MIME-Version: 1.0
References: <b3d80961-c2e5-38cc-6a32-61886cdf766d@gmail.com>
In-Reply-To: <b3d80961-c2e5-38cc-6a32-61886cdf766d@gmail.com>
From: Dilip Kumar <dilipbalaut@gmail.com>
Date: Sat, 4 Nov 2023 14:22:01 +0530
Message-ID: <CAFiTN-v1yDvU=X+hwfJ+55=sbgDj=_kuvbduEG-F7=BjpWcnuw@mail.gmail.com>
Subject: Re: Parallel Bitmap Heap Scan reports per-worker stats in EXPLAIN ANALYZE
To: David Geier <geidav.pg@gmail.com>
Cc: PostgreSQL Developers <pgsql-hackers@lists.postgresql.org>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Archived-At: <https://www.postgresql.org/message-id/CAFiTN-v1yDvU%3DX%2BhwfJ%2B55%3DsbgDj%3D_kuvbduEG-F7%3DBjpWcnuw%40mail.gmail.com>
Precedence: bulk

On Fri, Jan 20, 2023 at 2:04=E2=80=AFPM David Geier <geidav.pg@gmail.com> w=
rote:
>
> Hi hackers,
>
> EXPLAIN ANALYZE for parallel Bitmap Heap Scans currently only reports
> the number of heap blocks processed by the leader. It's missing the
> per-worker stats. The attached patch adds that functionality in the
> spirit of e.g. Sort or Memoize. Here is a simple test case and the
> EXPLAIN ANALYZE output with and without the patch:
>

> With the patch:
>
>   Gather (actual rows=3D99501 loops=3D1)
>     Workers Planned: 2
>     Workers Launched: 2
>     ->  Parallel Bitmap Heap Scan on foo (actual rows=3D33167 loops=3D3)
>           Recheck Cond: ((col0 > 900) OR (col1 =3D 1))
>           Heap Blocks: exact=3D98
>           Worker 0:  Heap Blocks: exact=3D171 lossy=3D0
>           Worker 1:  Heap Blocks: exact=3D172 lossy=3D0


else
  {
+ if (planstate->stats.exact_pages > 0)
+    appendStringInfo(es->str, " exact=3D%ld", planstate->stats.exact_pages=
);
+ if (planstate->stats.lossy_pages > 0)
+ appendStringInfo(es->str, " lossy=3D%ld", planstate->stats.lossy_pages);
      appendStringInfoChar(es->str, '\n');
  }
  }
....
+ for (int n =3D 0; n < planstate->shared_info->num_workers; n++)
+ {
....
+ "Heap Blocks: exact=3D"UINT64_FORMAT" lossy=3D" INT64_FORMAT"\n", +
si->exact_pages, si->lossy_pages);

Shouldn't we use the same format for reporting exact and lossy pages
for the actual backend and the worker? I mean here for the backend you
are showing lossy pages only if it is > 0 whereas for workers we are
showing 0 lossy pages as well.

--=20
Regards,
Dilip Kumar
EnterpriseDB: http://www.enterprisedb.com