Re: Tweaking the planner's heuristics for small/empty tables

"Kevin Grittner" <Kevin(dot)Grittner(at)wicourts(dot)gov> writes:
> Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:
>> Another thing that struck me while looking at the code is that the
>> curpages clamp is applied to indexes too, which seems like a
>> thinko. A table occupying a few pages wouldn't likely have an
>> index as big as the table itself is.

> But not zero pages, either.

Huh? I think you misread the test. Keep in mind that the starting
value of curpages is the actual current size of the relation (as
reported by lseek(SEEK_END)). The clamp action is triggering if
pg_class.relpages is zero, indicating (approximately) that we've never
yet run vacuum on the relation.

BTW, in some quick testing it seems like a newly-created index will
start out with relpages nonzero anyway, making the point moot; so adding
a relkind test here is really more documentation than anything else.

regards, tom lane