Обсуждение: Wanted: way to limit logging

Поиск
Список
Период
Сортировка

Wanted: way to limit logging

От
"Jeff Boes"
Дата:
We've enabled logging from our database during development, using the
option -l on the pg_ctl command.  However, we get a lot of useless data:

   NOTICE:  InvalidateSharedInvalid: cache state reset

shows up a lot, and whenever an INSERT statement is executed, we get the
statement as well as all the values.  Is there a way to reduce the
information spewed to the log? Ideally, we'd like to see only ERROR and
DEBUG messages, not NOTICE; for statements, we'd like to see the SQL
code, but have no interest in the values actually applied (some of our
INSERT statements receive 10K of data).


--
Jeff Boes                                             vox 616.226.9550
Database Engineer                                     fax 616.349.9076
Nexcerpt, Inc.                                      jboes@nexcerpt.com

Re: Wanted: way to limit logging

От
Herbert Rabago Ambos
Дата:
start postgres with

     -d 0

option,  this time it would only display limited info such as DEBUG,
NOTICEs, ERROR: ON LINE NUM's... or check postgresql.conf file for more
option.  In my configuration, i use

    silent_mode = false

    log_connections = true
    log_timestamp = true
    log_pid = true

    debug_level = 0 #equivalent to -d 0

to monitor the logs.


Herbert

/**
 * Database Administrator/Faculty
 * MSU-Iligan Institute of Technology
 * Philippines
 */

On Wed, 10 Oct 2001, Jeff Boes wrote:

> We've enabled logging from our database during development, using the
> option -l on the pg_ctl command.  However, we get a lot of useless data:
>
>    NOTICE:  InvalidateSharedInvalid: cache state reset
>
> shows up a lot, and whenever an INSERT statement is executed, we get the
> statement as well as all the values.  Is there a way to reduce the
> information spewed to the log? Ideally, we'd like to see only ERROR and
> DEBUG messages, not NOTICE; for statements, we'd like to see the SQL
> code, but have no interest in the values actually applied (some of our
> INSERT statements receive 10K of data).
>
>
>