[ please keep the list cc'd ] Alexander Voytsekhovskyy <young.inbox@xxxxxxxxx> writes: > Sorry again > here is both links: > https://explain.depesz.com/s/AEWj > https://explain.depesz.com/s/CHwF Don't think I believe that those are the same query --- there's a CTE in the second one that doesn't appear in the first, and it is eating a lot of time too. I have a vague recollection that there are cases where optimizations are possible in plain SELECT but not in data-modifying queries. So maybe this isn't pilot error but something triggered by that. We'd need to see a self-contained test case to verify that though. regards, tom lane
