cxf-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Clark, Chris" <>
Subject Services provided by DOSGi
Date Thu, 25 Feb 2010 22:09:28 GMT
Hi there,

I was just wondering, after struggling to incorporate CXF DOSGI into a long-running 'playground'
project of mine, whether it is possible to use the DOSGI system provided with CXF for other
front protocols - so can OSGi services be exposed via Hessian/Burlap/RMI/XML-RPC/etc... using
this system.

The whole idea that I had was that we had a load of OSGi services, but would like to be able
to expose them dynamically, without any worry about the front-end protocol (whether it be

Does CXF DOSGI support this? and is there any documentation about it?

Chris Clark  

"Audio, Video, Disco"

-----Original Message-----
From: Daniel Kulp [] 
Sent: Thursday, February 25, 2010 10:51 PM
Cc: easternwahoo
Subject: Re: How to remove ws-security tokens from message, on provider side?

On Thu February 25 2010 10:30:07 am easternwahoo wrote:
> When using ws-security headers with CXF, it is easy to configure logging
> interceptors to log the complete message. Unfortunately, this means the
> ws-security headers are displayed in the log.     Once we have
> authenticated the user's credentials, could we remove the ws-security
> headers from the message, and would this solve the problem?

You would need to write a "different" set of logging interceptors, but that's 
probably not a big deal.   The current logging interceptors operate at the 
stream level which is way before any processing is done.   Since the security 
stuff operates on an SAAJ model, this wouldn't be too hard.   You could write 
an interceptor that lives immediately after the WSS4JInInterceptor stuff, 
grabs the SAAJ object from the message, pulls off the Security header, and 
then calls saaj.writeTo(...) or similar to have it write itself to a log.  Not 
a huge amount of work.

Daniel Kulp

View raw message