Re: "No transaction in progress" warning

On 07/05/09 16:56, Adrian Klaver wrote:
>> The server startup messages were in this file (which is why I
>> assumed it was the correct one), but for some reason after a
>> restart the logs were created as
>>
>> /var/log/postgresql/postgresql-8.3-main.log.1241706461
[...]
> is done in the zone specified by log_timezone.) <<If no %-escapes are
> present, PostgreSQL will append the epoch of the new log file's
> creation time. For example, if log_filename were server_log, then the
> chosen file name would be server_log.1093827753 for a log starting at
> Sun Aug 29 19:02:33 2004 MST.>> This parameter can only be set in the
> postgresql.conf file or on the server command line.

Thanks, I had missed that in the manual.

I'm beginning to feel incredibly dense now, but this actually brought my
original problem back. When I do specify log_filename and log_directory,
and restart Postgres, the test you suggested ('select 1/0;') shows up as
an error in the log file

/var/log/postgresql/postgresql-8.3-main.log.{timestamp}

When I leave both log_directory and log_filename commented out (my
original settings), then restart postgres, it creates the file

/var/log/postgresql/postgresql-8.3-main.log

This contains three lines about SSL certificates, but the warning from
'select 1/0' will instead be written to this file:

/var/lib/postgresql/8.3/main/pg_log/postgresql-2009-05-07_170932.log

So there are two log files :-|

Is there any way to configure Postgres to always append to the same file
(/var/log/postgresql/postgresql-8.3-main.log), even if it doesn't
contain strftime escapes? I guess it must be possible, because that's
the way it used to work (before I screwed up my kernel memory settings
in an unrelated SNAFU, preventing Postgres from starting on boot).

Thanks for your patience.
- Conrad