Re: Two different execution plans for similar requests - Mailing list pgsql-performance

From Joby Joba
Subject Re: Two different execution plans for similar requests
Date
Msg-id AANLkTinyvW7Uck8z4_iyZji0Bi8bQ-c=_ze8NsS-7SEf@mail.gmail.com
Whole thread Raw
In response to Re: Two different execution plans for similar requests  (Joby Joba <jobyjoba59@gmail.com>)
Responses Re: Two different execution plans for similar requests  (Maciek Sakrejda <msakrejda@truviso.com>)
List pgsql-performance
Me again !  I have checked this question of 'explain analyze' and I understand now.

When the problem occured I have run a 'EXPLAIN'

I have run the request again to open this case using 'EXPLAIN ANALYZE' but I didn't reproduce the case. That's why I sent the other trace but it doesn't countain the information you are asking for.
So Tomas, you were right !

All apologies !!!!! (I'm confused)


But the problem still not resolved .....


2011/3/1 Joby Joba <jobyjoba59@gmail.com>
Sorry ! The command I use is 'EXPLAIN ANALYZE'
I can't do better ...

2011/3/1 <tv@fuzzy.cz>
> I've already used an 'EXPLAIN ANALYZE' to post the message. So I don't

> clearly understand what you are expecting for, when you tell me to provide
> 'EXPLAIN ANALYZE'  (please excuse me for the misunderstood)

No, you haven't. You've provided 'EXPLAIN' output, but that just prepares
an execution plan and displays it. So it shows just estimates of row
counts etc. and not actual values.

Do the same thing but use 'EXPLAIN ANALYZE' instead of 'EXPLAIN' - it will
run the query and provide more details about it (run time for each node,
actual number of rows etc.).

Anyway the sudden changes of estimated costs are suspicious ...

Tomas



pgsql-performance by date:

Previous
From: Joby Joba
Date:
Subject: Re: Two different execution plans for similar requests
Next
From: Tom Lane
Date:
Subject: Re: inheritance: planning time vs children number vs column number