RTC v4.0 Advisor not registering in process configuration
I am in a bit of a sticky situation here and I am hoping that someone from the community will be able to give me some hints, tips and tricks to overcome this issue.
We are developing some process extensions in the form of advisors and participants. The production environment has RTC deployed on Websphere Application Server. I have configured 2 development environments on my computer, one being a replica of the production configuration and the other one using Tomcat and Derby. Well, the participant I am writing deploys on the server and is registered in the process configuration. I am able to attach it to the work-item save operation and it gets called and is doing what it is supposed to do. The advisor on the other hand is stubborn and has attitude. On Tomcat is deploys fine and it is registered in the process configuration. On WAS on the other hand, it refuses to show up in the process configuration, event though the deployment is executed without error. I can see this in the hidden admin pages that show component status and provision status. So, my question to the community is: Has anyone managed to deploy an advisor successfully on JTS with WAS? If yes, how? What did you do? Did it go smoothly or were some custom configuration required? Did I say HELP? Thank you very much!
showing 5 of 6
show 1 more comments
|
Accepted answer
|
Problem solved! Websphere Application Server caches needed to be cleared and the server restarted. After that the advisor was correctly registered in the process configuration.
For instructions on how to clear the WAS cache, please see the document @ URL : http://www-01.ibm.com/support/docview.wss?uid=swg21607887 Ralph Schoon selected this answer as the correct answer
|
Your answer
Dashboards and work items are no longer publicly available, so some links may be invalid. We now provide similar information through other means. Learn more here.
Comments
and you did the server resetReset command after configuring the advisor plugin?
Yes I meant CCM and the sequence I am doing ritually is :
- server reset
- server stop
- site update
- server start
- check provision status logs
- check component status
- check availability of preconditions and postconditions using Eclipse client.
and the systems are at the same RTC level, correct?
you could have a dependency in your plugin that is ok for the test systems but not for the production system. I had that happen once..
can u find the ccm log file?
The dev environments are both located on my computer and have been provisioned by me. So, yes, they are identical from a software version point of view as well as application configuration.
The ccm.log does not contain any error or indication whatsoever as to whether there might be an error. I have checked that several times.
I developed a number of advisors on Tomcat, and they were deployed successfully to Websphere in production..
the folder names were different for the configuration ini files, but nothing else.
I did not touch the production system
And this is the same for me as far as the participant is concerned. But not the advisor. The funny thing is that the advisor is created by making a copy of the participant project and modifying the plugin, the package and the code to match the requirements of an advisor.