The system running the python has around 70gb of physical RAM available at the point of the error. We do have a largish transaction and run a lot of batched inserts, but the same dataset can be inserted into another database with no issues. Unfortunately the error seem to indicate problem with ODBC memory and not the database.
Has anyone seen it before? How one can make sure that it's not some sort of weird interpretation of the backend/server error that happens in the ODBC drivers? And if it is truly ODBC error, what would be the next step? At this point the error is pretty repeatable.
3.10.0-327.28.3.el7.x86_64 #1 SMP Thu Aug 18 19:05:49 UTC 2016 x86_64 x86_64 x86_64 GNU/Linux