Re: performace review - Mailing list pgsql-general

From Alexander Staubo
Subject Re: performace review
Date
Msg-id 9D9CFDF3-D6A0-4D01-A8ED-B5BF51E11301@purefiction.net
Whole thread Raw
In response to performace review  ("Tomi NA" <hefest@gmail.com>)
Responses Re: performace review  (Jonathan Vanasco <postgres@2xlp.com>)
List pgsql-general
On Oct 7, 2006, at 20:06 , Tomi NA wrote:

> I was just reading http://www.opencrx.org/faq.htm where RDBMS engines
> are one of the questions and see pgsql bashed sentence after sentence.
> Can anyone offer any insight as to weather it's fact or FUD?

As with any use of a database, it is useless and/or impossible to
evaluate blanket statements about performance without also evaluating
the application's use of the database. One person's slow join may be
another person's fast join.

Also, note how that the FAQ says that "the default distribution of
openCRX includes some indexes", implying that the schema must be
adapted to one's database to achieve adequate performance. The FAQ
does not mention whether the cited scaling limits include any
attempts at tuning PostgreSQL.

I don't see PostgreSQL being "bashed sentence after sentence",
however -- the two "known limitations" listed for PostgreSQL are
"slow (even for small datasets)" and "jokes [sic] on 3-table-joins"
-- and among the open-source databases mentioned, PostgreSQL is
described as scaling the highest.

With a running OpenCRX installation you could turn on PostgreSQL's
query logging to identify slow queries or bad query patterns. I'm
sure both the open-source communities would appreciate the feedback.

Alexander.


pgsql-general by date:

Previous
From: Rafal Pietrak
Date:
Subject: Re: time conversion fuinctions
Next
From: Matthew Peter
Date:
Subject: increment row number function question