On Fri, 25 Feb 2005, Vig, Sandor (G/FI-2) wrote:
> Another "funny" thing: I have a query which runs
> on (Linux) PostgreSQL 7.4.x under 10 sec. I tried
> to run it on (Windows) PostgreSQL 8.0 yesterday.
> It didn't finished at all! (I shoot it down after 10 minutes)
> I made various tests and I figured out something interesting:
> The same query with:
> A, "history.undo_action_id > 0" runs in 10 sec.
> B, "history.undo_action_id is not null" runs in 10 sec.
> C, "history.undo_action_id is null" runs forever (?!)
> I used EXPLAIN but I couldn't figure out what the problem was.
EXPLAIN ANALYZE would be more useful. My first guess would be that the IS
NULL is returning many more than the estimated 1 row and as such a nested
loop is a bad plan. How many history rows match type_id=6 and
undo_action_id is null?
In response to
pgsql-performance by date
|Next:||From: Tom Lane||Date: 2005-02-25 16:05:10|
|Subject: Re: Possible interesting extra information for explain analyze? |
|Previous:||From: Jeff||Date: 2005-02-25 13:49:23|
|Subject: Possible interesting extra information for explain analyze?|