May 2009 - jboss-dev-forums - Jboss List Archives

[Design of AOP on JBoss (Aspects/JBoss)] - Re: Creation of method tables in ClassAdvisor

by flavia.rainone＠jboss.com

Jaikiran wrote : What is the purpose of this code? I guess, there is some specific scenario where this comes into picture Yes, there is a scenario where this comes into picture. If you navigate through the call hierarchy, you will see that this method is being invoked by both createMethodTables and populateMethodTables, but you will see that createMethodTables invokes populateMethodTables for the superclass of the advised classe: protected void createMethodTables() | throws Exception | { | initAdvisedMethodsMap(); | populateMethodTables(clazz.getSuperclass()); | addDeclaredMethods(clazz); | } Plus, populateMethodTables is recursive on the superclass: private void populateMethodTables(Class<?> superclass) | throws Exception | { | if (superclass == null) return; | if (superclass.equals(Object.class)) return; | | populateMethodTables(superclass.getSuperclass()); | ... | So, we will first verify if the superclass has the unadvised method, if it has, we will use it. There are specific scenarios where the superclass may have the unadvised method, such as when there is a method in the superclass that matches a pointcut. In this scenario, the invocation to getDeclaredMethod that you mentioned is not going to fail. View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4230855#4230855 Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4230855

16 years, 11 months

1
0
0 / 0

[Design of JBoss ESB] - Re: JmsConnectionPool: Allowing control over the number of S

by Kevin.Conner＠jboss.com

"tfennelly" wrote : ... what I meant was that this seems like a bug to me (JmsSession.releaseResources aside), because it seems to be happening at the wrong time. Shouldn't the resources be released/cleaned-up before the session is returned to the "free" pool? Shouldn't returning it to the pool be the last step in the process. But the method is synchronized and, with the exception of the releaseResources behaviour, will be atomic. Kev View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4230853#4230853 Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4230853

16 years, 11 months

1
0
0 / 0

[Design of JBoss ESB] - Re: DefaultJMSPropertiesSetter, DefaultESBPropertiesSetter a

by Kevin.Conner＠jboss.com

"beve" wrote : This is to support pass-through of JMS properties. Previously we just passed through all properties which caused problems with some JMS Providers. But should this not be the responsibility of the gateway and EPR to handle? "beve" wrote : The issue we had was when working with IBMQ in combination with JBossMQ. JBossMQ throws an exception if a provider specific jms property, one starting with "JMS_verndor_name" was not recognized as it's own. This was seen with a single provider, oracle, we were not passing these between different providers. "beve" wrote : Well this might make our life easier but perhaps this could hurt performance. These properties could be optimization properties that the provider can take advantage of. Okay, so it sounds like we need a more flexible solution. Do you have any examples of the properties which you are allowed to set? Kev View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4230851#4230851 Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4230851

16 years, 11 months

1
0
0 / 0

[Design of JBoss ESB] - Re: JmsConnectionPool: Allowing control over the number of S

by tfennelly

Just to clarify... when I said... "tfennelly" wrote : The handleCloseSession method returns the session to the "free" pool before releasing it's resources. Is that a potential bug? ... what I meant was that this seems like a bug to me (JmsSession.releaseResources aside), because it seems to be happening at the wrong time. Shouldn't the resources be released/cleaned-up before the session is returned to the "free" pool? Shouldn't returning it to the pool be the last step in the process. View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4230850#4230850 Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4230850

16 years, 11 months

1
0
0 / 0

[Design the new POJO MicroContainer] - Re: DESCRIBE phase - Dependency builders for MC Beans

by flavia.rainone＠jboss.com

Kabir wrote : Changing it would be a big job, but it could be made pluggable. Flavia did some stuff for her masters which we had planned to add to AOP later, which I think should speed this up. I'll let Flavia provide more info on that. Yes, I did some stuff that optimizes the pointcut matching, replacing the matching by a search on a joinpoint graph. This replacement is performed only on dynamic aop operations, because it is cheaper to do the standard way (i.e., iterate through each joinpoint of a class) when loading new classes or when running aopc. There is already a branch with my work integrated, but since this branch was created before we went 2.0.GA, I would need to perform more integration, so I could add this to JBoss AOP trunk. View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4230848#4230848 Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4230848

16 years, 11 months

1
0
0 / 0

[Design of JBoss ESB] - Re: DefaultJMSPropertiesSetter, DefaultESBPropertiesSetter a

by beve

anonymous wrote : why does the ESB aware need to handle property mappings? This is to support pass-through of JMS properties. Previously we just passed through all properties which caused problems with some JMS Providers. The issue we had was when working with IBMQ in combination with JBossMQ. JBossMQ throws an exception if a provider specific jms property, one starting with "JMS_verndor_name" was not recognized as it's own. JBoss Messaging would simply ignore an unrecognized provider property. The use case was a jms gateway that was connected to IBMQ. A JMS Message would in this case contain provider specific properties to IBMQ. By filtering out these properties they would never be seen by JBossMQ. We created this for cases where other JMS Providers might do the same as JBossMQ and added the ability to filter out properties. anonymous wrote : should the DefaultJmsPropertiesSetter also be filtering? Yeah, it probably should. If there is a mix of JMS Provders this will probably cause troubles like the one you are seeing at the moment :( anonymous wrote : should we be ignoring JMS_XXX on a global basis? Well this might make our life easier but perhaps this could hurt performance. These properties could be optimization properties that the provider can take advantage of. /Daniel View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4230846#4230846 Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4230846

16 years, 11 months

1
0
0 / 0

[Design of JBoss ESB] - DefaultJMSPropertiesSetter, DefaultESBPropertiesSetter and f

by Kevin.Conner＠jboss.com

DefaultESBPropertiesSetter now includes the ability to filter some of the properties as they are imported from JMS, but the current implementation is only usable in certain circumstances. One area which cannot use the mapping is the JMSCourier, as the creation of the Setters are defaulted. I have been helping someone get this to work on Oracle AQ (10.2.3) which does not allow JMS_OracleXXX properties to be set (although 11g does). Tracking this through shows that we can filter on the gateway but, when the listener obtains the ESB aware message through the courier, these headers are reintroduced. So the questions I have are - why does the ESB aware need to handle property mappings? - should the DefaultJmsPropertiesSetter also be filtering? - should we be ignoring JMS_XXX on a global basis? Any thoughts? Thanks Kev View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4230834#4230834 Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4230834

16 years, 11 months

1
0
0 / 0

[Design the new POJO MicroContainer] - Re: Naming, @JNDI and OnDemand

by alesj

"alesj" wrote : "adrian(a)jboss.org" wrote : | | We'll maybe I'm dreaming and didn't actually implement it properly (yet ;-) | | | I guess we need something like this: | https://jira.jboss.org/jira/browse/JBMICROCONT-431 View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4230833#4230833 Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4230833

16 years, 11 months

1
0
0 / 0

[Design of AOP on JBoss (Aspects/JBoss)] - Creation of method tables in ClassAdvisor

by jaikiran

Need some input on a couple of things in the org.jboss.aop.ClassAdvisor http://anonsvn.jboss.org/repos/jbossas/projects/aop/branches/Branch_2_1/a.... During initialization and creation of method tables, the ClassAdvisor in it's addDeclaredMethod does this: | protected void addDeclaredMethods(Class<?> superclass) throws Exception | { | Method[] declaredMethods = superclass.getDeclaredMethods(); | for (int i = 0; i < declaredMethods.length; i++) | { | if (ClassAdvisor.isAdvisable(declaredMethods[ i])) | { | long hash = MethodHashing.methodHash(declaredMethods[ i]); | advisedMethods.put(hash, declaredMethods[ i]); | try | { | Method m = declaredMethods[ i]; | Method un = superclass.getDeclaredMethod(ClassAdvisor.notAdvisedMethodName(superclass.getName(), | m.getName()), | m.getParameterTypes()); | un.setAccessible(true); | unadvisedMethods.put(hash, un); | } | catch (NoSuchMethodException ignored) | { | } | } | } | } | | public static String notAdvisedMethodName(String className, | String methodName) | { | return className.replace('.', '$') + "$" + methodName + | NOT_TRANSFORMABLE_SUFFIX; | } | | /** | * Suffix added to unadvised methods. | */ | public static final String NOT_TRANSFORMABLE_SUFFIX = "$aop"; | | | Assuming i have this expression: | <domain name="XXX" extends="YYY" inheritBindings="true"> | <bind pointcut="execution(public * *->*(..))"> | ... | </bind> | | package org.myapp; | | | public class SomeClass | { | public void method1() | { | .. | } | | public void method2() | { | .. | } | } | 1) So the addDeclaredMethod piece of code, above, is going to generate for each method an "unadvised method name" as: | org$myapp$SomeClass$method1$aop | org$myapp$SomeClass$method2$aop | and later on when the | superclass.getDeclaredMethod(ClassAdvisor.notAdvisedMethodName(superclass.getName(), m.getName()),m.getParameterTypes()); | is called it's always going to throw a NoSuchMethodException, isn't it? What is the purpose of this code? I guess, there is some specific scenario where this comes into picture 2) The MethodHashing.methodHash() does some very specific/involved logic. How is it different from a normal method.hashCode()? View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4230825#4230825 Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4230825

16 years, 11 months

1
0
0 / 0

[Design of JBoss ESB] - Re: AS5 changes

by beve

anonymous wrote : Does that sound reasonable? Sounds good to me. I'll do this now so if anyone objects speak up :) Thanks Kev. View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4230820#4230820 Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4230820

16 years, 11 months

1
0
0 / 0

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

jboss-dev-forums May 2009