Обсуждение: Low disk performance?

Поиск
Список
Период
Сортировка

Low disk performance?

От
Martin Chlupac
Дата:
Hello everybody,
having this SQL query:

--------------
select
variable_id,float_value,ts,good_through,interval,datetime_value,string_value,int_value,blob_value,history_value_type
from "records_437954e9-e048-43de-bde3-057658966a9f" where variable_id
in (22727) and (ts >= '2010-10-02 11:19:55' or good_through >=
'2010-10-02 11:19:55') and (ts <= '2010-10-14 11:19:55' or
good_through <= '2010-10-14 11:19:55')
union all
select
variable_id,float_value,ts,good_through,interval,datetime_value,string_value,int_value,blob_value,history_value_type
from "records_1d115712-e943-4ae3-bb14-b56a95796111" where variable_id
in (24052) and (ts >= '2010-10-02 11:19:55' or good_through >=
'2010-10-02 11:19:55') and (ts <= '2010-10-14 11:19:55' or
good_through <= '2010-10-14 11:19:55')
order by ts
limit 2501 offset 0

---------------

and these two results:

1st run:
http://explain.depesz.com/s/1lT

2nd run:
http://explain.depesz.com/s/bhA

is there anything I can do about the speed? Only buying faster
hard-disk seems to me as the solution... Am I right?

Thank you in advance
  Martin

Re: Low disk performance?

От
tv@fuzzy.cz
Дата:
Hi, what is the size of the table and index (in terms of pages and
tuples)? Try something like

SELECT relpages, reltuples FROM pg_class WHERE relname = 'table or index
name';

And what indexes have you created? It seems to me there's just index on
the variable_id. It might be useful to create index on (variable_id, ts)
or even (variable_id, ts, good_through).

Tomas

> Hello everybody,
> having this SQL query:
>
> --------------
> select
> variable_id,float_value,ts,good_through,interval,datetime_value,string_value,int_value,blob_value,history_value_type
> from "records_437954e9-e048-43de-bde3-057658966a9f" where variable_id
> in (22727) and (ts >= '2010-10-02 11:19:55' or good_through >=
> '2010-10-02 11:19:55') and (ts <= '2010-10-14 11:19:55' or
> good_through <= '2010-10-14 11:19:55')
> union all
> select
> variable_id,float_value,ts,good_through,interval,datetime_value,string_value,int_value,blob_value,history_value_type
> from "records_1d115712-e943-4ae3-bb14-b56a95796111" where variable_id
> in (24052) and (ts >= '2010-10-02 11:19:55' or good_through >=
> '2010-10-02 11:19:55') and (ts <= '2010-10-14 11:19:55' or
> good_through <= '2010-10-14 11:19:55')
> order by ts
> limit 2501 offset 0
>
> ---------------
>
> and these two results:
>
> 1st run:
> http://explain.depesz.com/s/1lT
>
> 2nd run:
> http://explain.depesz.com/s/bhA
>
> is there anything I can do about the speed? Only buying faster
> hard-disk seems to me as the solution... Am I right?
>
> Thank you in advance
>   Martin
>
> --
> Sent via pgsql-performance mailing list (pgsql-performance@postgresql.org)
> To make changes to your subscription:
> http://www.postgresql.org/mailpref/pgsql-performance
>