On Tue, Sep 03, 2019 at 11:04:43AM +1200, Thomas Munro wrote:
On Tue, Sep 3, 2019 at 5:20 AM Tomas Vondra
wrote:
FWIW it's not clear to me why the cost would need to be recomputed after
constructing the parallel version of the plan? My understanding is that
the idea is to do cost-based planning
On Tue, Sep 3, 2019 at 5:20 AM Tomas Vondra
wrote:
> FWIW it's not clear to me why the cost would need to be recomputed after
> constructing the parallel version of the plan? My understanding is that
> the idea is to do cost-based planning for the serial plan, and then just
> "mechanically" constr
On Mon, Sep 02, 2019 at 02:19:15PM +1200, Thomas Munro wrote:
On Sat, Aug 24, 2019 at 3:19 AM Tomas Vondra
wrote:
> Although “The Wei Hong Optimizer” was designed in the context of
> Postgres, it became the standard approach for many of the parallel
> query optimizers in industry."
I assume th
On Sat, Aug 24, 2019 at 3:19 AM Tomas Vondra
wrote:
> > Although “The Wei Hong Optimizer” was designed in the context of
> > Postgres, it became the standard approach for many of the parallel
> > query optimizers in industry."
>
> I assume this quote is from 30 years ago. I wonder if the claim is
On Thu, Aug 22, 2019 at 11:41:45AM +1200, Thomas Munro wrote:
Hello,
After rereading some old papers recently, I wanted to share some
thoughts about XPRS and modern PostgreSQL. XPRS stood for "eXtended
Postgres on RAID and Sprite", and was a research project done nearly
three deca
Hello,
After rereading some old papers recently, I wanted to share some
thoughts about XPRS and modern PostgreSQL. XPRS stood for "eXtended
Postgres on RAID and Sprite", and was a research project done nearly
three decades ago at Berkeley by the POSTGRES group working with
operat
