Question

Hi I am using Serial port communication to one of my device using python codes. It is sending a set of Hex code, Receives a set of data. process it.

This data has to be stored in to a database.

I have another script that has MYSQLdb library pushing it in to the database.

If I do that in sequentially in one script i lose a lot in sampling rate. I can sample up to 32 data sets per second if I dont connect to a database and insert it in to the table.

If I use Multiprocessing and try to run it my sampling rate goes to 0.75, because the parent process is waiting for the child to join. so how can i handle this situation.

Is it possible to run them independently by using a queue to fill data?

Answer 1

Use threading and Queue.

here is an example:

from Queue import Queue
import threading
import serial

def readline(ser,output):
    threadLock.acquire()# Get lock to synchronize threads
    output.put(ser.readline())
    threadLock.release()# Free lock to release next thread
    # this will go on forever until you kill the program

if __name__=='__main__':
    output = Queue()
    with serial.Serial(com,baudrate=115200,timeout=1) as ser:
        ser_thread = threading.Thread(target=readline, args=(ser,output,))
        ser_thread.start()
        while(True):
             threadLock.acquire()#wait until lock is free
             try:
                 data=output.get()
             threadLock.release()# Free lock to release next thread
             # do somthing with data