On Wed, Nov 9, 2016 at 11:19 AM, Pierre Ducroquet
wrote:
> Indeed the words in the query are correlated, but I do hope that the FTS
> indexing is able to cope with that.
If the query returns correct results in reasonable time it can. OTOH
the planner, and the statistics system, is another beast.
2016-11-09 11:19 GMT+01:00 Pierre Ducroquet :
> On Wednesday, November 9, 2016 10:40:10 AM CET Francisco Olarte wrote:
> > Pierre:
> >
> > On Wed, Nov 9, 2016 at 10:22 AM, Pierre Ducroquet
> >
> > wrote:
> > > The query does a few joins «after» running a FTS query on a main table.
> > > The FTS q
On Wednesday, November 9, 2016 10:51:11 AM CET Pavel Stehule wrote:
> 2016-11-09 10:40 GMT+01:00 Francisco Olarte :
> > Pierre:
> >
> > On Wed, Nov 9, 2016 at 10:22 AM, Pierre Ducroquet
> >
> > wrote:
> > > The query does a few joins «after» running a FTS query on a main table.
> > > The FTS que
2016-11-09 10:40 GMT+01:00 Francisco Olarte :
> Pierre:
>
> On Wed, Nov 9, 2016 at 10:22 AM, Pierre Ducroquet
> wrote:
> > The query does a few joins «after» running a FTS query on a main table.
> > The FTS query returns a few thousand rows, but the estimations are wrong,
> > leading the optimize
Pierre:
On Wed, Nov 9, 2016 at 10:22 AM, Pierre Ducroquet
wrote:
> The query does a few joins «after» running a FTS query on a main table.
> The FTS query returns a few thousand rows, but the estimations are wrong,
> leading the optimizer to terrible plans compared to what should happen, and
> th
