On Wednesday, March 23, 2011, 10:50:46 AM, Michael Bayer wrote:
uh yeah you totally nailed that one, QueuePool and
SingletonThreadPool had similar issues.
it was quite hard to get into this. :)
you can get at these from hg via:
http://hg.sqlalchemy.org/sqlalchemy/archive/default.tar.gz
On Mar 23, 2011, at 1:25 PM, Jaimy Azle wrote:
On Wednesday, March 23, 2011, 10:50:46 AM, Michael Bayer wrote:
uh yeah you totally nailed that one, QueuePool and
SingletonThreadPool had similar issues.
it was quite hard to get into this. :)
you can get at these from hg via:
On Thursday, March 24, 2011, 12:33:37 AM, Michael Bayer wrote:
the pool_timeout parameter manages this from a client perspective,
in that a stale connection won't be used. An application with a
decent amount of activity should clean these connections out
regularly as they are closed the
Dear All,
I am using SQLAlchemy with Jython, I found there is a different
SQLAlchemy behaviour between CPython and Jython. In Jython, discarded
overflow connection object does not automatically close the actual
database socket connection during gc, anybody could confirm this?
my temporary,
uh yeah you totally nailed that one, QueuePool and SingletonThreadPool had
similar issues. I think in Jython those connections would eventually be
closed when asynchronous gc collects them, but we definitely don't ever want to
rely on gc to close connections we're discarding.
This issue
