Re: Notice lock waits

From: Jeff Janes <jeff(dot)janes(at)gmail(dot)com>
To: Jim Nasby <Jim(dot)Nasby(at)bluetreble(dot)com>
Cc: pgsql-hackers <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: Notice lock waits
Date: 2016-08-31 22:59:02
Message-ID: CAMkU=1ygG4r9jw7t2bk5ZKgb3vOtfvfCkj7NPQporAb8Mr-m6A@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On Tue, Aug 9, 2016 at 5:17 PM, Jim Nasby <Jim(dot)Nasby(at)bluetreble(dot)com> wrote:

> On 8/5/16 12:00 PM, Jeff Janes wrote:
>
>> So I created a new guc, notice_lock_waits, which acts like
>> log_lock_waits but sends the message as NOTICE so it will show up on
>> interactive connections like psql.
>>
>
> I would strongly prefer that this accept a log level instead of being
> hard-coded to NOTICE. The reason is that I find the NOTICE chatter from
> many DDL commands to be completely worthless (looking at you %TYPE),

Perhaps we should do something about those notices? In 9.3 we removed ones
about adding implicit unique indexes to implement primary keys, and I think
that that was a pretty good call.

> so I normally set client_min_messages to WARNING in DDL scripts. I can
> work on that patch; would it essentially be a matter of changing
> notice_lock_waits to int lock_wait_level?

How would it be turned off? Is there a err level which would work for
that? And what levels would non-superusers be allowed to set it to?

And, I'd be happy if you were to work on a patch to implement it.

Cheers,

Jeff

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Erik Rijkers 2016-08-31 23:04:00 Re: Logical Replication WIP
Previous Message Corey Huinker 2016-08-31 22:31:44 Re: [Patch] Temporary tables that do not bloat pg_catalog (a.k.a fast temp tables)