Re: show all; - Mailing list pgsql-patches

From Bruce Momjian
Subject Re: show all;
Date
Msg-id 200106021436.f52EaAw08778@candle.pha.pa.us
Whole thread Raw
In response to Re: show all;  (Tom Lane <tgl@sss.pgh.pa.us>)
Responses Re: show all;
Re: show all;
List pgsql-patches
> Bruce Momjian <pgman@candle.pha.pa.us> writes:
> > The issue is that it fills the logs if used a lot:
>
> Well, that's a generic problem with all our "helpful" NOTICEs.
> ROLLBACK is far from being the only sinner.  ISTM if you don't
> like log bloat, the answer is to attack that head-on, not invent
> yet another nonstandard spelling for ROLLBACK.
>
> Some ideas to think about:
>
> 1. Invent SET variables to control the minimum elog severity level
> that gets into the logs or sent to the client, rather than hard-wiring
> these levels at DEBUG and NOTICE respectively.
>
> 2. Divide existing NOTICEs into more than one severity level.  We have
> a lot of messages like 'CREATE TABLE will create a sequence' that might
> be helpful for development but are just plain nuisances in production
> applications.  It would be nice to turn off the "picky" notices at will.
> #1 would take care of this if they had a lower severity level than other
> notices.

The problem is that they don't want to change the log levels.  They are
using BEGIN;COMMIT; as an automatic thing in the PHP interface to close
any open transaction before passing the persistent connection to another
user.  This is a special case.

--
  Bruce Momjian                        |  http://candle.pha.pa.us
  pgman@candle.pha.pa.us               |  (610) 853-3000
  +  If your life is a hard drive,     |  830 Blythe Avenue
  +  Christ can be your backup.        |  Drexel Hill, Pennsylvania 19026

pgsql-patches by date:

Previous
From: Tom Lane
Date:
Subject: Re: show all;
Next
From: Tom Lane
Date:
Subject: Re: show all;