On Thu, 2003-03-27 at 07:12, Loki wrote:
> We have a PostgreSQL database full of over 500,000 events.
> When the database has a small number of events, to
> retrieve a specific event stored in that db takes a mere
> few seconds.
> However, as the database grew in size to where it is now,
> it takes over 15-20 seconds to get information back from a
> query to that database. We have tried everything,
> including vacuum, which someone else recommended to me.
Sometimes it's the simple things. Did you do an explain on the query to
see its access plan? I don't see anything in your post about indexes.
> Am I correct in saying that 500,000 events shouldn't be a
> problem? Is there anyone out there with more than 500,000
> events in their DB than what we currently have?
PostgreSQL can give you speedy access into billions of rows of data
given proper indexing and maintenance. I'd like to see your query's
Senior Database Architect and author, tedia2sql (http://tedia2sql.tigris.org)
If this helped you, http://svcs.affero.net/rm.php?r=philovivero
In response to
pgsql-admin by date
|Next:||From: Andrew Biagioni||Date: 2003-03-27 22:02:44|
|Subject: Re: Upgrading: So now you tell me!!?!?|
|Previous:||From: Robert Treat||Date: 2003-03-27 20:13:51|
|Subject: Re: pg_dump problem|
pgsql-general by date
|Next:||From: Alvaro Herrera||Date: 2003-03-27 21:29:22|
|Subject: Re: About OIDs|
|Previous:||From: Tom Lane||Date: 2003-03-27 21:25:01|
|Subject: Re: file in data doesn't correspond to any oid in pg_class? |