Re: Change ereport level for QueuePartitionConstraintValidation - Mailing list pgsql-hackers

From David Rowley
Subject Re: Change ereport level for QueuePartitionConstraintValidation
Date
Msg-id CAKJS1f-GDByLEiY=TjvreZz+-tVuhkH6yJeggE-oTLGhN7ci+g@mail.gmail.com
Whole thread Raw
In response to Re: Change ereport level for QueuePartitionConstraintValidation  (Robert Haas <robertmhaas@gmail.com>)
Responses Re: Change ereport level for QueuePartitionConstraintValidation
Re: Change ereport level for QueuePartitionConstraintValidation
List pgsql-hackers
On Tue, 16 Jul 2019 at 03:13, Robert Haas <robertmhaas@gmail.com> wrote:
> I vote for changing it to NOTICE instead of DEBUG1.

Well, there are certainly other DDL commands that spit out NOTICES.

postgres=# create table z (a int);
CREATE TABLE
postgres=# create table x (a int) inherits(z);
NOTICE:  merging column "a" with inherited definition
CREATE TABLE

However, we did get rid of a few of those a while back. In 9.2 we used to have:

postgres=# create table a (a int primary key);
NOTICE:  CREATE TABLE / PRIMARY KEY will create implicit index
"a_pkey" for table "a"

I'm pretty keen for consistency. Having ATTACH PARTITION spit out an
INFO and merge attributes a NOTICE, and SET NOT NULL just a DEBUG1 is
pretty far from consistent.  I wouldn't object to making them all
NOTICE. I've only seen complaints about the INFO one.

Would anyone complain if we made them all INFO?

If we do that should we backpatch the change into PG12.   SET NOT NULL
using a constraint was new there.

-- 
 David Rowley                   http://www.2ndQuadrant.com/
 PostgreSQL Development, 24x7 Support, Training & Services



pgsql-hackers by date:

Previous
From: Tom Lane
Date:
Subject: Re: Parallel Append subplan order instability on aye-aye
Next
From: Bruce Momjian
Date:
Subject: Re: PG 12 draft release notes