Also, the *only* related stuff in the boot.log now is:
[alrubinger@localhost Branch_5_x_BootstrapLegacyRemoval]$ cat build/output/jboss-5.1.0.GA/server/default/log/boot.log |grep "AbstractManagedObjectFactory"
| 18:49:38,064 DEBUG [AbstractManagedObjectFactory] Set ICF for: interface org.jboss.beans.metadata.spi.BeanMetaData, to: org.jboss.deployers.plugins.managed.BeanMetaDataICF@7fefa490
| 18:49:38,261 DEBUG [AbstractManagedObjectFactory] Set ICF for: class org.jboss.system.metadata.ServiceMetaData, to: org.jboss.system.deployers.managed.ServiceMetaDataICF@288e85b5
S,
ALR
View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4232380#4232380
Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4232380
In boot.log of Branch_5_x we can clearly see a bunch of entries like:
18:02:25,585 DEBUG [AbstractManagedObjectFactory] Null ManagedObject created for: class org.jboss.beans.metadata.plugins.AbstractPropertyMetaData
...though in the branch where I'm upgrading the bootstrap impl, I'm not seeing MCServer, ServerInfo, ServerConfig registered anymore. These are MC beans that have the correct @ManagementObject annotations. In fact, the statements like those above have disappeared from the boot.log.
Anything jump out at you guys that would kill this processing? Some system property not set, some JAR no longer on the CP, something along those lines?
S,
ALR
View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4232379#4232379
Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4232379
So the answer is yes, and the current BasicKernelBus implementation that the KernelBusRuntimeComponentDispatcher uses has caused an issue described in https://jira.jboss.org/jira/browse/JBAS-6960.
The problem is that once a stop has been issued to a component, its no longer in the installed state and an attempt to call start on the component fails because the BasicKernelBus.execute method is only looking for installed contexts.
We have not properly considered invoking operations that correspond to lifecycle operations. We need to decide how allow for exposed lifecycle operations to be used.
Right now we can know whether a ManagedOperation has been tagged as a lifecycle operation, but there is no information about what context state transition it applies to.
View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4232340#4232340
Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4232340