I did a vacuum full and an analyse .... I just ran the query again some 20 minutes ago.
Guess what .... It's still running !!
So it's not that much faster for the moment.
I just want to update a single field in one table with a simple value (negative value of another field)
That can not be that hard ...
Or is it the MVCC that is responsible for this ?
It can't be indexes on other tables, right ?
That would be absolutely sick
On 13 Jun 2005, at 18:45, Yves Vindevogel wrote:
Ok, if all 21 are affected, I can understand the problem.
But allow me to say that this is a "functional error"
On 13 Jun 2005, at 18:02, Richard Huxton wrote:
Yves Vindevogel wrote:Met vriendelijke groeten,
I forgot cc
Begin forwarded message:
From: Yves Vindevogel <[EMAIL PROTECTED]>
Date: Mon 13 Jun 2005 17:45:19 CEST
To: Tom Lane <[EMAIL PROTECTED]>
Subject: Re: [PERFORM] Updates on large tables are extremely slow
Yes, but if I update one column, why should PG update 21 indexes ?
There's only one index affected !
No - all 21 are affected. MVCC creates a new row on disk.
--
Richard Huxton
Archonet Ltd
Bien à vous,
Kind regards,
Yves Vindevogel
Implements
<Pasted Graphic 2.tiff>
Mail: [EMAIL PROTECTED] - Mobile: +32 (478) 80 82 91
Kempische Steenweg 206 - 3500 Hasselt - Tel-Fax: +32 (11) 43 55 76
Web: http://www.implements.be
<x-tad-smaller>
First they ignore you. Then they laugh at you. Then they fight you. Then you win.
Mahatma Ghandi.</x-tad-smaller>
---------------------------(end of broadcast)---------------------------Met vriendelijke groeten,
TIP 5: Have you checked our extensive FAQ?
http://www.postgresql.org/docs/faq
Bien à vous,
Kind regards,
Yves Vindevogel
Implements
<<inline: Pasted Graphic 2.tiff>>
Mail: [EMAIL PROTECTED] - Mobile: +32 (478) 80 82 91
Kempische Steenweg 206 - 3500 Hasselt - Tel-Fax: +32 (11) 43 55 76
Web: http://www.implements.be
<x-tad-smaller>
First they ignore you. Then they laugh at you. Then they fight you. Then you win.
Mahatma Ghandi.</x-tad-smaller>
---------------------------(end of broadcast)--------------------------- TIP 9: the planner will ignore your desire to choose an index scan if your joining column's datatypes do not match