problem in implementing multiprocessing

[gopal mishra]

I create two heavy objects sequentially without using multipleProcessing
then creation of the objects takes 2.5 sec.if i create these two objects in
separate process then total time is 6.4 sec.

i am thinking it is happening due to the pickling and unpickling of the
objects.if i am right then what could be the sollution.

my system configuration:
dual-core processor
winXP
python2.6.1


gopal mishra wrote:
> Hello, 
> 
> I am trying to implement the multiprocessing in my application to take
> advantage of multiple cores. I have created two
> 
> Separate process something like this.
> 
> que = Queue
> Process(target = getData, args=(que , section, MdbFile,)).start()
> Process(target = getData, args=(que , section, MdbFile,)).start()
> 
> In getData function I create the object(max 7MB size) and add in to queue
> (que.put (object)).
> 
> After that I fetch the object using que.get () and use in my application.
> but it takes more time to get the data.

More time than what?  Than not getting the date?  Than getting it some 
other way?

CPU, OS, and Python version may also be relevant.

> Any one can help me out this problem.
> 
> Thanks,
> Sibtey
> 
> My code<<<<<
> from multiprocessing import Process, Queue
> 
> def getData(queue, section, mdbFile):
>     """
>     This function returns the gapappdata for the given mdb file.
>     """
>     app  = MdbFile(mdbFile)
>     mdbData = app.data#it is a heavy object
>     queue.put((section,mdbData))
> 
> def getData2(mdbFile):
>     """
>     This function returns the gapappdata for the given mdb file.
>     """
>     app  = MdbFile(mdbFile)
>     mdbData = app.data#it is a heavy object
>     return mdbData
> 
> def test_multipleProcess(fromMdbFile, toMdbFile):
>     #multipleProcess
>     t1 = time.time()
>     queue = Queue()
>     sections = ['From', 'To']
>     Process(target= getData_1, args=(queue, 'From',fromMdbFile,)).start()
>     Process(target= getData_1, args=(queue, 'To',toMdbFile,)).start()
>     section, gapAppData = queue.get()
>     section, gapAppData = queue.get()
>     t2 = time.time()
>     print "total time using multiProcessing:",t2-t1
>     d1 = getData2(fromMdbFile)
>     d2 = getData2(toMdbFile)
>     print "total time withought multiProcessing:", time.time()-t2
> 
> 
> 
> 
> 
> if __name__=='__main__':
>     f1 =r" a.mdb"
>     f2 =r"b.mdb"
>     test_multipleProcess(f1,f2)
> 
> --
> http://mail.python.org/mailman/listinfo/python-list
>