Menu
▾
▴
#296 Log file max size is ignored when log load is high
Logging using max file size does not work on Windows
when there is a lot of logging taking place.
We have configured the conf file using the following log configuration:
wrapper.console.format=M
wrapper.console.loglevel=INFO
wrapper.logfile=log.txt
wrapper.logfile.format=M
wrapper.logfile.loglevel=INFO
wrapper.logfile.maxsize=10m
wrapper.logfile.maxfiles=10
wrapper.syslog.loglevel=NONE
wrapper.console.loglevel=DEBUG
wrapper.logfile.loglevel=DEBUG
so expecting the log file to rollover at 10 Mb, but when logging is intensive
we got a log file > 2GB, which was not what we expected.
Hello
Thank you for your email.
I just tested using version 3.5.23 and the log file is rolling after its
size exceed 10M, even when the log file is written in the working directory
as specified in your configuration.
Is it possible that lower in your configuration you have the property
wrapper.logfile.rollmode set to NONE which would disable log rolling?
If you still encounter the issue, please send us your configuration file at
support@tanukisoftware.com.
Best Regards,
Maxime
On Wed, Feb 22, 2017 at 5:51 PM, J. Stolze jstolze10@users.sf.net wrote:
The version we are using is 3.5.25 at least that what's stated in the manifest file.
But this version was not an option in the tracker versions, that's why I choose
the highest available version.
The rolling over works when the log load is not intensive, but when it is it stops.
Manifest-Version: 1.0
Ant-Version: Apache Ant 1.7.1
Created-By: 1.4.2_17-b06 (Sun Microsystems Inc.)
Built-By: leif
Package: org.tanukisoftware.wrapper
Extension-Name: wrapper
Specification-Title: Java Service Wrapper
Specification-Vendor: Tanuki Software, Ltd.
Specification-Version: 3.5.25
Implementation-Title: org.tanukisoftware.wrapper
Implementation-Vendor: Tanuki Software, Ltd.
Implementation-Version: 3.5.25
Hello
As far as I could test, version 3.5.25 is rolling the log files without
problem even under intensive log (wrapper.debug=TRUE &
wrapper.ping.interval=1). On your side, is it systematically falling when
logging is intensive? Could you determine under which circumstances is it
failing exactly? Can you also reproduce the problem with the test sample
application provided with the Wrapper?
Please send us your configuration file so that we can further investigate
this issue.
If the problem persists, you may also try a newer version of the Wrapper.
Version 3.5.30 is the latest version available. Among many improvements, it
fixes a log rolling issue but this should only apply to Unix platforms...
I look forward to your reply.
Best Regards,
Maxime
On Wed, Feb 22, 2017 at 7:44 PM, J. Stolze jstolze10@users.sf.net wrote:
Related
Bugs: #296