Re: Table-level log_autovacuum_min_duration

From: Michael Paquier <michael(dot)paquier(at)gmail(dot)com>
To: Naoya Anzai <anzai-naoya(at)mxu(dot)nes(dot)nec(dot)co(dot)jp>
Cc: PostgreSQL mailing lists <pgsql-hackers(at)postgresql(dot)org>, Akio Iwaasa <iwaasa(at)mxs(dot)nes(dot)nec(dot)co(dot)jp>
Subject: Re: Table-level log_autovacuum_min_duration
Date: 2015-02-19 06:32:17
Message-ID: CAB7nPqStnDt5jd1uvdfrsocpRVspArb1wLrkV79odX9AQvthJA@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On Thu, Feb 19, 2015 at 2:13 PM, Naoya Anzai
<anzai-naoya(at)mxu(dot)nes(dot)nec(dot)co(dot)jp> wrote:
> As a result, I think you should not delete VACOPT_VERBOSE.

In v8 it is not deleted. It is still declared, and its use is isolated
in gram.y, similarly to VACOPT_FREEZE.

> According to the last mail I have posted, the difference of
> manual-vacuum log and auto-vacuum log exists clearly.

Did you test the latest patch v8? I have added checks in it to see if
a process is an autovacuum worker to control elevel and the extra logs
of v7 do not show up.

> So, at least you should not touch the mechanism of VACOPT_VERBOSE
> until both vacuum log formats are unified to a same format.

If you mean that we should have the same kind of log outputs for
autovacuum and manual vacuum, I think that this is not going to
happen. Autovacuum entries are kept less verbose on purpose, contract
that v7 clealy broke.

> If you agree my think, please undo your removing VACOPT_VERBOSE work.

Well, I don't agree :) And I am guessing that you did not look at v8
as well. Centralizing the control of logs using log_min_duration is
more extensible than simply having VACOPT_VERBOSE.
--
Michael

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Michael Paquier 2015-02-19 06:40:35 Re: [REVIEW] Re: Compression of full-page-writes
Previous Message Noah Misch 2015-02-19 06:19:26 Re: assessing parallel-safety