Is there a documented (and accepted) bug for this problem? This looks
close, but not reproducable:
http://nagoya.apache.org/bugzilla/show_bug.cgi?id=4090
I'm thinking of moving to DailyRollingFileAppender (with version 1.2.x),
and my few tests have not uncovered any problems. Maybe the bug results
from differences in JREs?!?
Also, I don't see any significant change to DailyRollingFileAppender in
the 1.3 code (although the docs do suggest using the new
org.apache.log4j.rolling.RollingFileAppender), so I'm curious if something
has been "fixed" in 1.3 alpha. Sure would have a better feeling if I knew
why 1.3 solves your problem (so I can make sure I don't have the same
problem with 1.2).
Thanks,
Tom
Yes, quite a lot of people are getting this. There are 2 ways we have
gotten around this:
1. Use log4j1.3-alpha
2. Compile log4j1.2.8 from source but change
DailyRollingFileAppender.rollOver so it contains the code
...
try {
org.apache.log4j.rolling.helpers.Util.rename(fileName,
scheduledFilename);
} catch (org.apache.log4j.rolling.RolloverFailure rfE) {
}
...
It works for us.
Allistair Crossley
> -----Original Message-----
> From: Balasubramanian, Lakshmi
> [mailto:[EMAIL PROTECTED]
> Sent: 26 August 2004 02:34
> To: [EMAIL PROTECTED]
> Subject: DailyRolling not Happening in log4j Urgent
> Importance: High
>
>
> Hi,
>
> I am using log4j and rotating the log file using
> DailyRollingFileAppender at mid night of every day.
>
> This appliation running in WebSphere 4.0.6 and we have
> started recently stop and start of the particular application
> server at 1.30 am every night using shellscript.
>
> once the job started on day onwards,log rotation stopped.
>
> Any body experiening on this?
>
> Thanks & Regards
> Bala L
