0 
0 I was completely stuck with Service Bus 12c project deployment from JDeveloper to the Service Bus run time. Every deployment met with the same fate: Conflicts found during publish – OSB-398016, Error loading the WSDL from the repository: The WSDL is not semantically valid: Failed to read wsdl file from url due to — java.net.MalformedURLException: Unknown protocol: servicebus.
I was completely lost and frustrated – not even a simple hello_world could make it to the server.
Then, Google and Daniel Dias from Link Consulting to the rescue: http://middlewarebylink.wordpress.com/2014/07/17/soa-12c-end-to-end-e2e-tutorial-error-deploying-validatepayment/. He had run into the same problem – and he had a fix for it! Extremely hard to find if you ask me, but fairly easy to apply.
It turns out this is a known bug (18856204). The bug description refers to BPM and SB being installed in the same domain.
The resolution:
Open the Administration Console for the WebLogic Domain. From the Services node, select service OSGi Frameworks:
Click on the bac-svnserver-osgi-framework link. Note: if you run in production mode, you will now first have to create an edit session.
Add felix.service.urlhandlers=false in the Init Properties field for the configuration of this service. Then press the Save button.
If you run in Production Mode, you now have to commit the edit session.
Then, in order have this modification make any difference, you have to restart the WebLogic (Admin) Server.
This resolved the issue for me – a weight was lifted of my shoulders. Thanks to Daniel from Link!
Edit (28th December 2014): See for more details and a patch from Oracle this article by Jan van Zoggel: http://jvzoggel.wordpress.com/2014/10/07/patching-the-oracle-service-bus-12-1-3-unknown-protocol-deployment-error/
Related posts:
SOA Suite 12c: Deploying shared artifacts to Run Time MDS – the official approach and the quick short-cut 
Best practices for implementing an efficient continuous delivery process 
XL Deploy: Simple Case of Custom Deployment 
OpenWorld 2015: 14 Continuous Delivery / DevOps recommended sessions 
XL Deploy: Inception of legacy processes using the rule system
Hi Lucas,
Do we have any locks to stall the deployments which already running on WLS 11g….??
If yes please let me know the scenarios and circumstances…
Thanks,
Mudhu Krishna
Guys, You saved the day!!! was getting gray hair of that error. easy fix 😀
in case my deployment there is no option of osgi framework
and the weblogic server debugging log i am getting the message there is “no protocol” and then referring to some jar i dont know what to do plz suggest something
This solution works fine for my system. Thanks for this valuable information.
It does not work. BPMComposer fail to start with this configuration. I found on the internet a workaround that explain this step must be replaced by the inactivation of OSGI framework on adminserver.
But it does not work anymore !!
Hi,
I opened a SR at Oracle and they changed note for the solution.
12c: Error when Running both BPM and OSB in one Domain, ‘java.net.MalformedURLException’ (Doc ID 1903573.1)
You must take off the felix.service.urlhandlers=false from OSGI Framework start and do this at setStartupEnv.sh
Change from (20 times maybe)
JAVA_OPTIONS=”${JAVA_OPTIONS} -Djava.protocol.handler.pkgs=oracle.mds.net.protocol|oracle.fabric.common.classloaderurl.handler|oracle.fabric.common.uddiurl.handler”
to
JAVA_OPTIONS=”${JAVA_OPTIONS} -Djava.protocol.handler.pkgs=oracle.mds.net.protocol|oracle.fabric.common.classloaderurl.handler|oracle.fabric.common.uddiurl.handler|oracle.bpm.io.fs.protocol|com.bea.wli.sb.resources.url”
Command in Linux
sed ‘s/|oracle.bpm.io.fs.protocol/|oracle.bpm.io.fs.protocol|com.bea.wli.sb.resources.url/g’ setStartupEnv.sh -i
@jpvadell
Thanks for your information Juan, It was very helpfull