Re: query is taking longer time after a while

On Sun, Oct 04, 2009 at 11:08:01AM -0400, Tom Lane wrote:
> Sam Mason <sam@samason.me.uk> writes:
> > On Sun, Oct 04, 2009 at 01:44:30AM -0700, tomrevam wrote:
> >> ->  Bitmap Index Scan on session_allocation_info_status_idx (cost=0.00..5.28 rows=1 width=0) (actual
time=1619.652..1619.652rows=51025 loops=1) 
> >> Index Cond: ((status)::text = 'active'::text)
> >> ->  Bitmap Index Scan on session_allocation_info_status_idx (cost=0.00..5.28 rows=1 width=0) (actual
time=806.770..806.770rows=46601 loops=1) 
> >> Index Cond: ((status)::text = 'setup'::text)
> >> Total runtime: 4819.990 ms
>
> > Wow, that's quite a change in run time!  Are you sure planner stats are
> > being kept up to date?
>
> It's not the planner's fault.  Note that the parent BitmapHeapScan is
> still returning the same number of rows.

Sorry, I chopped out too much context.  Here's the "early" run, the
estimated and real row counts look good to me:

On Sun, Oct 04, 2009 at 01:44:30AM -0700, tomrevam wrote:
>          ->  Bitmap Index Scan on session_allocation_info_status_idx (cost=0.00..48.93 rows=1555 width=0) (actual
time=0.244..0.244rows=1557 loops=1) 
>                Index Cond: ((status)::text = 'active'::text)
>          ->  Bitmap Index Scan on session_allocation_info_status_idx (cost=0.00..48.93 rows=1555 width=0) (actual
time=0.181..0.181rows=1609 loops=1) 
>                Index Cond: ((status)::text = 'setup'::text)
>  Total runtime: 2.193 ms

Or did I missing something else?

--
  Sam  http://samason.me.uk/