i suggest to work with threads in the mining process

[Sanix-Darker]

When a mining started, it can take too long depending on how the difficulty have been configured, so when that operation start the api is waiting the process end before return a response, it's same when trying to access the api with any other endpoint /chain,... etc, so i suggest to use threads as i do:

from threading import Thread

def mine():
    .... mine() content....

...... some code here...
@app.route.....
def request_for_mine():
    .... some code here....
    # Launch the thread for the mine process, so that the API will be still available 
    Thread(target = mine).start() 

    return response

[satwikkansal]

Nice idea to perform the execution on different threads. But some API responses are dependent on the entire execution (for example, /chain endpoint), so they can't be dealt this way.

Also, expected behavior for a client is to call /mine, and then get the chain, which again can't be dealt on different threads (the chain won't be updated until mine is finished).

Edit: But yes, the mining figuring out the nonce process can be run parallely using multiprocessing module which can take advantage of multiple CPU cores.

[gerardkm94]

A nice approach for handling operations as mining will be the use of RabbitMQ or event data driven architectures. Of course is a big enhacement, but definetly will be necessary in a real time implementation.