Homework 1 forum

Clean stop of gossiper

Picture of Pasindu Nivanthaka Tennage
Re: Clean stop of gossiper
by Pasindu Nivanthaka Tennage - Friday, 23 October 2020, 15:27


Thanks for the question.

Can you let us know which particular test is failing? Or is it the entire test set?

We did not observe this behavior in our reference implementation, which we used to implement tests. So I suggest that your gossiper should use more parallelism and improve the speed at which it processes messages.

So the next step I can suggest you is, do a profiling in your code in performance sense, and try to identify bottlenecks. Then try to parallelize things and see if you get good enough speed.