David Nguyen – 103 752 978Jonathan Nguy – 633 799 761Project 1For client.c,1st, we initialized the variables using the arguments the user inputted. Since there can be more than just one set of [count | pong packet size | filling character], we used arrays for those values, with each index holding the corresponding set of values.Next, we got the socket number and outputted an error if the socket was not assigned. We used the inputted address to set up the server info. We checked to see if we were successful at connecting to the server. Once we knew we knew we were connected, we went through each set of [count | pong packet size | filling character] and sent them to the server. Once we got the responses, we printed out the buffer and finally, closed the socket.For server.c,1st, we got the port number from the input, and if there was no argument, we set a default port. We then created a listening socket and bound the socket addresses to the listening socket, then started to listen.We continued to listen until we got a connection and then accepted it. If the version numberwas 1, we used the regular input, and if the version was 2, we responded with different characters using the formula in the project specs. We closed the connection and continued to wait for new connections indefinitely.Evaluation question: To test the RTT for our project, we found the time before we went through the for loop using gettimeofday (). We collected the data with a count of 4 and then divided the final time difference by 4. We did this 5 times per payload size, and calculated the average of the 5.3Graphing the average values with the payload size in bytes, we get the following graph:Payload Byte SizeSeconds Elapsed AveragePayload Byte Size Seconds Elapsed AveragePayload Byte Size Seconds Elapsed Average1 0.002507 0.00324 16 0.002539 0.0020928 256 0.005237 0.00392441 0.002081 16 0.001891 256 0.0078191 0.004272 16 0.002199 256 0.002315 1 0.005677 16 0.001793 256 0.002361 1 0.001663 16 0.002042 256 0.00189 2 0.006265 0.0043282 32 0.002229 0.0036618 512 0.004649 0.0151092 0.009601 32 0.001689 512 0.059816 2 0.001801 32 0.002149 512 0.002519 2 0.002185 32 0.003899 512 0.002234 2 0.001789 32 0.008343 512 0.006327 4 0.002672 0.002193 64 0.001869 0.0021666 4 0.001945 64 0.00181 4 0.002082 64 0.003057 4 0.001665 64 0.0021 4 0.002601 64 0.001997 8 0.001731 0.0026214 128 0.005681 0.0058532 8 0.005167 128 0.00276 8 0.001531 128 0.00198 8 0.00219 128 0.010724 8 0.002488 128 0.008121 0 100 200 300 400 500 6000000.010.010.010.010.010.02As we expected, the more bytes we needed to send, the longer it took for the server to respond. It was not possible to split the time into the 4 types of delay. The processing delay can’t be determined because after send the code, we cannot tell what the router does to process the code. In our case, we send the packets and wait for a response. Similarly, we cannot determine the queuing or transmission delays. We do not control these delays, but instead the router or network card do. Lastly, the propagation delay isn’t something we control, either, and we can’t determine when the actual packets are in the propagation phase.Referenceshttp://cs.baylor.edu/~donahoo/practical/CSockets/textcode.htmlhttp://rabbit.eng.miami.edu/info/functions/time.html(for gettimeofday
View Full Document
Unlocking...