Re: [HACKERS] why not parallel seq scan for slow functions

On Thu, Sep 14, 2017 at 3:19 PM, Amit Kapila <amit.kapila16@gmail.com> wrote:
> The attached patch fixes both the review comments as discussed above.

This cost stuff looks unstable:

test select_parallel          ... FAILED

!  Gather  (cost=0.00..623882.94 rows=9976 width=8)
     Workers Planned: 4
!    ->  Parallel Seq Scan on tenk1  (cost=0.00..623882.94 rows=2494 width=8)
  (3 rows)

  drop function costly_func(var1 integer);
--- 112,120 ----
  explain select ten, costly_func(ten) from tenk1;
                                   QUERY PLAN
  ----------------------------------------------------------------------------
!  Gather  (cost=0.00..625383.00 rows=10000 width=8)
     Workers Planned: 4
!    ->  Parallel Seq Scan on tenk1  (cost=0.00..625383.00 rows=2500 width=8)
  (3 rows)