Persistent Queue implementations?

Armin Steinhoff a-steinhoff at web.de
Wed Dec 25 04:57:02 EST 2002


aahz at pythoncraft.com (Aahz) wrote in message news:<auaja2$sod$1 at panix2.panix.com>...
> In article <au833j$a7f$1 at baldur.whoi.edu>,
> Karl A. Krueger <kkrueger at example.edu> wrote:
> >
> >Using pickle in the most obvious way (pickling the whole queue) would
> >have to be done in every put() and get() call, since the copy on disk
> >needs to reflect the copy in memory.  And pickling the whole queue would
> >be -slow-, since it would have to write the whole queue each time --
> >keeping it locked for read, and thus blocking all the threads that want
> >to write to it.
> 
> Just how much ACID do you need, anyway?  If you really care, you should
> use a full-blown database.


Give gadfly a try ... it's completely written in Python.

Armin

>  If not, I'd take Kevin's advice to use
> cPickle, possibly using a separate thread to do the writes.



More information about the Python-list mailing list