"Richard Huxton" <[EMAIL PROTECTED]> writes:
>> ?? Knowing that your previous guess was wrong doesn't tell you what the
>> right answer is, especially not for the somewhat-different question that
>> the next query is likely to provide.
> Surely if you used a seqscan on "where x=1" and only got 2
You and Stephan hit it right on the nose - our table has been
maliciously propagated with thousands of faulty values - once gone
index are in use and DB is SPEEDING along 8)
Thanks for your help!!!
-r
On Thu, 10 May 2001 21:49:28 + (UTC), in
comp.databases.postgresql.general you wrote:
>
"Richard Huxton" <[EMAIL PROTECTED]> writes:
> Why doesn't PG (or any other system afaik) just have a first guess, run the
> query and then if the costs are horribly wrong cache the right result.
?? Knowing that your previous guess was wrong doesn't tell you what the
right answer is, especially n
Chris Jones <[EMAIL PROTECTED]> writes:
>> Ah. You must have a few values that are far more frequent (like tens of
>> thousands of occurrences?) and these are throwing off the planner's
>> statistics.
> I had a similar situation, where I had a lot of rows with 0's in
> them. Changing those to N
On Thu, May 10, 2001 at 05:22:07PM -0400, Tom Lane wrote:
> [EMAIL PROTECTED] writes:
> > No the query usually returns between 0 and 5 rows. Usually not zero -
> > most often 1.
>
> Ah. You must have a few values that are far more frequent (like tens of
> thousands of occurrences?) and these a
[EMAIL PROTECTED] writes:
> No the query usually returns between 0 and 5 rows. Usually not zero -
> most often 1.
Ah. You must have a few values that are far more frequent (like tens of
thousands of occurrences?) and these are throwing off the planner's
statistics.
7.2 will probably do better
No the query usually returns between 0 and 5 rows. Usually not zero -
most often 1.
-r
On Thu, 10 May 2001 19:47:32 + (UTC), [EMAIL PROTECTED] ("Mitch
Vincent") wrote:
>Does that query really return 9420 rows ? If so, a sequential scan is
>probably better/faster than an index scan..
>
>-Mi
On Thu, May 10, 2001 at 01:22:56PM +, [EMAIL PROTECTED] wrote:
> I vacuum every half hour! Here is the output from EXPLAIN:
>
> NOTICE: QUERY PLAN:
>
> Seq Scan on pa_shopping_cart (cost=0.00..7237.94 rows=9420 width=296)
>
> EXPLAIN
>
> Thanks!
Then try
set enable_seqscan to off;
exp
Does that query really return 9420 rows ? If so, a sequential scan is
probably better/faster than an index scan..
-Mitch
- Original Message -
From: <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>
Sent: Thursday, May 10, 2001 9:22 AM
Subject: Re: Query not using index
> I vacuum every half
I vacuum every half hour! Here is the output from EXPLAIN:
NOTICE: QUERY PLAN:
Seq Scan on pa_shopping_cart (cost=0.00..7237.94 rows=9420 width=296)
EXPLAIN
Thanks!
On Thu, 10 May 2001 18:19:16 + (UTC),
[EMAIL PROTECTED] (Stephan Szabo) wrote:
>
>Have you vacuum analyzed recently and
10 matches
Mail list logo
