... | CR | LF |
- - - - - - - - - - - - - - - - - - - -
(Note: the command includes a total of 6 characters plus the characters of the file name). The server replies by sending:
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
| + | O | K | B1 | B2 | B3 | B4 | T1 | T2 | T3 | T4 | filecontent... |
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Note that this message is composed of 5 characters followed by the number of bytes of the requested file (a 32-bit unsigned integer in network byte order - bytes B1 B2 B3 B4 in the figure), then by the timestamp of the last file modification (Unix time, i.e. number of seconds since the start of epoch, represented as a 32-bit unsigned integer in network byte order -bytes T1 T2 T3 T4 in the figure) and then by the bytes of the requested file.
The client can request more files using the same TCP connection, by sending many GET commands, one after the other. When it intends to terminate the communication it sends:
- - - - - - - - - - - - - -
| Q | U | I | T | CR | LF |
- - - - - - - - - - - - - -
(6 characters) and then it closes the communication channel.
In case of error (e.g. illegal command, non-existing file) the server always replies with:
- - - - - - - - - - - - - -
| - | E | R | R | CR | LF |
- - - - - - - - - - - - - -
(6 characters) and then it closes the connection with the client.
When implementing your client and server, use the directory structure included in the zip file provided with this text. After having unzipped the archive, you will find a directory named lab2.3, which includes a source folder with a nested server1 subfolder where you will write the server and a client1 subfolder where you will write the client. Skeleton empty files (one for the client and one for the server) are already present. Simply fill these files with your programs without moving them. You can use libraries of functions (e.g. the ones provided by Stevens). The C sources of such libraries must be copied into the source folder (do not put them into the client1 or server1 subdirectories, as such directories must contain only your own code!). Also, remember that if you want to include some of these files in your sources you have to specify the ”..” path in the include directive).
Your code will be considered valid if it can be compiled with the following commands, issued from the source folder:
gcc -std=gnu99 -o server server1/*.c *.c -Iserver1 -lpthread -lm
gcc -std=gnu99 -o client client1/*.c *.c -Iclient1 -lpthread -lm
The lab2.3 folder also contains a subfolder named tools, which includes some testing tools, among which you can find the executable files of a reference client and server that behave according to the specified protocol and that you can use for interoperability tests. Try to connect your client with the reference server, and the reference client with your server for testing interoperability (note that the executable files are provided for both 32bit and 64bit architectures. Files with the _32 suffix are compiled for and run on 32bit Linux systems. Files without that suffix are for 64bit systems. Labinf computers are 64bit systems). If fixes are needed in your client or server, make sure that your client and server can communicate correctly with each other after the modifications. Finally you should have a client and a server that can communicate with each other and that can interoperate with the reference client and server.
Develop a concurrent TCP server (listening to the port specified as the first parameter of the command line, as a decimal integer) that, after having established a TCP connection with a client, accepts file transfer requests from the client and sends the requested files back to the client, following the same protocol used in the iterative version. The server must create processes on demand (a new process for each new TCP connection).
Try the transfer of a large binary file (100MB) and check that the received copy of the file is identical to the original one (using diff) and that the implementation you developed is efficient in transferring the file in terms of transfer time.
- While a connection is active try to activate a second client against the same server.
- Try to activate on the same node a second instance of the server on the same port.
- Try to connect the client to a non-reachable address.
- Try to connect the client to an existing address but on a port the server is not listening to. Try to stop the server (by pressing ^C in its window) while a client is connected.
When you have finished testing your client and server, you can use the test script provided in the lab2.3 folder in order to perform a final check that your client and server conform to the essential requirements and pass the mandatory tests necessary for submission. In order to run the script, just give the following command from the lab2.3 folder ./test.sh The script will tell you if your solution is acceptable. If not, fix the errors and retry until you pass the tests.
- C - The development language
- Jacopo Nasi - Jacopx
The property of this project belongs to the Polytechnic of Turin since this project isn't delivered yet. After that it will be published under not-already defined license.