----- Original Message -----
From: "Scott Marlowe" <scott.marlowe@gmail.com>
To: "Ati Rosselet" <ati.rosselet@gmail.com>
Cc: <pgsql-general@postgresql.org>
Sent: Saturday, October 25, 2008 12:04 PM
Subject: Re: [GENERAL] again...
> On Sat, Oct 25, 2008 at 9:12 AM, Ati Rosselet <ati.rosselet@gmail.com>
> wrote:
>> I'm still getting a lot of these entries in my eventlog whenever I have a
>> reasonably large amount of logging:
>>
>> Event Type: Error
>> Event Source: PostgreSQL
>> Event Category: None
>> Event ID: 0
>> Date: 10/22/2008
>> Time: 9:36:28 AM
>> User: N/A
>> Computer: ----------
>> Description:
>> could not write to log file: Bad file descriptor
>>
>> production server 8.3.3 (tried with 8.3.4 - same thing), win 2003 server.
>> I'd send this to psql-bugs, but I can't figure out how to reproduce it..
>> and
>> apparently noone else has seen this? Or is noone else running postgres
>> on
>> win2003??? (survey says??)
>>
>> The timing of each appears to be when log is rolled over (size limited to
>> 10MB) and postgres attempts to write a log entry at the same time:
>>
>> Any ideas? help? How to proceed???
>
> Maybe you've got a virus scanner that's locking your log files on you?
Hi Ati,
Inconsistent misbehavior is always fun ;). You may need to increase the
probability that the misbehavior shows itself. I'd change the 10 MB limit
to 1MB and document your findings. Additionally, you may take the size
limit out of the picture and lower the time limit to something ridiculously
low. (120 minutes?)
Just a suggestion... Good luck!
Jeff
>
> --
> Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
> To make changes to your subscription:
> http://www.postgresql.org/mailpref/pgsql-general
>
