On Saturday 11 April 2009 00:41:54 Jasen Betts wrote:
> On 2009-04-02, Alvaro Herrera <[email protected]> wrote:
> > James Kitambara wrote:
> >> Dear Srikanth,
> >> You can solve your problem by doing this
> >>
> >> THE SQL IS AS FOLLOWS
> >> ASSUME TIME INTERVAL 2008-12-07 07:59:59 TO 2008-12-07 08:58:59 AND THE
> >> TABLE NAME time_interval
> >>
> >> COUNT (*) FROM
> >> (select customer_id, log_session_id, start_ts, end_ts ,
> >> end_ts-start_ts as "Interval" from time_interval
> >> where end_ts-start_ts >= '1 hour'
> >> and '2008-12-07 07:59:59' between start_ts and end_ts)
> >> AS COUNT ;
> >
> > Another way to phrase the WHERE clause is with the OVERLAPS operator,
> > something like this:
> >
> > WHERE (start_ts, end_ts) OVERLAPS ('2008-12-07 07:59:59', '2008-12-07
> > 08:59:59')
> >
> > What I'm not so sure about is how optimizable this construct is.
> >
>
> http://www.postgresql.org/docs/8.3/interactive/xindex.html
> if you gave the apropriate GIST index on (start_ts, end_ts) the
> overlaps may be optimisable. the subquery will run to completion
> and count will count the results. - but this form gives different results.
>
> beter to do
>
> select COUNT (*) AS COUNT FROM time_interval
> WHERE (start_ts, end_ts) OVERLAPS ('2008-12-07 07:59:59', '2008-12-07
> 08:59:59')
>
> or
>
> select COUNT (*) AS COUNT FROM time_interval
> where end_ts-start_ts >= '1 hour'
> and '2008-12-07 07:59:59' between start_ts and end_ts;
I only managed to get this to use the gist-index, and not with the overlaps
operator. I had to install the contrib-module btree_gist in order to be able to
create a gist index on the timestamps.
This is my index:
CREATE index origo_tart_end_time_idx on onp_crm_activity_log using gist
(start_time, end_time) ;
start_time and end_time are both timestamps.
Here are the EXPLAIN outputs:
andreak=# EXPLAIN ANALYZE select id from onp_crm_activity_log where end_time -
start_time >= '1 hour' AND '2008-12-07 07:59:59' between start_time and
end_time;
QUERY
PLAN
---------------------------------------------------------------------------------------------------------------------------------------------------------------
Bitmap Heap Scan on onp_crm_activity_log (cost=10.56..232.62 rows=76 width=4)
(actual time=0.175..0.175 rows=0 loops=1)
Recheck Cond: (('2008-12-07 07:59:59'::timestamp without time zone >=
start_time) AND ('2008-12-07 07:59:59'::timestamp without time zone <=
end_time))
Filter: ((end_time - start_time) >= '01:00:00'::interval)
-> Bitmap Index Scan on origo_tart_end_time_idx (cost=0.00..10.54 rows=229
width=0) (actual time=0.168..0.168 rows=0 loops=1)
Index Cond: (('2008-12-07 07:59:59'::timestamp without time zone >=
start_time) AND ('2008-12-07 07:59:59'::timestamp without time zone <=
end_time))
Total runtime: 0.274 ms
(6 rows)
andreak=# EXPLAIN ANALYZE select id from onp_crm_activity_log where
(start_time, end_time) OVERLAPS('2008-11-07 07:59:59'::timestamp, '2008-12-07
08:59:59'::timestamp);
QUERY PLAN
----------------------------------------------------------------------------------------------------------------------------------------------------
Seq Scan on onp_crm_activity_log (cost=0.00..319.29 rows=2968 width=4)
(actual time=14.542..15.794 rows=83 loops=1)
Filter: "overlaps"(start_time, end_time, '2008-11-07 07:59:59'::timestamp
without time zone, '2008-12-07 08:59:59'::timestamp without time zone)
Total runtime: 16.129 ms
(3 rows)
Is it possible to make the overlaps operator use the index? I'd prefer the
overlaps-syntax as I find it cleaner.
--
Andreas Joseph Krogh <[email protected]>
Senior Software Developer / CEO
------------------------+---------------------------------------------+
OfficeNet AS | The most difficult thing in the world is to |
Rosenholmveien 25 | know how to do a thing and to watch |
1414 Trollåsen | somebody else doing it wrong, without |
NORWAY | comment. |
| |
Tlf: +47 24 15 38 90 | |
Fax: +47 24 15 38 91 | |
Mobile: +47 909 56 963 | |
------------------------+---------------------------------------------+
--
Sent via pgsql-sql mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-sql
