This PR enable the use of multiple threads to import issues. This is speeding up the sending of JSON-RPC requests to fedorahosted, by allowing more than 1 request at the time to be sent.
I always wonder what is the best approach between Thread and Multiprocessing ( http://blog.pingoured.fr/index.php?post/2014/11/27/Python-multiprocessing-and-queue )
So I just asked duckduckgo and landed on http://stackoverflow.com/questions/3044580/multiprocessing-vs-threading-python#3046201
So this is also looking good :)
Thanks for pointing out the Multiprocessing module (did not know about it). I might play around with it ;)
Aren't network calls(RPCs or get/post) non-CPU bound? (i don't know for sure but, this will be my guess). In that case, according to the links, threading should do better.
threading
Network calls are definitely not CPU bound and much more I/O bound :)
Good catch @vivekanand1101 I missed this last bullet when checking this page.
So :thumbsup: for me
close as this will need to be refactored after #105 is merged
Pull-Request has been closed by cverna
This PR enable the use of multiple threads to import issues. This is speeding up the sending of JSON-RPC requests to fedorahosted, by allowing more than 1 request at the time to be sent.