On Wed, 2007-01-10 at 10:46 -0500, Tom Lane wrote:
> "Simon Riggs" <[EMAIL PROTECTED]> writes:
> > Presumably we'd need to teach the Materialize node to pass straight
> > through when the node does not receive any of EXEC_FLAG_MARK,
> > EXEC_FLAG_REWIND or EXEC_FLAG_BACKWARD.
>
> It does that alre
"Simon Riggs" <[EMAIL PROTECTED]> writes:
> Presumably we'd need to teach the Materialize node to pass straight
> through when the node does not receive any of EXEC_FLAG_MARK,
> EXEC_FLAG_REWIND or EXEC_FLAG_BACKWARD.
It does that already.
> The Materialize node would have to communicate with the
On Wed, 2007-01-10 at 10:10 -0500, Tom Lane wrote:
> "Simon Riggs" <[EMAIL PROTECTED]> writes:
> > Merge Joins require us to potentially Mark and Restore positions in the
> > tuples arriving from executor sub-nodes.
>
> I came across an old note to myself suggesting that we handle this by
> interp
"Simon Riggs" <[EMAIL PROTECTED]> writes:
> Merge Joins require us to potentially Mark and Restore positions in the
> tuples arriving from executor sub-nodes.
I came across an old note to myself suggesting that we handle this by
interposing a Materialize node, and then teaching Material that if it
Simon Riggs wrote:
> On Sat, 2007-01-06 at 17:06 -0500, Bruce Momjian wrote:
> > I saw no replies to this.
>
> Me neither.
>
> I take it that means its a very good idea and we should add a TODO
Added to TODO:
* Improve merge join performance by allowing mark/restore of
tuple sources
http:/
On Mon, Jan 08, 2007 at 10:37:25AM +, Heikki Linnakangas wrote:
> >Simon Riggs wrote:
> >>Implementing the variable mark/restore buffer as a dumb Tuplestore would
> >>mean that the space usage of the Sort could in worst case go as high as
> >>x2 total space. The worst case is where the inner sc
Bruce Momjian wrote:
I saw no replies to this.
Sounds like a good idea to me.
(further comments below)
Simon Riggs wrote:
Implementing the variable mark/restore buffer as a dumb Tuplestore would
mean that the space usage of the Sort could in worst case go as high as
x2 total space. The wors
On Sat, 2007-01-06 at 17:06 -0500, Bruce Momjian wrote:
> I saw no replies to this.
Me neither.
I take it that means its a very good idea and we should add a TODO
--
Simon Riggs
EnterpriseDB http://www.enterprisedb.com
---(end of broadcast)-
I saw no replies to this.
---
Simon Riggs wrote:
> Merge Joins require us to potentially Mark and Restore positions in the
> tuples arriving from executor sub-nodes.
>
> This currently means that if the tuples arrive from a
