tansly / ceng435 Goto Github PK

The destination script simply discards any out-of-sequence packets. Buffering future packets may significantly improve performance.

After obtaining the results and plotting these figures; report the figures and write your comments about the figures. You are expected to write at least one paragraph per figure. Ensure that the figures are self-explanatory: labels, captions, and units are clearly stated in the figures. Use the paragraph to draw the attention to the most significant points of the figure, give specifics but do not repeat details.

In your reports try to analyze and demonstrate your results.

Labels/captions are OK. Figures are self-explanatory.

Need to explain the figures and analyze the results.

Before submission, I'll set -DNDEBUG and -O2 in compile flags, and remove -ggdb.

We will use MD5. This is urgent and should be done before we start the experiments.

Implementation details of the broker and the effects of our design decisions must be explained in the report.

#1 and #2 implement the timeout mechanism in the broker in two separate ways. #1 uses SIGALRM, #2 uses a timer thread. Need to evaluate the pros and cons and pick one.

The broker became a huge mess. Locking mechanism is just an example. What have I done wrong? A better design would most likely avoid lots of the problems that we've encountered.

I will improve it if we finish everything early and have a lot of free time (which is nearly impossible). Maybe in the holiday.

The recv() in broker call that receives data from the source may return prematurely (in case of receiving a signal etc.) This may be a problem if it returns when no data is available. In this case, I guess recv() returns -1, which may cause broker to freak out. What would actually happen, I don't really know.

This is not urgent because I don't think we will hit such a case. But it would be good to keep this issue in mind.