Re: Dynamic Partitioning using Segment Visibility Maps

On Tue, Jan 08, 2008 at 01:08:52AM +0100, Markus Schiltknecht wrote:
>
> Uh, which key are you talking about? AFAIU Simon's proposal, he suggests
> maintaining min/max values for all columns of the table.

Right, but I think that's just because that approach is automatable. Only
some use cases are going to be approproate to this.

> Yeah, and if only *one* tuple in the 1G segment has:
>
> some_date <= '1998-12-31' OR some_date >= '2001-01-01'
>
> Segment Exclusion can't exclude that segment. That's all I'm saying.

Correct.

> Huh? I'm certainly not the one asking for it. Quite the opposite, I'm
> warning from over-estimating the use of SE.

Right; I think one should be clear that there are many -- maybe most --
uses of PostgreSQL where the proposal will be of no use. I just think we
need to be clear that for the areas where it _can_ be useful, it could be
very useful indeed.

A