hi
i found 2 issues with latest SA:
* r3449 introduces a memory leak if SA is used and then all cleaned
(clear_mappers() etc) - ArgSingleton is not cleared anymore and
grows. Same applies to orm.session._sessions, but i dont know whats
kept there.
$ python _test_leak.py repeat=10 leak
or
$
On Oct 25, 2007, at 1:21 PM, Yannick Gingras wrote:
Michael Bayer [EMAIL PROTECTED] writes:
If you were setting ctx.current.flush to something for every new
session, the approach here would be the same (Session ().flush =
abort_ro). if you were only setting ctx.current.flush to
3665 fixes ArgSingleton. as for session._sessions, that has to be
something on your end..its just WeakValue to a Session object, SA
doesnt reference those anywhere (unless, youre using ScopedSession,
are creating many new threads and disposing them, since its using a
TL. nothing new
endless loop fixed in 3666
On Oct 26, 2007, at 5:35 AM, svilen wrote:
hi
i found 2 issues with latest SA:
* r3449 introduces a memory leak if SA is used and then all cleaned
(clear_mappers() etc) - ArgSingleton is not cleared anymore and
grows. Same applies to orm.session._sessions, but
oops, one more ArgSingleton i didnt realize was being used, r3667.
On Oct 26, 2007, at 5:35 AM, svilen wrote:
hi
i found 2 issues with latest SA:
* r3449 introduces a memory leak if SA is used and then all cleaned
(clear_mappers() etc) - ArgSingleton is not cleared anymore and
grows.