"Jeremy M. Guthrie" <jeremy(dot)guthrie(at)berbee(dot)com> writes:
> I looked back at my code and I also need to reclarify something. The delete
> at the end is multiple delete statements within a transaction.
I think you are going to have to show us all the details of what you're
doing in between these queries. I was expecting to see a difference in
query plans, but you've got the exact same plan in all three runs ---
so it's not the planner making the difference here, nor the ANALYZE
statistics. My best conjecture now is something odd about the way you
are deleting the old data or loading the new.
regards, tom lane