[FELIX-5959] Exception on changecount timer timeout

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Fix Version/s: scr-2.1.12
• Component/s: Declarative Services (SCR)
• Assignee: Raymond Augé
• Reporter: Raymond Augé
• Created: 10/Oct/18 16:35
• Updated: 22/Oct/18 11:05
• Resolved: 11/Oct/18 16:42

Description

When running a large system with complex and deeply nested component graph, the changecount timer can time out resulting in exceptions like:

    [junit] java.lang.IllegalStateException: Timer already cancelled.
    [junit] 	at java.util.Timer.sched(Timer.java:397)
    [junit] 	at java.util.Timer.schedule(Timer.java:193)
    [junit] 	at org.apache.felix.scr.impl.ComponentRegistry.updateChangeCount(ComponentRegistry.java:722)
    [junit] 	at org.apache.felix.scr.impl.BundleComponentActivator.updateChangeCount(BundleComponentActivator.java:777)
    [junit] 	at org.apache.felix.scr.impl.manager.AbstractComponentManager.setState(AbstractComponentManager.java:1420)
    [junit] 	at org.apache.felix.scr.impl.manager.AbstractComponentManager.deactivateInternal(AbstractComponentManager.java:806)
    [junit] 	at org.apache.felix.scr.impl.manager.DependencyManager.deactivateComponentManager(DependencyManager.java:2330)
    [junit] 	at org.apache.felix.scr.impl.manager.DependencyManager.access$400(DependencyManager.java:56)
    [junit] 	at org.apache.felix.scr.impl.manager.DependencyManager$SingleStaticCustomizer.removedService(DependencyManager.java:1125)
    [junit] 	at org.apache.felix.scr.impl.manager.DependencyManager$SingleStaticCustomizer.removedService(DependencyManager.java:1007)
    [junit] 	at org.apache.felix.scr.impl.manager.ServiceTracker$Tracked.customizerRemoved(ServiceTracker.java:1242)
    [junit] 	at org.apache.felix.scr.impl.manager.ServiceTracker$Tracked.customizerRemoved(ServiceTracker.java:1137)
    [junit] 	at org.apache.felix.scr.impl.manager.ServiceTracker$AbstractTracked.untrack(ServiceTracker.java:997)
    [junit] 	at org.apache.felix.scr.impl.manager.ServiceTracker$Tracked.serviceChanged(ServiceTracker.java:1176)
    [junit] 	at org.apache.felix.scr.impl.BundleComponentActivator$ListenerInfo.serviceChanged(BundleComponentActivator.java:125)
    [junit] 	at org.eclipse.osgi.internal.serviceregistry.FilteredServiceListener.serviceChanged(FilteredServiceListener.java:109)
    [junit] 	at org.eclipse.osgi.internal.serviceregistry.ServiceRegistry.publishServiceEventPrivileged(ServiceRegistry.java:891)
    [junit] 	at org.eclipse.osgi.internal.serviceregistry.ServiceRegistry.publishServiceEvent(ServiceRegistry.java:804)
    [junit] 	at org.eclipse.osgi.internal.serviceregistry.ServiceRegistrationImpl.unregister(ServiceRegistrationImpl.java:222)
    [junit] 	at org.eclipse.osgi.internal.serviceregistry.ServiceRegistry.unregisterServices(ServiceRegistry.java:668)
    [junit] 	at org.eclipse.osgi.internal.framework.BundleContextImpl.close(BundleContextImpl.java:96)
    [junit] 	at org.eclipse.osgi.internal.framework.EquinoxBundle.stopWorker0(EquinoxBundle.java:1018)
    [junit] 	at org.eclipse.osgi.internal.framework.EquinoxBundle$EquinoxModule.stopWorker(EquinoxBundle.java:359)
    [junit] 	at org.eclipse.osgi.container.Module.doStop(Module.java:636)
    [junit] 	at org.eclipse.osgi.container.Module.stop(Module.java:498)
    [junit] 	at org.eclipse.osgi.container.ModuleContainer$ContainerStartLevel.decStartLevel(ModuleContainer.java:1723)
    [junit] 	at org.eclipse.osgi.container.ModuleContainer$ContainerStartLevel.doContainerStartLevel(ModuleContainer.java:1642)
    [junit] 	at org.eclipse.osgi.container.ModuleContainer$ContainerStartLevel.dispatchEvent(ModuleContainer.java:1555)
    [junit] 	at org.eclipse.osgi.container.ModuleContainer$ContainerStartLevel.dispatchEvent(ModuleContainer.java:1)
    [junit] 	at org.eclipse.osgi.framework.eventmgr.EventManager.dispatchEvent(EventManager.java:230)
    [junit] 	at org.eclipse.osgi.framework.eventmgr.EventManager$EventThread.run(EventManager.java:340)

There are two issues:

• the timeout is not configurable (hard coded to 5000L)
• the exception is not handled in a safe way (i.e. logged to an appropriate log which ends up in the system log)