server-side cursors (i.e. named) for all postgres functionality by
default is currently in the trunk, and you can turn it off by saying
client_side_cursors=True in create_engine(). however one user
reported that table reflection breaks, which I cannot reproduce. so
can some folks please try out
Hi Mike,
I've been lurking here in the shadows for a while...this particular topic looks
very interesting to me and I'd love to upgrade to HEAD to try it out. However,
I'm wondering what the risk factor is in doing that? Have there been any other
major (potentially destabilizing) changes
it should be pretty stable, we havent had much dramatic going on. i
moved some things around a bit in r2174 with regards to eager loading
but i think those changes have ironed out pretty well.
On Jan 10, 2007, at 10:39 PM, Daniel Miller wrote:
Hi Mike,
I've been lurking here in the
On Jan 9, 2007, at 5:24 AM, Michele Petrazzo wrote:
Because, like what I wrote on the psycopg2 ml (lists.initd.org/
pipermail/psycopg/2007-January/005250.html)
without the server side cursor, the memory usage on the client
side is proportionally at the number of records returned by the
Michael Bayer wrote:
On Jan 9, 2007, at 5:24 AM, Michele Petrazzo wrote:
Because, like what I wrote on the psycopg2 ml (lists.initd.org/
pipermail/psycopg/2007-January/005250.html)
without the server side cursor, the memory usage on the client
side is proportionally at the number of
OK well this is all quite disturbing !
I would want to examine this more closely. but, lets assume this is
really how it is, since thats how it looks (eesh).
So, you dont really want to use named cursors, you just want psycopg2
to use cursors in its underlying operations so that it has