In <2.0.4, we logged stack traces for all the exceptions and everyone
complained about it as they didn't want stack traces for the
exceptions they were throwing from their implementation as they are
rightfully supposed to be mapped to faults. Logs were filling very
fast, etc... Everyone hated it.
In 2.0.5 we changed it so all the exceptions thrown from the
implementation only got a single log line. However, I think we went
too far here. Most likely, we should only do this for CHECKED
exceptions. For the unchecked exceptions (like your NPE), we
probably need the full stack trace. I'll change the code over to do
that. That should hopefully come closer to making everyone happy, or
at least a good compromise. (I hope)
Dan
On May 22, 2008, at 7:00 AM, John-M Baker wrote:
Hello,
CXF is swallowing exceptions:
22-May-2008 11:59:29 org.apache.cxf.phase.PhaseInterceptorChain
doIntercept
INFO: Application has thrown exception, unwinding now:
java.lang.NullPointerException: null
I think it should provide a full stack trace for all throwable
exceptiosn,
because when they happen in production (which they will!), log
levels will
be set to INFO and if it's reported as anything else then it is
missed.
John Baker
--
Web SSO
IT Infrastructure
Deutsche Bank London
URL: http://websso.cto.gt.intranet.db.com
---
This e-mail may contain confidential and/or privileged information.
If you are not the intended recipient (or have received this e-mail
in error) please notify the sender immediately and delete this e-
mail. Any unauthorized copying, disclosure or distribution of the
material in this e-mail is strictly forbidden.
Please refer to http://www.db.com/en/content/eu_disclosures.htm for
additional EU corporate and regulatory disclosures.
---
Daniel Kulp
[EMAIL PROTECTED]
http://www.dankulp.com/blog
