https://issues.apache.org/bugzilla/show_bug.cgi?id=52833
--- Comment #13 from Konstantin Kolinko <knst.koli...@gmail.com> 2012-03-20 16:53:34 UTC --- (In reply to comment #8) (In reply to comment #9) AbstractRunMojo#startContainer() creates a "new Catalina()". I do not see it ever calling setParentClassLoader() and Catalina#parentClassLoader defaults to its own classloader. Is Catalina.class.getClassLoader() the one that does not have system CL in its parents chain? > What I don't understand is why in the parent classLoader chain having > SystemClassLoader is mandatory! Anyway as far as WebappClassLoader class goes 1) It has to follow its documentation, [1] in 7.0 and in more details in 5.5, where shared class loader was actually used. There is lookup sequence, "Therefore, from the perspective of a web application, class or resource loading looks in the following repositories, in this order: (...)" - see [1] [1] http://tomcat.apache.org/tomcat-7.0-doc/class-loader-howto.html 2) WebappClassLoader has its own explicit call to getSystemClassLoader() and treats the system class loader specially and separately from its parent, to follow its specific delegating pattern of calling A system CL B webapp CL C parent CL. Therefore for the WebappClassLoader class it does not matter where its "parent CL" is rooted. It does not impact its behaviour. So using shared CL that does not have system CL as a parent does not seriously impact Tomcat. I do not mind a (cl != null) check to be added in the loop in r1242101. It seems that the code in r1242101 is the first one that actually looks into the shared classloader and is outside of a webapp. Such work is usually scoped to the web application. The shared CL is actually used in WebappLoader#createClassLoader(). See also my comments in "Re: r1242101" on dev@. http://markmail.org/message/rgd7tzz5zks6l5wn (In reply to comment #12) > > Caused by: org.apache.catalina.LifecycleException: An invalid Lifecycle > transition was attempted ([before_stop]) for component > [org.apache.catalina.deploy.NamingResources@7f1f91ac] in state [INITIALIZED] Interesting... A small issue though. It tries to stop a component that has not been started? -- Configure bugmail: https://issues.apache.org/bugzilla/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are the assignee for the bug. --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org For additional commands, e-mail: dev-h...@tomcat.apache.org
