Re: Why PG uses nested-loop join when no indexes are available?

Thank you for your fast response!

I'm sorry, my vocabulary may be not correct and my "french approach" to
explain my problem is not efficient ;).

But the underestimation problem in complex queries is still there? ... and
generates potential "dangerous" plans (nested loop).

We thought at the beginning we were alone but it seems to be a problem of
most database systems.
What do you think about the paragraph 4.1 of this paper
http://www.vldb.org/pvldb/vol9/p204-leis.pdf ?

Regards,
-------------------
David Grelaud,
Ideolys.

2016-01-13 16:02 GMT+01:00 Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>:

> David Grelaud <dgrelaud(at)ideolys(dot)com> writes:
> > Statistics are not propagated when Common Table Expressions (CTE) are
> used.
> > The cardinality of a CTE is equal to 1 most of the time
>
> Really?
>
> The rest of this seems to be proceeding from completely false assumptions.
>
> regards, tom lane
>