Re: postgres 7.4 vs 8.x redux: query plans

On 4/3/07, Alex Deucher <alexdeucher(at)gmail(dot)com> wrote:
> On 4/3/07, Merlin Moncure <mmoncure(at)gmail(dot)com> wrote:
> > On 4/3/07, Alex Deucher <alexdeucher(at)gmail(dot)com> wrote:
> > (('{AB6698130,AB7076908,AB6499382,AB6438888,AB6385893,AB6378237,AB7146973,AB7127138,AB7124531,AB7124513,AB7123427,AB7121183,AB7121036,AB7110101,AB7100321,AB7089845,AB7088750,AB7031384,AB7021188,AB7006144,AB6988331,AB6973865,AB6966775,AB6935066,AB6931779,AB6923412,AB6902405,AB6892488,AB6886288,AB6880467,AB6874269,AB6871439,AB6868615,AB6819495,AB6807740,AB6799138,AB6796038,AB6769347,AB6732987,AB6722076,AB6718130,AB6717543,AB6714564,AB6701821,AB6667761,AB6666630,AB6655069,AB6648287,AB6643969,AB6636412}'::character
> > > varying[])::text[]))
> > > -> Bitmap Index Scan on t1_pkey (cost=0.00..216.69 rows=50
> > > width=0) (actual time=198.188..198.188 rows=50 loops=1)
> > > Index Cond: ((num)::text = ANY
> >
> > bitmap scan:
> > * did you run analyze?
>
> yes.
>
> > * is effective_cache_size set properly?
>
> It should be. I based it on the output of `free`. It's set to 988232.
> The system has 8 GB of ram.
>
> > * if nothing else works, try disable bitmap scan and running query.
>
> I'll give that a try and post the results.
>

Turning off bitmapscan ends up doing a sequential scan. Turning off
both bitmapscan and seqscan results in a bitmap heap scan. It doesn't
seem to want to use the index at all. Any ideas?

Alex