Jesse Zhang writes:
> For more kicks, I don't even think this is restricted to volatile
> functions only. To stir the pot, it's conceivable that planner might
> produce the following plan
> Seq Scan on pg_temp_3.foo
> Output: foo.a
> Filter: (SubPlan 1)
> SubPlan 1
> -> WindowAgg
>
Hi Tom and Zhenghua,
On Thu, Jul 16, 2020 at 8:18 AM Tom Lane wrote:
>
> Jesse Zhang writes:
> > You are right, no they are not consistent. But Neither plans is
> > incorrect:
>
> > 1. In the first query, it's semantically permissible to evaluate
> > timeofday() for each pair of (c1, c2), and th
Jesse Zhang writes:
> You are right, no they are not consistent. But Neither plans is
> incorrect:
> 1. In the first query, it's semantically permissible to evaluate
> timeofday() for each pair of (c1, c2), and the plan reflects that.
> (Notice that the parallel nature of the plan is just noise h
: Volatile Functions in Parallel Plans
Hi Zhenghua,
On Wed, Jul 15, 2020 at 5:44 AM Zhenghua Lyu wrote:
>
> Hi,
> I test some SQL in the latest Postgres master branch code (we find these
> issues when
> developing Greenplum database in the PR
> https://nam04.safelinks.protectio
Hi Zhenghua,
On Wed, Jul 15, 2020 at 5:44 AM Zhenghua Lyu wrote:
>
> Hi,
> I test some SQL in the latest Postgres master branch code (we find these
> issues when
> developing Greenplum database in the PR
> https://github.com/greenplum-db/gpdb/pull/10418,
> and my colleague come up with the f
___
From: Amit Kapila
Sent: Thursday, July 16, 2020 12:07 PM
To: Zhenghua Lyu
Cc: PostgreSQL Hackers
Subject: Re: Volatile Functions in Parallel Plans
On Wed, Jul 15, 2020 at 6:14 PM Zhenghua Lyu wrote:
>
>
> The first plan:
>
> Finalize Aggregate
>
On Wed, Jul 15, 2020 at 6:14 PM Zhenghua Lyu wrote:
>
>
> The first plan:
>
> Finalize Aggregate
>-> Gather
> Workers Planned: 2
> -> Partial Aggregate
>-> Nested Loop
> Join Filter: (t3.c1 = t4.c1)
> -> Parallel
