> On 30 Dec 2015, at 21:12, Tom Lane <t...@sss.pgh.pa.us> wrote: > > Emre Hasegeli <e...@hasegeli.com> writes: >>> I don’t see how to solve this problem without changing explain analyze >>> output to accommodate for “unknown” value. I don’t think “0” is a >>> non-confusing representation of “unknown” for most people, and from the >>> practical standpoint, a “best effort” estimate is better than 0 (i.e. I >>> will be able to estimate how efficient BRIN index is for my tables in terms >>> of the number of tuples retrieved/thrown away) > > We do already have a nearby precedent for returning zero when we don't > have an accurate answer: that's what BitmapAnd and BitmapOr plan nodes > do. (This is documented btw, at the bottom of section 14.1.)
+1 for following a precedent. > >> The number of retrieved and thrown away rows are already available on >> the upper part of the plan. Bitmap Index Scan should provide the rows >> that matched the index. > > It doesn't have that information. > >> Another alternative would be just returning >> the number of matching pages (by not multiplying with 10). It might >> be better understood. > > No, it would not, at least not unless we found a way to explicitly mark > the output as being blocks not rows (which would doubtless break a lot of > existing client-side code). Zero is fairly clearly an impossible value, > whereas anything that's not zero is going to be taken at face value by > many users. But is it? Is it impossible for the BRIN bitmap index scan to return 0 rows (say, if the value being matched is outside the min/max boundary for every block range?) Granted, if we document that it always returns 0 and should be ignored, then confusing the actual 0 with the 0 as a representation of “unknown” would be less a problem. -- Oleksii
