Andrei Zubkov <zubkov(at)moonset(dot)ru> writes:
> Thank you for your attention and for the problem resolution. However
> I'm worry a little about possible performance issues related to
> monitoring solutions performing regular sampling of statistic views to
> find out the most intensive objects in a database.
There's no actual evidence that the new formulation is meaningfully
worse for such cases. Sure, it's probably *somewhat* less efficient,
but that could easily be swamped by pgstat or other costs. I wouldn't
care to worry about this unless some evidence is presented that we've
created a big problem.
regards, tom lane