Hi, I have performance problem with queries on uncommited data. I run huge import, with commit at each 10000 items. During import I make lot of selection queries (to protect data integrity), but they are very slow. With commit every 100 items, queries run lot of faster.
Profiler told me that most of time is spend in dataobjects and enhancer methods (get double etc). I think OpenJPA is making linear scan on all uncommited items. I played with IgnoreChanges settings, but it does not seems to help. So my question is: can I speed up those selection queries? Thanks, Jan
