April 2008 - jboss-dev-forums - Jboss List Archives

[Design of JBoss Build System] - Re: Build fail on test errors

by pgier

This feature is not available as far as I know. But I agree it would be useful. I created a jira issue for it here: http://jira.codehaus.org/browse/MDEPLOY-76 View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4145166#4145166 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4145166

18 years, 1 month

1
0
0 / 0

[Design of JBoss Profiler] - Re: JBoss Profiler 2 in JBoss AS 4.2.2 - Does not gather any

by justinwalsh

ok - I think I was on a wild good chase with the manifest, although I suspect it does need some tidying up. Looking at the source for: org.jboss.profiler.agent.Agent.redefine(Class[] clz, boolean instrument) This method appears to the be part involved in instrumenting classes. I assume that without instrumentation, we cannot gather stats. Now the confusing part: redefine is called from the same class at three points: addClasses() removeClasses() setEnabled() each of which is called from the handler. I don't want to add and remove classes manually each time, so I'll concern myself with the enabled method call. Take a look at the method:public static void setEnabled(boolean e) { | if (isRepository()) { | List<Class> l = new ArrayList<Class>(); | | for (Class c : instrumentation.getAllLoadedClasses()) { | String className = c.getName().replace(".", "/"); | | boolean added = false; | for (int i = 0; !added && i < includeList.size(); i++) { | String classes = includeList.get(i); | if (className.startsWith(classes)) { | l.add(c); | added = true; | } | } | } | | Class[] clz = new Class[l.size()]; | for (int i = 0; i < l.size(); i++) { | clz = l.get(i); | } | | if (enabled) { | redefine(clz, true); | } else { | redefine(clz, false); | } | } | } Code will never be instrumented if the repository is disabled (which is the default value) No instrumentation = no stats. So I do the following: configure jboss.profiler.properties with: core=asm (javaassist doesn't seem to work) repository=yes Then: 1) start the application server, and run through my test case. I need to do this to get the classes loaded by the classloader (is there a way to circumvent this?) 2) Using the jmx-console/socket client I enable the profiler. This causes the classes which are in the classloader to be reloaded. 3) I then run my test again. 4) stopProfiler 5) getSnapshot 1 voila One big caveat - I had to change the logic in : private static void redefine(Class[] clz, boolean instrument) to handle exception on a per class basis so that if loading a class fails we don't bail out of the whole operation: | if (clz != null) { | for (Class c : clz) { | try { | byte[] buffer = null; | if (instrument) { | buffer = transformer.loadInstrumented(c.getName(), c.getClassLoader()); | } else { | buffer = transformer.loadNonInstrumented(c.getName(), c.getClassLoader()); | } | | if (buffer != null) { | ClassDefinition cd = new ClassDefinition(c, buffer); | l.add(cd); | } | } | catch (Throwable t) { | System.err.println("Failed to load [" + c.getName() + "]: " + t.getMessage()); | } | } | } | Has anyone else had a similar experience or can shed some light on the issues? View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4145160#4145160 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4145160

18 years, 1 month

1
0
0 / 0

[Design of JBoss Build System] - Re: Build fail on test errors

by ALRubinger

"pgier" wrote : Yes. Is it possible to further restrict this such that "mvn deploy" is prohibited unless your local changes are committed AND your local wc is current with "svn info"'s "Last Changed Revision"? Otherwise we're overwriting each other's snapshots. :) S, ALR View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4145156#4145156 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4145156

18 years, 1 month

1
0
0 / 0

[Design of JBoss Build System] - Re: Build fail on test errors

by pgier

Actually "false" is the default behaviour anyway, so you shouldn't have to specify that you want it false. Just don't put it in the surefire config at all and it should fail when there are failures. Carlo, what do you mean "fails if the workspace is unclean"? Do you mean no local uncommitted changes? View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4145153#4145153 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4145153

18 years, 1 month

1
0
0 / 0

[Design of JBoss Remoting, Unified Invokers] - jmx invoker in trunk

by adrian＠jboss.org

I'm wondering why the JMX invoker in trunk is still using the jrmp invoker instead of the unified invoker? http://viewvc.jboss.org/cgi-bin/viewvc.cgi/jbossas/trunk/server/src/resou... View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4145128#4145128 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4145128

18 years, 1 month

1
0
0 / 0

[Design of JBossXB] - Re: Element order exception

by alex.loubyansky＠jboss.com

Sorry, no, those were not logged. Confused with the @XmlAttribute processing. For the changes see http://jira.jboss.com/jira/browse/JBXB-139. Specifically jbossxb/trunk/src/main/java/org/jboss/xb/builder/JBossXBNoSchemaBuilder.java View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4145103#4145103 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4145103

18 years, 1 month

1
0
0 / 0

[Design of JBossXB] - Re: Element order exception

by alesj

"alex.loubyansky(a)jboss.com" wrote : The attributes are bound to the target namespace. You would see it if you enabled trace. You mean this: | 11:33:23,625 TRACE [BeanHandler] startElement {http://java.sun.com/xml/ns/persistence}property bean=org.jboss.xb.builder.runtime.DefaultMapEntry parent=org.jboss.metadata.jpa.spec.PersistenceUnitMetaData@6460907 | "alex.loubyansky(a)jboss.com" wrote : That's a bug. I am committing the tests and the fix to the xb. I copied the Map usage from your test in JBossXB. What's different there, so that it works? View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4145101#4145101 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4145101

18 years, 1 month

1
0
0 / 0

[Design of JBossXB] - Re: Element order exception

by alex.loubyansky＠jboss.com

The attributes are bound to the target namespace. You would see it if you enabled trace. That's a bug. I am committing the tests and the fix to the xb. View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4145099#4145099 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4145099

18 years, 1 month

1
0
0 / 0

[Design of JBoss Build System] - Re: Build fail on test errors

by alesj

Is this not it? | <configuration> | <testFailureIgnore>false</testFailureIgnore> | </configuration> | View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4145086#4145086 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4145086

18 years, 1 month

1
0
0 / 0

[Design of AOP on JBoss (Aspects/JBoss)] - Re: Per Instance Aspects Overhead

by kabir.khan＠jboss.com

Hi Flavia, I think what you suggest sounds fine. Try it out and see if our tests still pass. I committed some code I had lying around to compare JoinPointInfos. Take a look at Advisor.hasSameMethodAspects(). This is completely untested, so please modify as you see fit. View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4145083#4145083 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4145083

18 years, 1 month

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 April 2008