2012/2/9 Daniel Kulp <[email protected]>:
> On Thursday, February 09, 2012 4:23:11 PM Aki Yoshida wrote:
>> 2012/2/9 Daniel Kulp <[email protected]>:
>> > On Thursday, February 09, 2012 3:13:47 PM Blue Diamond wrote:
>> >> Thanks for your response Glen.
>> >>
>> >> The WSDL is correctly set. And yes we are running inside Karaf.
>> >> We also tried to put the WSDL file in the application classpath so that
>> >> it
>> >> is visible to both CXF & our app and gave the wsdlLocation="myfile.wsdl"
>> >> ... still the same problem. the service starts, but ?wsdl is not
>> >> responding
>> >> correctly and CXF tries to handle it like a SOAP request intended for my
>> >> service.
>> >
>> > Are there policy fragments in the WSDL that CXF may not understand?
>> > Thats
>> > what it looks like the issue is.   Since the wsdl is returned from an
>> > interceptor on the chain, and the policies are processed very early in the
>> > chain, they need to be resolvable.
>>
>> I also suspected this. But I was not sure what the semantic or rule
>> must be for the WSDL retrieval for a particular endpoint that enforces
>> policy assertions. If you can't look up the WSDL, you won't know what
>> assertions your request must satisfy in advance.
>>
>> In other words, shouldn't the WSDL retrieval call be handled
>> differently from the normal soap binding soap request calls?
>
> Well, this is slightly different (and I'm almost wondering if this check
> should be done once at startup).   What this is showing is that there is NO
> SOAP request that can be processed by this server based on the policies it
> has.   Thus, even if the client had a valid wsdl and was able to create a SOAP
> message based on the policies there, the server would still be unable to do
> anything with it as it cannot figure out a policy alternative that it could
> work with.

Dan,
Ah, okay,  that makes sense.
thanks for the explanation.
regards, aki

>
> Like I mentioned, we really should try and dig in and see if the initial
> server side policy calculations can occur at startup.   It's actually a bad
> thing that the server can start up, but not actually do anything.  Something
> to look into.
>
> --
> Daniel Kulp
> [email protected] - http://dankulp.com/blog
> Talend Community Coder - http://coders.talend.com

Reply via email to