On 23 May 2015 09:19:15 BST, Rainer Jung <rainer.j...@kippdata.de> wrote: >Am 22.05.2015 um 16:55 schrieb ma...@apache.org: >> Author: markt >> Date: Fri May 22 14:55:06 2015 >> New Revision: 1681138 >> >> URL: http://svn.apache.org/r1681138 >> Log: >> Log errors processing LifecycleListener events and continue rather >than allowing uncaught exception to propagate >> >> Modified: >> tomcat/trunk/java/org/apache/catalina/util/LifecycleBase.java >> >tomcat/trunk/java/org/apache/catalina/util/LocalStrings.properties >> >> Modified: >tomcat/trunk/java/org/apache/catalina/util/LifecycleBase.java >> URL: >http://svn.apache.org/viewvc/tomcat/trunk/java/org/apache/catalina/util/LifecycleBase.java?rev=1681138&r1=1681137&r2=1681138&view=diff >> >============================================================================== >> --- tomcat/trunk/java/org/apache/catalina/util/LifecycleBase.java >(original) >> +++ tomcat/trunk/java/org/apache/catalina/util/LifecycleBase.java Fri >May 22 14:55:06 2015 >> @@ -117,7 +117,13 @@ public abstract class LifecycleBase impl >> LifecycleEvent event = new LifecycleEvent(this, type, >data); >> LifecycleListener interested[] = listeners; >> for (int i = 0; i < interested.length; i++) { >> - interested[i].lifecycleEvent(event); >> + try { >> + interested[i].lifecycleEvent(event); >> + } catch (Throwable t) { >> + ExceptionUtils.handleThrowable(t); >> + log.error(sm.getString("lifecycleBase.listenerFail", >> + type, interested[i].getClass().getName()), >t); >> + } >> } >> } > >Currently the unit test testWebappListenerConfigureFail() in >TestStandardContext fails due to the context startup with a >FailingLifecycleListener no longer throwing an exception. > >Regards, > >Rainer > > >--------------------------------------------------------------------- >To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org >For additional commands, e-mail: dev-h...@tomcat.apache.org
Feel free to revert my change (I won't be near my laptop for a few days.). The bz issue needs a more careful look. It might end up as won't fix. Throw during start, log during stop might work but I can't remember how this interacts with the error state. Mark
