From: Tom Lane <tgl@sss.pgh.pa.us>
To: Julien Rouhaud <rjuju123@gmail.com>
cc: Fujii Masao <masao.fujii@oss.nttdata.com>, Sergei Kornilov <sk@zsrv.org>,
        "imai.yoshikazu@fujitsu.com" <imai.yoshikazu@fujitsu.com>,
        legrand legrand <legrand_legrand@hotmail.com>,
        "pgsql-hackers@postgresql.org" <pgsql-hackers@postgresql.org>
Subject: Re: Planning counters in pg_stat_statements (using pgss_store)
In-reply-to: <20210725163637.qkslpe24mgzen73t@nol>
References: 
 <CAOBaU_YMDofrfRrA4c_kNfFd4ZtwtoTy1XmnBom7rwyVGeFTDw@mail.gmail.com>
 <8c6449d4-5fd7-f544-182b-a4bb7da1d18c@oss.nttdata.com>
 <20200331060307.e7kypfynz7jous62@nol>
 <ac44fd08-5923-c0f7-742d-952a6f5a14a3@oss.nttdata.com>
 <20200331073321.bhnx2p63gp5px37f@nol>
 <f2b4e098-1b96-e1a8-f601-a0908c9ff8c4@oss.nttdata.com>
 <20200331184217.tf3x534uz4vusgmp@nol>
 <bd27f039-5285-1fb7-5ed4-57f870c8a5a4@oss.nttdata.com>
 <13923268-bad9-80f9-2542-5fd5e8b8c500@oss.nttdata.com>
 <42557.1627229005@sss.pgh.pa.us> <20210725163637.qkslpe24mgzen73t@nol>
Comments: In-reply-to Julien Rouhaud <rjuju123@gmail.com>
	message dated "Mon, 26 Jul 2021 00:36:37 +0800"
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <81234.1627232375.1@sss.pgh.pa.us>
Content-Transfer-Encoding: quoted-printable
Date: Sun, 25 Jul 2021 12:59:35 -0400
Message-ID: <81235.1627232375@sss.pgh.pa.us>
Archived-At: 
 <https://www.postgresql.org/message-id/81235.1627232375%40sss.pgh.pa.us>
Precedence: bulk

Julien Rouhaud <rjuju123@gmail.com> writes:
> On Sun, Jul 25, 2021 at 12:03:25PM -0400, Tom Lane wrote:
>> So AFAICS this test is inherently unstable and there is no code bug
>> to be fixed.  We could drop the "plans" column from this query, or
>> print something approximate like "plans > 0 AND plans <=3D calls".
>> Thoughts?

> I think we should go with the latter.  Checking for a legit value, even =
if it's
> a bit imprecise is still better than nothing.

> Would it be worth to split the query for the prepared statement row vs t=
he rest
> to keep the full "plans" coverage when possible?

+1, the same thought occurred to me later.  Also, if we're making
it specific to the one PREPARE example, we could get away with
checking "plans >=3D 2 AND plans <=3D calls", with a comment like
"we expect at least one replan event, but there could be more".

Do you want to prepare a patch?

			regards, tom lane