[logback-user] Making logging failures fatal.

[Thorbjørn Ravn Andersen]

I need to make logback report problems back as exceptions.

 

Our general deployment scenario is one where programs are restarted every
day, and are tedious to monitor unless they explicitly make themselves
heard.  Hence logfiles are important and a failure in the logging framework
configuration should cause the program to stop there and then.

 

To my knowledge things like “cannot write to log file X/Y/Z” or “bad
configuration” are currently just logged using logbacks internal mechanism
and the application does not know anything about it.

 

I’d prefer not having logback specific code in my application as we strive
to be logging framework neutral, but if it is unavoidable to do this I can
make an exception.

 

Thanks for any help

 

/Thorbjørn 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.qos.ch/pipermail/logback-user/attachments/20120809/22496d9d/attachment.html>