I bet EXPLAIN shows a seq. scan b/c of the date field. index the date
field too.
eric
> -----Original Message-----
> From: Brian Avis [mailto:brian.avis@searhc.org]
> Sent: Monday, November 19, 2001 4:06 PM
> To: pgsql-general@postgresql.org
> Subject: [GENERAL] Need more speed from this.
>
>
> I hope someone can help me out with this.
>
> Postgres 7.1.3
> Solaris 2.8
>
> I have a table called system_data setup like this.
>
>
> | server_id | date | time | data_type_id | data |
> |___________|________|________|______________|______|
> | | | | | |
> | 26 | a date | a time | 8 | 98 |
> | | | | | |
> -----------------------------------------------------
>
>
> This table is indexed on server_id and data_type_id.
>
> It currently has about 674,456 rows and is growing on a daily
> basis.
>
> The problem is that selects from this table are really slow.
>
> For instance if I do this.
>
> SELECT AVG(data) FROM system_data
> WHERE date BETWEEN CAST('2001-02-01' AS DATE)
> AND CAST ('2001-02-28' AS DATE)
> AND
> server_id = 26
> AND
> data_type_id = 8;
>
> It will take several seconds to complete this select.
>
> I need this type of data retrieval to be much faster if
> possible.
>
> Thanks in advance for any assistance.
>
>
> ---------------------------(end of
> broadcast)---------------------------
> TIP 6: Have you searched our list archives?
>
> http://archives.postgresql.org
>
