On Tue, Sep 8, 2020 at 6:33 PM Magnus Hagander wrote:
Per his session list, 11.2.
11.2 is on my personal testing machine. The real tests run with 12.1
Oh, now *cursors* is definitely something I didn't think of. And especially in the context of ODBC, I wonder if it might be creating cursors transparently, and that this somehow causes the problems.
Michael, do you know if that might be the case? Or try enabling log_statements to check if it is?
This is very interesting. I'll try to check it. I'll post here a reply if we do have open cursors. Thanks.