Index: [Article Count Order] [Thread]

Date:  Sun, 17 Sep 2006 19:29:18 -0400
From:  "Brian N. Smith" <brian (at mark) nuonce.net>
Subject:  [coba-e:07030] Re: Weird logs
To:  <coba-e (at mark) bluequartz.org>
Message-Id:  <014a01c6dab1$194a3cc0$1e64a8c0@moe>
References:  <004901c6daa7$9f4c01f0$6400a8c0@YOUR4105E587B6>
X-Mail-Count: 07030

Darrell &Everyone else who is wondering.

The reason that the logs in /var/log go to '0' are because Logrotate 
is broken, duh.

Now, it can generally be fixed, but you need to figure out wtf is 
causing it to break.  The best way is to run it manually and see what 
the output says.

/usr/sbin/logrotate /etc/logrotate.conf

The /etc/logrotate.conf should be very non complex.

The following:
/var/log/boot.log
/var/log/cron
/var/log/messages
/varl/log/secure

These files are created by "syslog".  If you do;

service syslog restart

That should be good enough to get them working again, but IT won't fix 
it 100%.  You should NOT have to do that every day.

The error that Darrell got;

/etc/cron.daily/logrotate:
error: error running prerotate script, leaving old log in place

Doesn't tell us jack shit, accept that something broke.

You need to run the "logrotate" command above and see what the output 
says.  The most common I have seen has been this;

error: httpd:1 duplicate log entry for /var/log/httpd/access_log

That is easily by doing this;
rm -f /etc/logrotate.d/httpd

Please post the error