[EMAIL PROTECTED] wrote: > On Feb 27, 3:32 pm, Tim Golden <[EMAIL PROTECTED]> wrote: >> [EMAIL PROTECTED] wrote: >>> The problem I have is that since I import WMI, it takes a long time >>> and we have users complaining about it. So I stuck the import >>> statement into a separate thread and set it to a daemon so it could do >>> its thing in the background and the rest of the script would finish >>> and exit. >> Two things: >> >> 1) If you run WMI in a thread, you'll need to call >> pythoncom.CoInitialize first: >> >> <code> >> import pythoncom >> import wmi >> >> pythoncom.CoInitialize () >> c = wmi.WMI () >> # >> # do things >> # >> pythoncom.CoUninitialize () >> </code> >> >> 2) If you need a bit of speed running WMI, see the post >> I sent a few days ago to someone else: >> >> http://mail.python.org/pipermail/python-win32/2007-February/005550.html >> >> TJG > > Thanks! This works for my problem. It appears to cut the real time > required for my script to run by 30-50%. I tried to figure out how to > apply your answer to the other fellow, but I am actually querying WMI > for the amount of RAM and the CPU type and I just don't see how to use > your example in these cases. I am new to the WMI paradigm.
If you want to post some specific code examples, I'm happy to talk you through possible optimisations. TJG -- http://mail.python.org/mailman/listinfo/python-list
