Allen Bierbaum wrote:
On Tue, Feb 24, 2009 at 4:44 AM, Chris Miles miles.ch...@gmail.com wrote:
On Feb 22, 6:08 am, Allen Bierbaum abierb...@gmail.com wrote:
Python 2.5 and later will free up garbage collected memory, handing it
back to the system. Previous versions of Python would never
I hadn't described the details be there is not much to the work around
and pretty application specific.
The summary is that I moved some application level filtering that was
being done in python code on the results into the query so less
results are returned. This saves a great deal of memory
On Feb 22, 6:08 am, Allen Bierbaum abierb...@gmail.com wrote:
The python process. The number of objects seems to remain fairly
controlled. But the amount of resident memory used by the python
process does not decrease. I had expected that by calling
gc.collect(2) python would reclaim any
Allen Bierbaum schrieb:
The python process. The number of objects seems to remain fairly
controlled. But the amount of resident memory used by the python
process does not decrease. I had expected that by calling
gc.collect(2) python would reclaim any objects that could be freed and
free
On Tue, Feb 24, 2009 at 4:44 AM, Chris Miles miles.ch...@gmail.com wrote:
On Feb 22, 6:08 am, Allen Bierbaum abierb...@gmail.com wrote:
The python process. The number of objects seems to remain fairly
controlled. But the amount of resident memory used by the python
process does not
I spent some more time and came up with a completely standalone test
application. (attached) (requires SA 0.5.2 and some database backend).
I have tested this with postgres and with sqlite and both of them
exhibit the behavior I was talking about. Namely the memory keeps
increasing and stays
are you referring to memory of the python process or number of
unclaimed objects still present ? psycopg2 will buffer the full
results from the database in memory before fetchone() is ever called,
which will force the python process to grow to accomodate it. a
python process generally
The python process. The number of objects seems to remain fairly
controlled. But the amount of resident memory used by the python
process does not decrease. I had expected that by calling
gc.collect(2) python would reclaim any objects that could be freed and
free all memory associated with
i dont. you may want to look into using the server_side_cursors=True
option we provide on the postgres dialect, which prevents psycopg2
from fully pre-fetching result sets.
On Feb 21, 2009, at 2:08 PM, Allen Bierbaum wrote:
The python process. The number of objects seems to remain
