Re: Frequent Update Project: Design Overview of HOT Updates

From: Martijn van Oosterhout <kleptog(at)svana(dot)org>
To: Simon Riggs <simon(at)2ndquadrant(dot)com>
Cc: pgsql-hackers(at)postgresql(dot)org
Subject: Re: Frequent Update Project: Design Overview of HOT Updates
Date: 2006-11-09 17:49:45
Message-ID: 20061109174945.GC23352@svana.org
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Nice idea, just one question:

On Thu, Nov 09, 2006 at 05:13:16PM +0000, Simon Riggs wrote:
> Behavioural Characteristics
> ---------------------------
>
> With HOT, it is easily possible that the chain of prior versions spans
> many blocks. The chain always starts with the block of the root tuple
> but possibly includes many overflow relation blocks.
>
> A SeqScan of a HOT table will turn into a large number of
> random accesses to the overflow relation, which could be considerably
> more expensive than sequential I/O. Clearly very large tables would not
> be able to be HOT enabled without additional cost, so we make HOT a
> table-level option: WITH (freqUPDATE = on) [discuss...]

It seems to me that bitmap index scans will get these same
characteristics also, right? The bitmap scan will have to follow the
chain of any possibly matching tuple in any of the blocks that are in
the bitmap, right?

Have a ncie day,
--
Martijn van Oosterhout <kleptog(at)svana(dot)org> http://svana.org/kleptog/
> From each according to his ability. To each according to his ability to litigate.

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Simon Riggs 2006-11-09 17:57:17 Re: Frequent Update Project: Design Overview of HOTUpdates
Previous Message Simon Riggs 2006-11-09 17:13:16 Frequent Update Project: Design Overview of HOT Updates