directory-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Niclas Hedhman <nic...@hedhman.org>
Subject Re: [jira] Commented: (DIREVE-171) add log4j config patch
Date Mon, 27 Jun 2005 07:31:23 GMT
On Monday 27 June 2005 14:51, Emmanuel Lecharny wrote:
> I don't think so, but I may be wrong. I do think that debug is for debug
> purpose, when writing code (I mean, when in production, it's too late to
> run in debug mode). So if you have a problem, you should isolate the
> data that produce it, and then, debug it on a special environment,
> setting the level to debug.

Well, it may not be easy to "isolate" without turning debugging on. I am in 
the camp who have far too much debug statements turned on "always" to have a 
chance to figure out "what went wrong" afterwards. Of course, sometimes that 
is not feasible, and that is when the dynamic settings of Level is useful.


> A few years ago, when working on a system that crashed in prod, I
> suggested to put on the debug level. I got 2,5 Gb of logs in half an
> hour. I can tell by experience (I spent hours reading those useless
> lines of logs) that it does not help at all...

:o)  Well, Chainsaw could have made your life a lot easier. Also, it wouldn't 
be that much data if 'selective enabling' is done, such as a handful of 
Loggers at a time.

The "logger.isDebugEnabled()" is cheap enough to be used for each invocation, 
and provides the possibility for dynamic level settings in the logging 
system.


Cheers
Niclas

Mime
View raw message