From: | Stephan Szabo <sszabo(at)megazone23(dot)bigpanda(dot)com> |
---|---|
To: | Rupert <rminnett(at)rsmas(dot)miami(dot)edu> |
Cc: | <pgsql-general(at)postgresql(dot)org> |
Subject: | Re: Cursors with Large, Ordered Result Sets |
Date: | 2003-04-01 18:27:35 |
Message-ID: | 20030401095839.N51133-100000@megazone23.bigpanda.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
On 27 Mar 2003, Rupert wrote:
> Thanks for the quick reply and sorry for the slow response.
>
> Yes, this is very similar to what we are currently doing and it seems
> to be working rather well - much to my surprise. However, I still have
> the same questions regarding the actual steps being taken by the DBMS
> to order a massive result set. Doesn't it need to have the entire
> result in memory before it can return the first records? If so, and
Whether or not you need to sort depends on the query.
For example:
select * from foo order by a;
could choose to use an index on foo.a to do the ordering rather than a
sort. It'll choose based on what it thinks will be the best plan.
> the result is larger than the RAM, does it dump it to disk and then
> sort?
It's based on the sort_mem GUC parameter and will go to disk if necessary.
From | Date | Subject | |
---|---|---|---|
Next Message | Joe Conway | 2003-04-01 18:46:26 | Re: create type problem! |
Previous Message | scott.marlowe | 2003-04-01 17:26:27 | Re: Dates in inserts |