Re: Database connection caching
Thanks everybody for their replies. Lorenzo -- http://mail.python.org/mailman/listinfo/python-list
Re: Database connection caching
[EMAIL PROTECTED] wrote: Hi all, is there an alternative way of: - create a connection object - open the connection - close the connection psycopg, a Postgresql database adapter does connection pooling automatically http://initd.org/projects/psycopg1 Most Zope database adapters also have implicit connection pooling. Istvan. -- http://mail.python.org/mailman/listinfo/python-list
Re: Database connection caching
On 18 Mar 2005 04:52:03 -0800, [EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote: > is there an alternative way of: > > - create a connection object > - open the connection > - close the connection > > every time one has to run a query. It's actually morte like: create connection create cursor execute SQL process cursor close cursor close connection The bit that you can avoid it the creation of the connection - that's an intensive process. If your system is single threaded, or if it's a very small app, you can just create a single open connection and keep using that. I usually find that I need a pool of open connections, though. -- Cheers, Simon B, [EMAIL PROTECTED], http://www.brunningonline.net/simon/blog/ -- http://mail.python.org/mailman/listinfo/python-list
Re: Database connection caching
Lorenzo> is there an alternative way of: Lorenzo> - create a connection object Lorenzo> - open the connection Lorenzo> - close the connection Lorenzo> every time one has to run a query. Sure, create a Queue.Queue object and stuff a number of connections into it. When you want a connection call the queue's .get() method. When you're done with it .put() it back. This has the other nice feature that a program with a large number of threads can't overwhelm your database. Skip -- http://mail.python.org/mailman/listinfo/python-list
Re: Database connection caching
On 18 Mar 2005 04:52:03 -0800, [EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote: > Hi all, > > is there an alternative way of: > > - create a connection object > - open the connection > - close the connection > > every time one has to run a query. Why not use cursor objects with a single connection object? You can have any number of cursor objects and can run with a single connection object A good introduction is at http://www.amk.ca/python/writing/DB-API.html Regards, -- Swaroop C H Blog: http://www.swaroopch.info Book: http://www.byteofpython.info -- http://mail.python.org/mailman/listinfo/python-list
Database connection caching
Hi all, is there an alternative way of: - create a connection object - open the connection - close the connection every time one has to run a query. I assume that big sites do something to prevent the performance hit given by these operations, don't they? Would you kindly point me to some examples of good practices in this sense beside http://www.danga.com/words/2004_oscon/oscon2004.pdf ? Thanks, Lorenzo -- http://mail.python.org/mailman/listinfo/python-list
