[JBoss-dev] [ jboss-Bugs-643732 ] MainDeployer can't undeploy bad EARs
Bugs item #643732, was opened at 2002-11-25 20:36 You can respond by visiting: https://sourceforge.net/tracker/?func=detailatid=376685aid=643732group_id=22866 Category: JBossServer Group: CVS HEAD Status: Closed Resolution: Fixed Priority: 5 Submitted By: Michael Newcomb (mnewcomb) Assigned to: Nobody/Anonymous (nobody) Summary: MainDeployer can't undeploy bad EARs Initial Comment: Facts about MainDeployer: - all deployments (DeploymentInfo), both good or bad, go into a List called deploymentList when MainDeployer.init() completes (in a finally clause) - all deployments that that do not fail their DeploymentInfo.deployer.init() calls go into a Map called deploymentMap (DeploymentInfo.url to DeploymentInfo) This can cause problems when the DeploymentInfo.deployer throws a DeploymentException. The DeploymentInfo does not get inserted into the deploymentMap and can therefore never be removed. Any call to MainDeployer.undeploy() fails quietly because there is no entry inside of deploymentMap (because the DeploymentInfo.deployer threw an exception skipping that deploymentMap.put() call). All attempts to redeploy the application (even if the errors were corrected) fall on deaf ears and the new DeploymentInfos just keep getting put into the deploymentList List (small memory leak). A simple test case to demonstrate this error: 1. create an empty EAR file (nothing in the jar file) 2. deploy it 3. watch EARDeployer throw a DeploymentException stating that META-INF/application.xml doesn't exist 4. delete the EAR file (nothing happens) 5. try to re-deploy the EAR file (valid or not) and you get a new exception about how it is waiting for an appropriate deployer The only way to redeploy the application is to restart JBoss. I fixed this problem by adding some additional code to MainDeployer.undeploy. If undeploy can't find the DeploymentInfo object in deploymentMap, it simply searches through deploymentList looking for a match. I've fixed this in CVS, just entering it here so we can track the bug. The above test should now always throw only the no META-INF/application.xml file found exception. Michael -- Comment By: Christian Riege (lqd) Date: 2002-11-26 09:00 Message: Logged In: YES user_id=176671 marking CLOSED -- Comment By: Michael Newcomb (mnewcomb) Date: 2002-11-25 20:46 Message: Logged In: YES user_id=382427 I fixed this in CVS. Can someone close this? I don't seem to have close priviledges. Thanks, Michael -- You can respond by visiting: https://sourceforge.net/tracker/?func=detailatid=376685aid=643732group_id=22866 --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
[JBoss-dev] [ jboss-Bugs-643673 ] ClassCastException
Bugs item #643673, was opened at 2002-11-25 18:40 You can respond by visiting: https://sourceforge.net/tracker/?func=detailatid=376685aid=643673group_id=22866 Category: JBossServer Group: None Status: Closed Resolution: Invalid Priority: 5 Submitted By: Alvaro Mota Goncalves (alvaromota) Assigned to: Nobody/Anonymous (nobody) Summary: ClassCastException Initial Comment: SO - Linux RedHat JVM 1.3.1-06 Sun , 1.3.1 - Blackdown , 1.4.1 Sun JBOSS Server 3.0.4 and 3.0.3 A session Bean packaged in a jar file (Session01.jar) is calling one entity bean packaged in another jar file (Entity01.jar) . Works great. I have one client in Tomcat in another machine. If I hot-deploy Session01.jar, it's work. But hot-deploy Entity01 get 20:34:59,449 ERROR [LogInterceptor] RuntimeException: java.lang.ClassCastException: $Proxy156 at com.da.motion.testeAlvaro.model.statelessEJB.SManterTesteEJB.find All(SManterTes teEJB.java:75) The solution is to hot-deploy Entity01 and Session01 in sequence works fine. Thanks Alvaro -- Comment By: Christian Riege (lqd) Date: 2002-11-26 09:05 Message: Logged In: YES user_id=176671 alvaro, simply put: you can't. scott explained this in great detail below: The type of a class is its fully qualified name + the class loader that loaded it.. Once you lookup the EntityBean from your SessionBean, the class is defined by the Unified ClassLoader (this is what UCL stands for btw) that loaded your Entity01.jar. On redeploy, this UCL will be gone and a new one is created. So the reference that the Session Bean has points to the old UCL which is now invalid. Redeploying Session01.jar helps b/c once you call the Finder, the class will be loaded again (Session01.jar doesn't know about it yet). Package both files in an .ear and redeploy that. Everything else WILL NOT WORK. -- Comment By: Alvaro Mota Goncalves (alvaromota) Date: 2002-11-26 00:17 Message: Logged In: YES user_id=354157 Mr Scott When the session doesn't reference the local entity, the entity redeploy works. But, when the session makes at least one lookup to Entity, the problem occurs. Since the lookup Session--Entity doesn't keep state, Why the link between the Session and the entity keeps the state? Regards Alvaro -- Comment By: Bill Wadley (wrwadley) Date: 2002-11-25 22:33 Message: Logged In: YES user_id=223475 My problem: ClassCastExceptions on results from finders: $Proxy123 (numbers different per invocation) - EJB's in .jar, Servlets in .war, all in .ear. - Hot-deploy or server restart, it didn't matter. - changed JVM's for server (1.3.1, 1.3.1_01, 1.4.1, 1.4.1_01); didn't matter My solution: Stop compiling with Jikes 1.15 and use the same java/javac that I'm running the server with (Sun JDK 1.4.1). Now all is well. -- Comment By: Alvaro Mota Goncalves (alvaromota) Date: 2002-11-25 21:25 Message: Logged In: YES user_id=354157 Ok I need to redeploy only the ejb-jar, without needing to redeploy the whole application (ear). How can i do this? Thanks Alvaro -- Comment By: Scott M Stark (starksm) Date: 2002-11-25 21:20 Message: Logged In: YES user_id=175228 Its not possible which is why these ejbs need to be deployed together in an ear so that redeployment cycles the classes consistently. -- Comment By: Alvaro Mota Goncalves (alvaromota) Date: 2002-11-25 21:17 Message: Logged In: YES user_id=354157 Hi Mr. Scott --Deployments with explicit references to each other must be deployed as a unit so that the Java type system remains consistent. How is possible to make an unexplicit reference to Entity Local in another ejb-jar.file(Entity01.jar).? Session -- Entity The source code is: private ECepTituloHome getHome() { ECepTituloHome home = null; try { //Implementar o getHome Local para o JBOSS InitialContext jndi = new InitialContext(); // delete all organizations home = (ECepTituloHome) jndi.lookup(ejb/entityEJB/ECepTituloBeanRef); } catch (Exception e) { System.err.println(e.getMessage()); } finally { return home; } } The client lookup source code is: private SManterTesteHome getHome() { SManterTesteHome home = null; try { Hashtable env = new Hashtable();
Re: [JBoss-dev] Re: JBoss/CMP Improvements
tisdagen den 26 november 2002 kl 03.08 skrev Stefan Arentz: I'd kill to have more test cases and especially performance test cases. I would love to have tests the spit out a XML file with timings. Then we could quickly determine the results of a change. And then create reports similar to the JUnit html output? Say a 'JBench' framework ... ? Together with some kind of versioning one could then deploy/bench cascades to a running jboss instance with different versions of the code to test/bench. Not fully sure how to do that tho ... ? ... /peter_f --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
[JBoss-dev] [ jboss-Bugs-643673 ] ClassCastException
Bugs item #643673, was opened at 2002-11-25 15:40 You can respond by visiting: https://sourceforge.net/tracker/?func=detailatid=376685aid=643673group_id=22866 Category: JBossServer Group: None Status: Closed Resolution: Invalid Priority: 5 Submitted By: Alvaro Mota Goncalves (alvaromota) Assigned to: Nobody/Anonymous (nobody) Summary: ClassCastException Initial Comment: SO - Linux RedHat JVM 1.3.1-06 Sun , 1.3.1 - Blackdown , 1.4.1 Sun JBOSS Server 3.0.4 and 3.0.3 A session Bean packaged in a jar file (Session01.jar) is calling one entity bean packaged in another jar file (Entity01.jar) . Works great. I have one client in Tomcat in another machine. If I hot-deploy Session01.jar, it's work. But hot-deploy Entity01 get 20:34:59,449 ERROR [LogInterceptor] RuntimeException: java.lang.ClassCastException: $Proxy156 at com.da.motion.testeAlvaro.model.statelessEJB.SManterTesteEJB.find All(SManterTes teEJB.java:75) The solution is to hot-deploy Entity01 and Session01 in sequence works fine. Thanks Alvaro -- Comment By: Alvaro Mota Goncalves (alvaromota) Date: 2002-11-26 10:00 Message: Logged In: YES user_id=354157 Ok, so I can't redeploy the two jar files separately. However, redeploying only my the ejb-jar with the session bean works fine, since it is not referenced by the entity bean. Now, my question is: can I pack the two jars into an ear, and redeploy only the session jar from inside the ear ? That is, when I need to redeploy the session bean, I can do it without redeploying also the entity; and when I need to redeploy the entity, I redeploy the whole ear. Is it possible ? Please note that in my real application the jar with entity beans is much bigger than in the example I mentioned. Thanks, Alvaro -- Comment By: Christian Riege (lqd) Date: 2002-11-26 06:05 Message: Logged In: YES user_id=176671 alvaro, simply put: you can't. scott explained this in great detail below: The type of a class is its fully qualified name + the class loader that loaded it.. Once you lookup the EntityBean from your SessionBean, the class is defined by the Unified ClassLoader (this is what UCL stands for btw) that loaded your Entity01.jar. On redeploy, this UCL will be gone and a new one is created. So the reference that the Session Bean has points to the old UCL which is now invalid. Redeploying Session01.jar helps b/c once you call the Finder, the class will be loaded again (Session01.jar doesn't know about it yet). Package both files in an .ear and redeploy that. Everything else WILL NOT WORK. -- Comment By: Alvaro Mota Goncalves (alvaromota) Date: 2002-11-25 21:17 Message: Logged In: YES user_id=354157 Mr Scott When the session doesn't reference the local entity, the entity redeploy works. But, when the session makes at least one lookup to Entity, the problem occurs. Since the lookup Session--Entity doesn't keep state, Why the link between the Session and the entity keeps the state? Regards Alvaro -- Comment By: Bill Wadley (wrwadley) Date: 2002-11-25 19:33 Message: Logged In: YES user_id=223475 My problem: ClassCastExceptions on results from finders: $Proxy123 (numbers different per invocation) - EJB's in .jar, Servlets in .war, all in .ear. - Hot-deploy or server restart, it didn't matter. - changed JVM's for server (1.3.1, 1.3.1_01, 1.4.1, 1.4.1_01); didn't matter My solution: Stop compiling with Jikes 1.15 and use the same java/javac that I'm running the server with (Sun JDK 1.4.1). Now all is well. -- Comment By: Alvaro Mota Goncalves (alvaromota) Date: 2002-11-25 18:25 Message: Logged In: YES user_id=354157 Ok I need to redeploy only the ejb-jar, without needing to redeploy the whole application (ear). How can i do this? Thanks Alvaro -- Comment By: Scott M Stark (starksm) Date: 2002-11-25 18:20 Message: Logged In: YES user_id=175228 Its not possible which is why these ejbs need to be deployed together in an ear so that redeployment cycles the classes consistently. -- Comment By: Alvaro Mota Goncalves (alvaromota) Date: 2002-11-25 18:17 Message: Logged In: YES user_id=354157 Hi Mr. Scott --Deployments with explicit references to each other must be deployed as a unit so that the Java type system remains consistent. How is possible to make an unexplicit reference to Entity Local in another ejb-jar.file(Entity01.jar).? Session -- Entity The source code is: private ECepTituloHome getHome() { ECepTituloHome home = null; try {
[JBoss-dev] [ jboss-Bugs-643673 ] ClassCastException
Bugs item #643673, was opened at 2002-11-25 18:40 You can respond by visiting: https://sourceforge.net/tracker/?func=detailatid=376685aid=643673group_id=22866 Category: JBossServer Group: None Status: Closed Resolution: Invalid Priority: 5 Submitted By: Alvaro Mota Goncalves (alvaromota) Assigned to: Nobody/Anonymous (nobody) Summary: ClassCastException Initial Comment: SO - Linux RedHat JVM 1.3.1-06 Sun , 1.3.1 - Blackdown , 1.4.1 Sun JBOSS Server 3.0.4 and 3.0.3 A session Bean packaged in a jar file (Session01.jar) is calling one entity bean packaged in another jar file (Entity01.jar) . Works great. I have one client in Tomcat in another machine. If I hot-deploy Session01.jar, it's work. But hot-deploy Entity01 get 20:34:59,449 ERROR [LogInterceptor] RuntimeException: java.lang.ClassCastException: $Proxy156 at com.da.motion.testeAlvaro.model.statelessEJB.SManterTesteEJB.find All(SManterTes teEJB.java:75) The solution is to hot-deploy Entity01 and Session01 in sequence works fine. Thanks Alvaro -- Comment By: Christian Riege (lqd) Date: 2002-11-26 13:43 Message: Logged In: YES user_id=176671 Redeploying only the Session Bean .jar works b/c the Entity Bean does not have any references to the Session Bean. You cannot redeploy something inside an .ear. It's all or nothing. Simply put: you can re-deploy the .jar containing the session beans as often as you want w/o having to redeploy the jar containing the entity beans. however, when you redeploy your entity-bean .jar, you HAVE TO redeploy your session-beans, too. this is very easily achieved by using appropriate targets if you're using ant for example. consult your tool vendor for further help. -- Comment By: Alvaro Mota Goncalves (alvaromota) Date: 2002-11-26 13:00 Message: Logged In: YES user_id=354157 Ok, so I can't redeploy the two jar files separately. However, redeploying only my the ejb-jar with the session bean works fine, since it is not referenced by the entity bean. Now, my question is: can I pack the two jars into an ear, and redeploy only the session jar from inside the ear ? That is, when I need to redeploy the session bean, I can do it without redeploying also the entity; and when I need to redeploy the entity, I redeploy the whole ear. Is it possible ? Please note that in my real application the jar with entity beans is much bigger than in the example I mentioned. Thanks, Alvaro -- Comment By: Christian Riege (lqd) Date: 2002-11-26 09:05 Message: Logged In: YES user_id=176671 alvaro, simply put: you can't. scott explained this in great detail below: The type of a class is its fully qualified name + the class loader that loaded it.. Once you lookup the EntityBean from your SessionBean, the class is defined by the Unified ClassLoader (this is what UCL stands for btw) that loaded your Entity01.jar. On redeploy, this UCL will be gone and a new one is created. So the reference that the Session Bean has points to the old UCL which is now invalid. Redeploying Session01.jar helps b/c once you call the Finder, the class will be loaded again (Session01.jar doesn't know about it yet). Package both files in an .ear and redeploy that. Everything else WILL NOT WORK. -- Comment By: Alvaro Mota Goncalves (alvaromota) Date: 2002-11-26 00:17 Message: Logged In: YES user_id=354157 Mr Scott When the session doesn't reference the local entity, the entity redeploy works. But, when the session makes at least one lookup to Entity, the problem occurs. Since the lookup Session--Entity doesn't keep state, Why the link between the Session and the entity keeps the state? Regards Alvaro -- Comment By: Bill Wadley (wrwadley) Date: 2002-11-25 22:33 Message: Logged In: YES user_id=223475 My problem: ClassCastExceptions on results from finders: $Proxy123 (numbers different per invocation) - EJB's in .jar, Servlets in .war, all in .ear. - Hot-deploy or server restart, it didn't matter. - changed JVM's for server (1.3.1, 1.3.1_01, 1.4.1, 1.4.1_01); didn't matter My solution: Stop compiling with Jikes 1.15 and use the same java/javac that I'm running the server with (Sun JDK 1.4.1). Now all is well. -- Comment By: Alvaro Mota Goncalves (alvaromota) Date: 2002-11-25 21:25 Message: Logged In: YES user_id=354157 Ok I need to redeploy only the ejb-jar, without needing to redeploy the whole application (ear). How can i do this? Thanks Alvaro -- Comment By: Scott M Stark (starksm) Date: 2002-11-25 21:20 Message: Logged In: YES user_id=175228
Re: [JBoss-dev] Re: JBoss/CMP Improvements
On Tuesday, Nov 26, 2002, at 03:40 Europe/Amsterdam, Dain Sundstrom wrote: On Monday, November 25, 2002, at 08:08 PM, Stefan Arentz wrote: I'd kill to have more test cases and especially performance test cases. I would love to have tests the spit out a XML file with timings. Then we could quickly determine the results of a change. And then create reports similar to the JUnit html output? Say a 'JBench' framework ... ? Never heard of it... have you used it It doesn't exist, except in my mind. I can start working on it. I've been looking for something ilke that to profile my own J2EE apps ... S. --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
[JBoss-dev] [ jboss-Bugs-644146 ] Problem with Deploy on demand
Bugs item #644146, was opened at 2002-11-26 15:15 You can respond by visiting: https://sourceforge.net/tracker/?func=detailatid=376685aid=644146group_id=22866 Category: None Group: None Status: Open Resolution: None Priority: 5 Submitted By: nicola degara (labradornic) Assigned to: Nobody/Anonymous (nobody) Summary: Problem with Deploy on demand Initial Comment: RED HAT 8.0 JDK 1.4 Hi! I've got a J2EE component (A) wich is called from another one (B) in this way: javax.naming.InitialContext jndiContext = new javax.naming.InitialContext(); Object ref = jndiContext.lookup(CISIContext); anthesi.isiportal.context.IISIContextHome home = (anthesi.isiportal.context.IISIContextHome)javax.rmi.PortableRemoteObject.narrow(ref,anthesi.isiportal.context.IISIContextHome.class); anthesi.isiportal.context.IISIContext context = home.create(ApplServer,ServerName); String strXml=context.getXML(param1, param2,...); -- If I re-deploy my J2EE component (A) the caller(B) will not work even if nothing has been changed (in A or B); the only way to correct this situation is to re-deploy my (B) component again! If the components are only 2 (caller ansd called) no problems!! But I've got N components and all one call my component (A) !!! So everytime I've got to re-deploy all my callers! -- or better restart JBOSS!!! What shall I do? Thank you Nicola Degara [EMAIL PROTECTED] -- You can respond by visiting: https://sourceforge.net/tracker/?func=detailatid=376685aid=644146group_id=22866 --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
Re: [JBoss-dev] Re: JBoss/CMP Improvements
tisdagen den 26 november 2002 kl 15.49 skrev Stefan Arentz: I'd kill to have more test cases and especially performance test cases. I would love to have tests the spit out a XML file with timings. Then we could quickly determine the results of a change. And then create reports similar to the JUnit html output? Say a 'JBench' framework ... ? Never heard of it... have you used it It doesn't exist, except in my mind. I can start working on it. I've been looking for something ilke that to profile my own J2EE apps ... http://www.yoda.arachsys.com/java/jbench/docs/index.html --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
[JBoss-dev] [ jboss-Bugs-644146 ] Problem with Deploy on demand
Bugs item #644146, was opened at 2002-11-26 16:15 You can respond by visiting: https://sourceforge.net/tracker/?func=detailatid=376685aid=644146group_id=22866 Category: None Group: None Status: Closed Resolution: Duplicate Priority: 5 Submitted By: nicola degara (labradornic) Assigned to: Nobody/Anonymous (nobody) Summary: Problem with Deploy on demand Initial Comment: RED HAT 8.0 JDK 1.4 Hi! I've got a J2EE component (A) wich is called from another one (B) in this way: javax.naming.InitialContext jndiContext = new javax.naming.InitialContext(); Object ref = jndiContext.lookup(CISIContext); anthesi.isiportal.context.IISIContextHome home = (anthesi.isiportal.context.IISIContextHome)javax.rmi.PortableRemoteObject.narrow(ref,anthesi.isiportal.context.IISIContextHome.class); anthesi.isiportal.context.IISIContext context = home.create(ApplServer,ServerName); String strXml=context.getXML(param1, param2,...); -- If I re-deploy my J2EE component (A) the caller(B) will not work even if nothing has been changed (in A or B); the only way to correct this situation is to re-deploy my (B) component again! If the components are only 2 (caller ansd called) no problems!! But I've got N components and all one call my component (A) !!! So everytime I've got to re-deploy all my callers! -- or better restart JBOSS!!! What shall I do? Thank you Nicola Degara [EMAIL PROTECTED] -- Comment By: Christian Riege (lqd) Date: 2002-11-26 16:57 Message: Logged In: YES user_id=176671 This looks like a duplicate of bug #643673, see https://sourceforge.net/tracker/?func=detailatid=376685aid=643673group_id=22866 -- You can respond by visiting: https://sourceforge.net/tracker/?func=detailatid=376685aid=644146group_id=22866 --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
[JBoss-dev] [ jboss-Bugs-641740 ] jboss-3.0.4 classloader bug
Bugs item #641740, was opened at 2002-11-21 01:37 You can respond by visiting: https://sourceforge.net/tracker/?func=detailatid=376685aid=641740group_id=22866 Category: JBossServer Group: v3.0 Rabbit Hole Status: Open Resolution: Later Priority: 5 Submitted By: Vladimir Korenev (vkorenev) Assigned to: Scott M Stark (starksm) Summary: jboss-3.0.4 classloader bug Initial Comment: The same bug was submitted against v.3.0.2 (see #606359) and it appeared again in v.3.0.4 -- Comment By: Scott M Stark (starksm) Date: 2002-11-26 08:14 Message: Logged In: YES user_id=175228 Can you run the ListJar program on the original ear and provide that so I can be sure of the structure of the ear? -- Comment By: Vladimir Korenev (vkorenev) Date: 2002-11-25 23:25 Message: Logged In: YES user_id=622967 CbUser and CurrentUser classes are not included in client-core-ejb.jar and admin-ejb.jar. They are in dao-common.jar and are referenced from manifest Class-Path attribute. As a workaround I deleted dao-common.jar from ear and placed it in server/default/lib directory. The application began working, but I don't think it is a good solution. -- Comment By: Scott M Stark (starksm) Date: 2002-11-25 11:58 Message: Logged In: YES user_id=175228 Looking more at the server.log I see: 2002-11-21 12:20:28,271 WARN [org.jboss.deployment.MainDeployer] The manifest e ntry in file:/D:/java/jboss- 3.0.4/server/default/tmp/deploy/server/default/deplo y/cboss.ear/66.cboss.ear-contents/client-core-ejb.jar references URL file:/D:/ja va/jboss- 3.0.4/server/default/tmp/deploy/server/default/deploy/cboss.ea r/66.cbos s.ear-contents/client-common.jar which could not be opened, entry ignored Run the attached ListJar class on the cboss.ear so I can see the complete structure and manifest references. You should be able to attach the output to this bug unless your ear is huge. -- Comment By: Scott M Stark (starksm) Date: 2002-11-25 11:39 Message: Logged In: YES user_id=175228 The attached ia.log is a distillation of the server.log and ucl.log sent to me by Vladimir. It shows the CurrentUser class being loaded from the client-core-ejb.jar and the CbUser being loaded from the admin-ejb.jar. Presumably the classes are redundantly included in both of the jars. In order for this to work with the current class loading scheme the common classes that use package protected or protected access must be loaded by the same class loader. This means these utility classes must be factored out into a seperate jar that is referenced by the ejb-jars manifest Class- Path attribute. In 3.2 we will be looking at either transforming the deployment into a canonical set of classes or using a single UnifiedClassLoader for all of the jars in the top level deployment to avoid this. -- Comment By: Vladimir Korenev (vkorenev) Date: 2002-11-21 01:43 Message: Logged In: YES user_id=622967 I have got an IllegalAccessError when accessing a field with package access. IMHO it is a classloader bug. I have attached an archive with logs an 2 source files. P.S. The source files are ugly (they are not mine), but they worked fine on Orion. -- You can respond by visiting: https://sourceforge.net/tracker/?func=detailatid=376685aid=641740group_id=22866 --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
[JBoss-dev] [ jboss-Bugs-641740 ] jboss-3.0.4 classloader bug
Bugs item #641740, was opened at 2002-11-21 12:37 You can respond by visiting: https://sourceforge.net/tracker/?func=detailatid=376685aid=641740group_id=22866 Category: JBossServer Group: v3.0 Rabbit Hole Status: Open Resolution: Later Priority: 5 Submitted By: Vladimir Korenev (vkorenev) Assigned to: Scott M Stark (starksm) Summary: jboss-3.0.4 classloader bug Initial Comment: The same bug was submitted against v.3.0.2 (see #606359) and it appeared again in v.3.0.4 -- Comment By: Vladimir Korenev (vkorenev) Date: 2002-11-26 19:50 Message: Logged In: YES user_id=622967 Yes, I have already done it. It is in ListJar.zip. -- Comment By: Scott M Stark (starksm) Date: 2002-11-26 19:14 Message: Logged In: YES user_id=175228 Can you run the ListJar program on the original ear and provide that so I can be sure of the structure of the ear? -- Comment By: Vladimir Korenev (vkorenev) Date: 2002-11-26 10:25 Message: Logged In: YES user_id=622967 CbUser and CurrentUser classes are not included in client-core-ejb.jar and admin-ejb.jar. They are in dao-common.jar and are referenced from manifest Class-Path attribute. As a workaround I deleted dao-common.jar from ear and placed it in server/default/lib directory. The application began working, but I don't think it is a good solution. -- Comment By: Scott M Stark (starksm) Date: 2002-11-25 22:58 Message: Logged In: YES user_id=175228 Looking more at the server.log I see: 2002-11-21 12:20:28,271 WARN [org.jboss.deployment.MainDeployer] The manifest e ntry in file:/D:/java/jboss- 3.0.4/server/default/tmp/deploy/server/default/deplo y/cboss.ear/66.cboss.ear-contents/client-core-ejb.jar references URL file:/D:/ja va/jboss- 3.0.4/server/default/tmp/deploy/server/default/deploy/cboss.ea r/66.cbos s.ear-contents/client-common.jar which could not be opened, entry ignored Run the attached ListJar class on the cboss.ear so I can see the complete structure and manifest references. You should be able to attach the output to this bug unless your ear is huge. -- Comment By: Scott M Stark (starksm) Date: 2002-11-25 22:39 Message: Logged In: YES user_id=175228 The attached ia.log is a distillation of the server.log and ucl.log sent to me by Vladimir. It shows the CurrentUser class being loaded from the client-core-ejb.jar and the CbUser being loaded from the admin-ejb.jar. Presumably the classes are redundantly included in both of the jars. In order for this to work with the current class loading scheme the common classes that use package protected or protected access must be loaded by the same class loader. This means these utility classes must be factored out into a seperate jar that is referenced by the ejb-jars manifest Class- Path attribute. In 3.2 we will be looking at either transforming the deployment into a canonical set of classes or using a single UnifiedClassLoader for all of the jars in the top level deployment to avoid this. -- Comment By: Vladimir Korenev (vkorenev) Date: 2002-11-21 12:43 Message: Logged In: YES user_id=622967 I have got an IllegalAccessError when accessing a field with package access. IMHO it is a classloader bug. I have attached an archive with logs an 2 source files. P.S. The source files are ugly (they are not mine), but they worked fine on Orion. -- You can respond by visiting: https://sourceforge.net/tracker/?func=detailatid=376685aid=641740group_id=22866 --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
[JBoss-dev] CLEAN BUILD FAILS
folks, please please please. A clean build of JBoss fails with buildfragments/tools.eng:29: taskdef class xdoclet.modules.jmx.JMXDocletTask cannot be found. What gives? Also it seems the nightly build from chris is gone? marcf xx Marc Fleury, Ph.D. President, Founder JBoss Group, LLC xx --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
Re: [JBoss-dev] CLEAN BUILD FAILS
Marc, pay attention to the dev list, and always explain what you are doing when you complain. cvs co jboss-all hasn't worked for weeks Try cvs co jboss-head cvs co -r Branch_3_2 jboss-3.2 or cvs co -r Branch_3_0 jboss-3.0 or cvs co -r Branch_3_0 jboss-all There were probably 50 to 100 messages on this when it changed. david On 2002.11.26 12:47:04 -0500 marc fleury wrote: folks, please please please. A clean build of JBoss fails with buildfragments/tools.eng:29: taskdef class xdoclet.modules.jmx.JMXDocletTask cannot be found. What gives? Also it seems the nightly build from chris is gone? marcf xx Marc Fleury, Ph.D. President, Founder JBoss Group, LLC xx --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
RE: [JBoss-dev] CLEAN BUILD FAILS
OK david jencks pointed out that I was checking out jboss-all instead of jboss-head never mind It remains that chris kimpton stuff seems gone? -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]] On Behalf Of marc fleury Sent: Tuesday, November 26, 2002 12:47 PM To: Jboss-Development@Lists. Sourceforge. Net Subject: [JBoss-dev] CLEAN BUILD FAILS folks, please please please. A clean build of JBoss fails with buildfragments/tools.eng:29: taskdef class xdoclet.modules.jmx.JMXDocletTask cannot be found. What gives? Also it seems the nightly build from chris is gone? marcf xx Marc Fleury, Ph.D. President, Founder JBoss Group, LLC xx --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
RE: [JBoss-dev] CLEAN BUILD FAILS
-Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]] On Behalf Of David Jencks Sent: Tuesday, November 26, 2002 1:13 PM To: [EMAIL PROTECTED] Subject: Re: [JBoss-dev] CLEAN BUILD FAILS Marc, pay attention to the dev list, and always explain what you are doing when you complain. cvs co jboss-all hasn't worked for weeks Try cvs co jboss-head cvs co -r Branch_3_2 jboss-3.2 or cvs co -r Branch_3_0 jboss-3.0 or cvs co -r Branch_3_0 jboss-all There were probably 50 to 100 messages on this when it changed. Right, I am doing sales and marketing these days almost full time. I am good at it but it takes all the time away from real coding as you can see. Anyway, I really want to implement the website changes so we can have blog/wiki to edit the webpages easily and get going with this website as a real repository of knowledge. I didn't follow this announcement and was lost in the cracks. So are probably 2458 other developers out there. marcf --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
[JBoss-dev] [ jboss-Bugs-644287 ] in a Filter, getServletPath() empty
Bugs item #644287, was opened at 2002-11-26 19:05 You can respond by visiting: https://sourceforge.net/tracker/?func=detailatid=376685aid=644287group_id=22866 Category: JBossWeb Group: v3.0 Rabbit Hole Status: Open Resolution: None Priority: 5 Submitted By: James Manning (jmm) Assigned to: Nobody/Anonymous (nobody) Summary: in a Filter, getServletPath() empty Initial Comment: using the servlet 2.3 API, in a defined filter (extends HttpServlet implements Filter), inside method public void doFilter(ServletRequest request, ServletResponse response, FilterChain filterChain) HttpServletRequest httpServletRequest = (HttpServletRequest)request; String current_file = httpServletRequest.getServletPath(); This works great under Tomcat 4.x but returns an empty string (not null, just empty) in jboss 3.0.4 Trying to put together the minimal webapp to show the bug now. Wanted to go ahead and get the bug posted just in case it's known or something -- You can respond by visiting: https://sourceforge.net/tracker/?func=detailatid=376685aid=644287group_id=22866 --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
[JBoss-dev] [ jboss-Bugs-644289 ] ClassLoading issue with package protecte
Bugs item #644289, was opened at 2002-11-26 11:10 You can respond by visiting: https://sourceforge.net/tracker/?func=detailatid=376685aid=644289group_id=22866 Category: JBossMX Group: v3.0 Rabbit Hole Status: Open Resolution: None Priority: 5 Submitted By: Scott M Stark (starksm) Assigned to: Scott M Stark (starksm) Summary: ClassLoading issue with package protecte Initial Comment: When package protected access is used with classes that are loaded by more than one classloader via manifest Class-Path references you can get IllegalAccessErrors when more than one jar references the jar containing the classes due to the use of a UCL per jar. This has been reported in various forms in bugs 565701, 606359, 641740 and is consolidated here with a trival testcase that demonstrates the problem. The org.jboss.test.classloader.test.CircularityUnitTestCase testPackageProtected unit test demonstrates the problem with a repository that contains a login.jar and a usrmgr.jar, both of which reference a cl-util.jar. The contents of the jars is: login.jar +- org/jboss/test/classloader/circularity/support/UserOfLogi nInfo.class +- META-INF/MANIFEST.MF Class-Path: cl-util.jar usrmgr.jar +- org/jboss/test/classloader/circularity/support/UserOfUsr Mgr.class +- META-INF/MANIFEST.MF Class-Path: cl-util.jar cl-util.jar +- org/jboss/test/classloader/circularity/support/LoginInfo.cl ass +- org/jboss/test/classloader/circularity/support/UsrMgr.cla ss +- META-INF/MANIFEST.MF The test first loads the UserOfLoginInfo class using the UCL associated with login.jar, and then loads the UserOfUsrMgr class using the UCL associated with usrmgr.jar. A changePassword method is then invoked on an instance of UserOfUsrMgr and this results in an IllegalAccessError because the LoginInfo class was seen to be loaded from the UCL associated with login.jar while the UsMgr class was loaded from the UCL associated with usrmgr.jar. Although both are in the same package and jar, two different class loaders are involved and so the access check fails. The only workaround at the moment requires moving the common jar to the server/xxx/lib directory so that the UCLs do not see common jar as part of their classes. -- You can respond by visiting: https://sourceforge.net/tracker/?func=detailatid=376685aid=644289group_id=22866 --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
[JBoss-dev] [ jboss-Bugs-644289 ] ClassLoader issue with protected access and manifest refs
Bugs item #644289, was opened at 2002-11-26 11:10 You can respond by visiting: https://sourceforge.net/tracker/?func=detailatid=376685aid=644289group_id=22866 Category: JBossMX Group: v3.0 Rabbit Hole Status: Open Resolution: None Priority: 5 Submitted By: Scott M Stark (starksm) Assigned to: Scott M Stark (starksm) Summary: ClassLoader issue with protected access and manifest refs Initial Comment: When package protected access is used with classes that are loaded by more than one classloader via manifest Class-Path references you can get IllegalAccessErrors when more than one jar references the jar containing the classes due to the use of a UCL per jar. This has been reported in various forms in bugs 565701, 606359, 641740 and is consolidated here with a trival testcase that demonstrates the problem. The org.jboss.test.classloader.test.CircularityUnitTestCase testPackageProtected unit test demonstrates the problem with a repository that contains a login.jar and a usrmgr.jar, both of which reference a cl-util.jar. The contents of the jars is: login.jar +- org/jboss/test/classloader/circularity/support/UserOfLogi nInfo.class +- META-INF/MANIFEST.MF Class-Path: cl-util.jar usrmgr.jar +- org/jboss/test/classloader/circularity/support/UserOfUsr Mgr.class +- META-INF/MANIFEST.MF Class-Path: cl-util.jar cl-util.jar +- org/jboss/test/classloader/circularity/support/LoginInfo.cl ass +- org/jboss/test/classloader/circularity/support/UsrMgr.cla ss +- META-INF/MANIFEST.MF The test first loads the UserOfLoginInfo class using the UCL associated with login.jar, and then loads the UserOfUsrMgr class using the UCL associated with usrmgr.jar. A changePassword method is then invoked on an instance of UserOfUsrMgr and this results in an IllegalAccessError because the LoginInfo class was seen to be loaded from the UCL associated with login.jar while the UsMgr class was loaded from the UCL associated with usrmgr.jar. Although both are in the same package and jar, two different class loaders are involved and so the access check fails. The only workaround at the moment requires moving the common jar to the server/xxx/lib directory so that the UCLs do not see common jar as part of their classes. -- You can respond by visiting: https://sourceforge.net/tracker/?func=detailatid=376685aid=644289group_id=22866 --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
[JBoss-dev] [ jboss-Bugs-641740 ] jboss-3.0.4 classloader bug
Bugs item #641740, was opened at 2002-11-21 01:37 You can respond by visiting: https://sourceforge.net/tracker/?func=detailatid=376685aid=641740group_id=22866 Category: JBossServer Group: v3.0 Rabbit Hole Status: Closed Resolution: Duplicate Priority: 5 Submitted By: Vladimir Korenev (vkorenev) Assigned to: Scott M Stark (starksm) Summary: jboss-3.0.4 classloader bug Initial Comment: The same bug was submitted against v.3.0.2 (see #606359) and it appeared again in v.3.0.4 -- Comment By: Scott M Stark (starksm) Date: 2002-11-26 11:14 Message: Logged In: YES user_id=175228 Ok, sorry I did not see that. I am closing this as a duplicate of 644289 bug I just opended as I have a trival testcase that demonstrates the problem so follow that bug for a resolution. -- Comment By: Vladimir Korenev (vkorenev) Date: 2002-11-26 08:50 Message: Logged In: YES user_id=622967 Yes, I have already done it. It is in ListJar.zip. -- Comment By: Scott M Stark (starksm) Date: 2002-11-26 08:14 Message: Logged In: YES user_id=175228 Can you run the ListJar program on the original ear and provide that so I can be sure of the structure of the ear? -- Comment By: Vladimir Korenev (vkorenev) Date: 2002-11-25 23:25 Message: Logged In: YES user_id=622967 CbUser and CurrentUser classes are not included in client-core-ejb.jar and admin-ejb.jar. They are in dao-common.jar and are referenced from manifest Class-Path attribute. As a workaround I deleted dao-common.jar from ear and placed it in server/default/lib directory. The application began working, but I don't think it is a good solution. -- Comment By: Scott M Stark (starksm) Date: 2002-11-25 11:58 Message: Logged In: YES user_id=175228 Looking more at the server.log I see: 2002-11-21 12:20:28,271 WARN [org.jboss.deployment.MainDeployer] The manifest e ntry in file:/D:/java/jboss- 3.0.4/server/default/tmp/deploy/server/default/deplo y/cboss.ear/66.cboss.ear-contents/client-core-ejb.jar references URL file:/D:/ja va/jboss- 3.0.4/server/default/tmp/deploy/server/default/deploy/cboss.ea r/66.cbos s.ear-contents/client-common.jar which could not be opened, entry ignored Run the attached ListJar class on the cboss.ear so I can see the complete structure and manifest references. You should be able to attach the output to this bug unless your ear is huge. -- Comment By: Scott M Stark (starksm) Date: 2002-11-25 11:39 Message: Logged In: YES user_id=175228 The attached ia.log is a distillation of the server.log and ucl.log sent to me by Vladimir. It shows the CurrentUser class being loaded from the client-core-ejb.jar and the CbUser being loaded from the admin-ejb.jar. Presumably the classes are redundantly included in both of the jars. In order for this to work with the current class loading scheme the common classes that use package protected or protected access must be loaded by the same class loader. This means these utility classes must be factored out into a seperate jar that is referenced by the ejb-jars manifest Class- Path attribute. In 3.2 we will be looking at either transforming the deployment into a canonical set of classes or using a single UnifiedClassLoader for all of the jars in the top level deployment to avoid this. -- Comment By: Vladimir Korenev (vkorenev) Date: 2002-11-21 01:43 Message: Logged In: YES user_id=622967 I have got an IllegalAccessError when accessing a field with package access. IMHO it is a classloader bug. I have attached an archive with logs an 2 source files. P.S. The source files are ugly (they are not mine), but they worked fine on Orion. -- You can respond by visiting: https://sourceforge.net/tracker/?func=detailatid=376685aid=641740group_id=22866 --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
[JBoss-dev] ClassLoader issue with protected access and manifest refs
Ok, I finally got a handle on the IllegalAccessError that has been reported for a while now and have a trivial testcase that demonstrates the problem. See the bug details here: [ 644289 ] ClassLoader issue with protected access and manifest refs http://sourceforge.net/tracker/index.php?func=detailaid=644289group_id=22866atid=376685 The real problem is the use of a UCL per jar for deployments. Because the java URLClassLoader keeps track of the Class-Path reference in the jar it owns, when more than one jar references a common jar the UCLs associated with the referencing jars are candidates for the defining class loader for the classes in the common jar. Effectively this means that the common jar classes are deployed multiple times even though only a single instance of a given class is created due to the ULR caching. I think the simplest fix is to introduce a common class loader associated with the parent deployment that contains the union of all jars contained in the deployment. In addition to solving this problem it also helps with issues of duplicates classes in a deployment since with a single class loader an ordering of the jars can be specified. I'm working on this issue in main and will backport fixes to 3.2 and 3.0 Scott Stark Chief Technology Officer JBoss Group, LLC --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
Re: [JBoss-dev] CLEAN BUILD FAILS
Hi, The hourly compiles are there still - although the URL has changed: http://jboss.kimptoc.net/ ...although I am not sure why there was no email about the problem earlier... The nightly tests currently do not complete - hence the title on the build page and the outstanding bug. I have switched the tests to once a week - until things get better... Chris - Original Message - From: marc fleury [EMAIL PROTECTED] To: [EMAIL PROTECTED] Cc: 'Chris Kimpton' [EMAIL PROTECTED] Sent: Tuesday, November 26, 2002 6:18 PM Subject: RE: [JBoss-dev] CLEAN BUILD FAILS OK david jencks pointed out that I was checking out jboss-all instead of jboss-head never mind It remains that chris kimpton stuff seems gone? -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]] On Behalf Of marc fleury Sent: Tuesday, November 26, 2002 12:47 PM To: Jboss-Development@Lists. Sourceforge. Net Subject: [JBoss-dev] CLEAN BUILD FAILS folks, please please please. A clean build of JBoss fails with buildfragments/tools.eng:29: taskdef class xdoclet.modules.jmx.JMXDocletTask cannot be found. What gives? Also it seems the nightly build from chris is gone? marcf xx Marc Fleury, Ph.D. President, Founder JBoss Group, LLC xx --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
[JBoss-dev] [ jboss-Bugs-644287 ] in a Filter, getServletPath() empty
Bugs item #644287, was opened at 2002-11-26 19:05 You can respond by visiting: https://sourceforge.net/tracker/?func=detailatid=376685aid=644287group_id=22866 Category: JBossWeb Group: v3.0 Rabbit Hole Status: Open Resolution: None Priority: 5 Submitted By: James Manning (jmm) Assigned to: Nobody/Anonymous (nobody) Summary: in a Filter, getServletPath() empty Initial Comment: using the servlet 2.3 API, in a defined filter (extends HttpServlet implements Filter), inside method public void doFilter(ServletRequest request, ServletResponse response, FilterChain filterChain) HttpServletRequest httpServletRequest = (HttpServletRequest)request; String current_file = httpServletRequest.getServletPath(); This works great under Tomcat 4.x but returns an empty string (not null, just empty) in jboss 3.0.4 Trying to put together the minimal webapp to show the bug now. Wanted to go ahead and get the bug posted just in case it's known or something -- Comment By: James Manning (jmm) Date: 2002-11-26 20:02 Message: Logged In: YES user_id=11485 this is the minimal test-case AFAICT. logging shows: 15:02:14,709 INFO [TestFilter] *** getServletPath gives -- You can respond by visiting: https://sourceforge.net/tracker/?func=detailatid=376685aid=644287group_id=22866 --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
[JBoss-dev] [ jboss-Bugs-644287 ] in a Filter, getServletPath() empty
Bugs item #644287, was opened at 2002-11-26 19:05 You can respond by visiting: https://sourceforge.net/tracker/?func=detailatid=376685aid=644287group_id=22866 Category: JBossWeb Group: v3.0 Rabbit Hole Status: Open Resolution: None Priority: 5 Submitted By: James Manning (jmm) Assigned to: Nobody/Anonymous (nobody) Summary: in a Filter, getServletPath() empty Initial Comment: using the servlet 2.3 API, in a defined filter (extends HttpServlet implements Filter), inside method public void doFilter(ServletRequest request, ServletResponse response, FilterChain filterChain) HttpServletRequest httpServletRequest = (HttpServletRequest)request; String current_file = httpServletRequest.getServletPath(); This works great under Tomcat 4.x but returns an empty string (not null, just empty) in jboss 3.0.4 Trying to put together the minimal webapp to show the bug now. Wanted to go ahead and get the bug posted just in case it's known or something -- Comment By: James Manning (jmm) Date: 2002-11-26 20:13 Message: Logged In: YES user_id=11485 on Tomcat 4.x 0 [HttpProcessor[8000][3]] INFO TestFilter - *** getServletPath gives /foo/bar/ack So this appears (afaict) to be a jboss breakage -- Comment By: James Manning (jmm) Date: 2002-11-26 20:02 Message: Logged In: YES user_id=11485 this is the minimal test-case AFAICT. logging shows: 15:02:14,709 INFO [TestFilter] *** getServletPath gives -- You can respond by visiting: https://sourceforge.net/tracker/?func=detailatid=376685aid=644287group_id=22866 --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
Re: [JBoss-dev] CLEAN BUILD FAILS
I setup anthill to do the builds in a automated fashion at our company, it's a lot more easier to setup than CC (which I tried without success :) ). You just have to make some tricks to keep all the ant logs and junit results in different dirs. If want some help to configure it to run jboss, just let me know. Chris Kimpton wrote: Hi, The hourly compiles are there still - although the URL has changed: http://jboss.kimptoc.net/ ...although I am not sure why there was no email about the problem earlier... The nightly tests currently do not complete - hence the title on the build page and the outstanding bug. I have switched the tests to once a week - until things get better... Chris - Original Message - From: marc fleury [EMAIL PROTECTED] To: [EMAIL PROTECTED] Cc: 'Chris Kimpton' [EMAIL PROTECTED] Sent: Tuesday, November 26, 2002 6:18 PM Subject: RE: [JBoss-dev] CLEAN BUILD FAILS OK david jencks pointed out that I was checking out jboss-all instead of jboss-head never mind It remains that chris kimpton stuff seems gone? -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]] On Behalf Of marc fleury Sent: Tuesday, November 26, 2002 12:47 PM To: Jboss-Development@Lists. Sourceforge. Net Subject: [JBoss-dev] CLEAN BUILD FAILS folks, please please please. A clean build of JBoss fails with buildfragments/tools.eng:29: taskdef class xdoclet.modules.jmx.JMXDocletTask cannot be found. What gives? Also it seems the nightly build from chris is gone? marcf xx Marc Fleury, Ph.D. President, Founder JBoss Group, LLC xx --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development -- | Emerson Cargnin | | Analista de Sistemas Sr. | | Tel : (051) 3358-4959| | SICREDI Serviços | | Porto Alegre - Brasil| |xx| --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
Re: [JBoss-dev] ClassLoader issue with protected access and manifest refs
Could we just write our own URLClassloader that handles this stuff correctly? We have the URL and Zip classes available. -dain On Tuesday, November 26, 2002, at 01:29 PM, Scott M Stark wrote: Ok, I finally got a handle on the IllegalAccessError that has been reported for a while now and have a trivial testcase that demonstrates the problem. See the bug details here: [ 644289 ] ClassLoader issue with protected access and manifest refs http://sourceforge.net/tracker/ index.php?func=detailaid=644289group_id=22866atid=376685 The real problem is the use of a UCL per jar for deployments. Because the java URLClassLoader keeps track of the Class-Path reference in the jar it owns, when more than one jar references a common jar the UCLs associated with the referencing jars are candidates for the defining class loader for the classes in the common jar. Effectively this means that the common jar classes are deployed multiple times even though only a single instance of a given class is created due to the ULR caching. I think the simplest fix is to introduce a common class loader associated with the parent deployment that contains the union of all jars contained in the deployment. In addition to solving this problem it also helps with issues of duplicates classes in a deployment since with a single class loader an ordering of the jars can be specified. I'm working on this issue in main and will backport fixes to 3.2 and 3.0 Scott Stark Chief Technology Officer JBoss Group, LLC --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
[JBoss-dev] Automated JBoss(Branch_3_0) Testsuite Results: 26-November-2002
Number of tests run: 995 Successful tests: 987 Errors:7 Failures: 1 [time of test: 2002-11-26.11-56 GMT] [java.version: 1.3.1] [java.vendor: Apple Computer, Inc.] [java.vm.version: 1.3.1_03-69] [java.vm.name: Java HotSpot(TM) Client VM] [java.vm.info: mixed mode] [os.name: Mac OS X] [os.arch: ppc] [os.version: 10.2.2] See http://users.jboss.org/~starksm/Branch_3_0/2002-11-26.11-56 for details of this test. NOTE: If there are any errors shown above - this mail is only highlighting them - it is NOT indicating that they are being looked at by anyone. Remember - if a test becomes broken after your changes - fix it or fix the test! Oh dear - still got some errors! Thanks for all your effort - we really do love you! --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
[JBoss-dev] [ jboss-Bugs-644287 ] in a Filter, getServletPath() empty
Bugs item #644287, was opened at 2002-11-26 19:05 You can respond by visiting: https://sourceforge.net/tracker/?func=detailatid=376685aid=644287group_id=22866 Category: JBossWeb Group: v3.0 Rabbit Hole Status: Open Resolution: None Priority: 5 Submitted By: James Manning (jmm) Assigned to: Nobody/Anonymous (nobody) Summary: in a Filter, getServletPath() empty Initial Comment: using the servlet 2.3 API, in a defined filter (extends HttpServlet implements Filter), inside method public void doFilter(ServletRequest request, ServletResponse response, FilterChain filterChain) HttpServletRequest httpServletRequest = (HttpServletRequest)request; String current_file = httpServletRequest.getServletPath(); This works great under Tomcat 4.x but returns an empty string (not null, just empty) in jboss 3.0.4 Trying to put together the minimal webapp to show the bug now. Wanted to go ahead and get the bug posted just in case it's known or something -- Comment By: James Manning (jmm) Date: 2002-11-26 21:09 Message: Logged In: YES user_id=11485 in case it helps, adding another check for getRequestURI shows: (under Tomcat 4.x) 10095 [HttpProcessor[8000][3]] INFO TestFilter - *** getServletPath gives /foo/bar/ack 10095 [HttpProcessor[8000][3]] INFO TestFilter - *** getRequestURI gives /filter/foo/bar/ack under jboss 3.0.4 16:09:28,870 INFO [TestFilter] *** getServletPath gives 16:09:28,870 INFO [TestFilter] *** getRequestURI gives /filter/foo/bar/ack Both are being hit by IE with a url of http://127.0.0.1:8080/filter/foo/bar/ack (port number changed to point to each one) -- Comment By: James Manning (jmm) Date: 2002-11-26 20:13 Message: Logged In: YES user_id=11485 on Tomcat 4.x 0 [HttpProcessor[8000][3]] INFO TestFilter - *** getServletPath gives /foo/bar/ack So this appears (afaict) to be a jboss breakage -- Comment By: James Manning (jmm) Date: 2002-11-26 20:02 Message: Logged In: YES user_id=11485 this is the minimal test-case AFAICT. logging shows: 15:02:14,709 INFO [TestFilter] *** getServletPath gives -- You can respond by visiting: https://sourceforge.net/tracker/?func=detailatid=376685aid=644287group_id=22866 --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
[JBoss-dev] where to post correction to jboss 3.0.4 adm. and devel. Second Edition?
where should I post correction to jboss 3.0.4 adm. and devel. Second Edition? by this list? by bug at source forge? -- | Emerson Cargnin | | Analista de Sistemas Sr. | | Tel : (051) 3358-4959| | SICREDI Serviços | | Porto Alegre - Brasil| |xx| --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
[JBoss-dev] Re: [JBoss-user] Using the same JAAS user to open connections
maybe Steven Coy could give a tip on this?? (he created the oracle-service.xml) David Jencks wrote: I admit I have never set it up and tried it... In the oracle-service.xml specify a jaas security domain In the jaas security domain setup in login-conf.xml use the CallerIdentityLoginModule. I am not entirely sure how to set up the login module so it will get the existing user/pw from the SecurityAssociation. The possibilities I would try are to have a separate security domain for the caller identity login module and also to include that login module in your primary security domain. (and mention this one in your oracle-service.xml) how to metion this one (the CallerIdentityLoginModule i suppose) at the oracle-services.xml? Do I have to take the user/pass out of the atribute ManagedConnectionFactoryProperties ?? Please report on your success. thanks david jencks On 2002.11.25 15:17:48 -0500 Emerson Cargnin - SICREDI Serviços wrote: We use Jaas Ldap LoginModule to make the authentication/authorization. Can someone explain me (or tell me where can I find at forum/maillist/jboss 3.0.4 admbook) how to set the oracle-services.xml to use the same user that's logged at the session bean (a session facade)? -- | Emerson Cargnin | | Analista de Sistemas Sr. | | Tel : (051) 3358-4959| | SICREDI Serviços | | Porto Alegre - Brasil| |xx| --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ JBoss-user mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-user --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ JBoss-user mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-user -- | Emerson Cargnin | | Analista de Sistemas Sr. | | Tel : (051) 3358-4959| | SICREDI Serviços | | Porto Alegre - Brasil| |xx| --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ JBoss-user mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-user --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
[JBoss-dev] What is the status of hsqldb 1.7 in JBoss 3.2?
I'm trying to get PetStore 1.3.1 running with 3.2 but its DAO code uses ResultSet.absolute() which is not supported by hsqldb 1.6 but is by 1.7.1 Will the changes to support 1.7.1 be backported to 3.2 before its release? thanks jeremy --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
Re: [JBoss-dev] ClassLoader issue with protected access and manifest refs
Won't this eliminate the possiblity of redeploying a subpackage of say an unpacked .ear deployment? In fact, I have been able to redeploy subpackage of a packaged .ear using emacs ability to modify a .jar file's contents and the main deployers ability to redeploy packages explicitly. I take it this behavior occurs even though the jar referenced on the manifest classpath has been loaded in its own ucl, so the only other approach would be to disable the manifest-classpath tracking behavior of URLClassloader? I suppose one possiblity would be to rewrite the manifest to omit the classpath when we make the local copy. thanks david jencks On 2002.11.26 14:29:03 -0500 Scott M Stark wrote: Ok, I finally got a handle on the IllegalAccessError that has been reported for a while now and have a trivial testcase that demonstrates the problem. See the bug details here: [ 644289 ] ClassLoader issue with protected access and manifest refs http://sourceforge.net/tracker/index.php?func=detailaid=644289group_id=22866atid=376685 The real problem is the use of a UCL per jar for deployments. Because the java URLClassLoader keeps track of the Class-Path reference in the jar it owns, when more than one jar references a common jar the UCLs associated with the referencing jars are candidates for the defining class loader for the classes in the common jar. Effectively this means that the common jar classes are deployed multiple times even though only a single instance of a given class is created due to the ULR caching. I think the simplest fix is to introduce a common class loader associated with the parent deployment that contains the union of all jars contained in the deployment. In addition to solving this problem it also helps with issues of duplicates classes in a deployment since with a single class loader an ordering of the jars can be specified. I'm working on this issue in main and will backport fixes to 3.2 and 3.0 Scott Stark Chief Technology Officer JBoss Group, LLC --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
Re: [JBoss-dev] where to post correction to jboss 3.0.4 adm. and devel. Second Edition?
Post a patch on sf.net --jason On Tuesday, November 26, 2002, at 01:21 PM, Emerson Cargnin - SICREDI Serviços wrote: where should I post correction to jboss 3.0.4 adm. and devel. Second Edition? by this list? by bug at source forge? -- | Emerson Cargnin | | Analista de Sistemas Sr. | | Tel : (051) 3358-4959| | SICREDI Serviços | | Porto Alegre - Brasil| |xx| --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
Re: [JBoss-dev] What is the status of hsqldb 1.7 in JBoss 3.2?
I am not sure, but I believe the changes to back port might be small... you could post a patch... getting you closer to a RW... if you want it that is. --jason On Tuesday, November 26, 2002, at 01:55 PM, Jeremy Boynes wrote: I'm trying to get PetStore 1.3.1 running with 3.2 but its DAO code uses ResultSet.absolute() which is not supported by hsqldb 1.6 but is by 1.7.1 Will the changes to support 1.7.1 be backported to 3.2 before its release? thanks jeremy --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
[JBoss-dev] [ jboss-Patches-644394 ] support for cmr+cmp not null fields
Patches item #644394, was opened at 2002-11-26 17:04 You can respond by visiting: https://sourceforge.net/tracker/?func=detailatid=376687aid=644394group_id=22866 Category: JBossCMP Group: v3.0 Rabbit Hole Status: Open Resolution: None Priority: 5 Submitted By: Scott Walters (scottw512) Assigned to: Nobody/Anonymous (nobody) Summary: support for cmr+cmp not null fields Initial Comment: This patch addresses issues with fields that exist as both CMP fields and as part of a CMR field, particularly when the field is not null. There are two different issues addressed by these changes. First, there are duplicate column names in the insert and create table sql commands that jboss generates. Second, jboss attempts to insert NULL values for these fields, even when they are part of the primary key. The first issue is dealt with by building a list of unique fields to use for the create table and insert statements. This list of fields contains all the CMP fields followed by the CMR fields that are not in the CMP field list. The second issue is taken care of by the insertion of code to populate the bean instance values for the duplicated CMR fields before the sql insert parameters are bound. This way, when the parameter binding code executes, the correct values are pulled from the bean instance as if they had been set in ejbCreate. This is accomplished by invoking the get method on the local interface argument passed into ejbCreate. The way it works is by querying the duplicated CMR field's related field and extracting it's field name and it's local interface name. Then, the arguments to ejbCreate are inspected for a matching local interface name and a matching get method for the field name. When it finds a match, the get method is invoked and the return value is assigned to the bean instance. For this change to work properly, the local interfaces for the related fields of the duplicated CMR fields must be passed in as arguments to ejbCreate. Also, since the get method is invoked and the bean instance value is set after the ejbCreate call, any code in ejbCreate that sets the field's bean instance value will be overridden. This is the sequence of events. 1) ejbCreate is called normally and returns. 2) the patch code is called to populate duplicated CMR field instance values 3) the primary key is extracted from the bean and stored in the container context 4) the sql insert statement is prepared and executed. 5) ejbPostCreate is called. -- You can respond by visiting: https://sourceforge.net/tracker/?func=detailatid=376687aid=644394group_id=22866 --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
[JBoss-dev] [ jboss-Patches-644397 ] error in JBOSS adm and dev. Sec. Ed.
Patches item #644397, was opened at 2002-11-26 21:09 You can respond by visiting: https://sourceforge.net/tracker/?func=detailatid=376687aid=644397group_id=22866 Category: JBossDoc Group: None Status: Open Resolution: None Priority: 5 Submitted By: Emerson Cargnin (echofloripa) Assigned to: Nobody/Anonymous (nobody) Summary: error in JBOSS adm and dev. Sec. Ed. Initial Comment: There an error in the page 146 of jboss adm and dev. Sec Edt. in the item 4 : It says : The Proxy is translates requests made through the strongly typed... I think there's either an error or a confusion in the sentence.. -- You can respond by visiting: https://sourceforge.net/tracker/?func=detailatid=376687aid=644397group_id=22866 --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
[JBoss-dev] [ jboss-Patches-644398 ] typing error in jboss 3.0.4 adm/dev book
Patches item #644398, was opened at 2002-11-26 21:11 You can respond by visiting: https://sourceforge.net/tracker/?func=detailatid=376687aid=644398group_id=22866 Category: JBossDoc Group: None Status: Open Resolution: None Priority: 5 Submitted By: Emerson Cargnin (echofloripa) Assigned to: Nobody/Anonymous (nobody) Summary: typing error in jboss 3.0.4 adm/dev book Initial Comment: The re a error in the jboss 3.0.4 adm/dev book sec. edt. in the page 149 : The last paragraph, second sentence says : An Inovker simply represents..;. should be (of course :) ) : An Invoker simply represents..;. -- You can respond by visiting: https://sourceforge.net/tracker/?func=detailatid=376687aid=644398group_id=22866 --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
Re: [JBoss-dev] where to post correction to jboss 3.0.4 adm. anddevel. Second Edition?
i did it in JBossDoc category. correct me if it was the wrong place. Jason Dillon wrote: Post a patch on sf.net --jason On Tuesday, November 26, 2002, at 01:21 PM, Emerson Cargnin - SICREDI Serviços wrote: where should I post correction to jboss 3.0.4 adm. and devel. Second Edition? by this list? by bug at source forge? -- | Emerson Cargnin | | Analista de Sistemas Sr. | | Tel : (051) 3358-4959| | SICREDI Serviços | | Porto Alegre - Brasil| |xx| --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development -- | Emerson Cargnin | | Analista de Sistemas Sr. | | Tel : (051) 3358-4959| | SICREDI Serviços | | Porto Alegre - Brasil| |xx| --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
[JBoss-dev] new direction for JBoss AOP
Code is under jboss-head/aop It compiles but that's about it. Its incomplete. The new direction is shown in the below xml if you can follow along. Don't have much time for explanation before Thanksgiving, but the idea is POJOs and DynamicProxies married into one framework. Field access, Constructors and methods interceptable on POJOs. More funtionality can be inferred if you read the following XML. !-- Interceptor definition. This is just a template name is the string reference to the interceptor classname is its classname scope defines how the interceptor is instantiated instance - one interceptor instance per object instance class - one instance per class global - one instance for entire VM config - defines any arbitrary xml to pass to component for initialization -- interceptor name=service1 classname= scope=instance,class,global config/ /interceptor-service !-- Extensions are classes that can be attached to an object to provide extended APIs. name is the string reference to the extention classname is its classname scope defines how the extension is instantiated instance - one extension instance per object instance class - one instance per class global - one instance for entire VM interfaces - are a list of interfaces to extend the aspected object. Any methods invoked on the aspect that pertain to these interfaces will be delegated directly to the extension class config - defines any arbitrary xml to pass to component for initialization i.e. Class MyClass {} interface api { void doSometing(); } Class MyExtension implements api { public void doSomething(){ ... } } You could do in code { MyClass instance = new MyClass(); api a = (api)instance; a.doSomething(); } Even though MyClass doesn't implement the api interface, or has the code for it you can cast and invoke on it anyways. This is the extension. -- extension nameextension1 classname= scope=instance,class,global interfaces/ config/ /extension !-- chain of interceptor definition -- stack name=stack1 interceptor-ref name=service1/ interceptor name=int1 classname= config/ /interceptor /stack !-- interceptor pointcut defines what classes, methods,fields must be intercepted and by what -- interceptor-pointcut name= class=.*pattern* method-pointcut expr=*set*/ !-- Don't intercept method defined in these interfaces -- method-pointcut excluded-interfaces/ /method-pointcut !-- Intercept methods defined in all these interfaces -- method-pointcut included-interfaces/ /method-pointcut field-pointcut expr=*variable*/ field-pointcut expr={PUBLIC,PRIVATE,PROTECTED}/ interceptors stack-ref name=stack1/ interceptor-ref name= config/ /interceptor-ref interceptors/ /interceptor-pointcut !-- Extension pointcut defines what classes should be extended -- extension-pointcut name= class=.* extensions extension-ref name=extension1/ extension/ /extensions /extension-pointcut !-- A proxy is a template for creating Dynamic Proxies. -- proxy name= interfaces/ interceptor-pointcuts/ extensions/ /proxy !-- Add a pointcut to created proxy -- interceptor-pointcut name= proxy= method-pointcut included-interfaces/ /method-pointcut interceptors stack-ref name=stack1/ interceptor-ref name= config/ /interceptor-ref interceptors/ /interceptor-pointcut --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
[JBoss-dev] RE: [jboss-group] new direction for JBoss AOP
Field access, Constructors and methods interceptable on POJOs. With bytecode generation on the framework? well done my friend you are a king, JB4 is the mother of java middleware design marcf --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
[JBoss-dev] jboss-development@lists.sourceforge.net
tisdagen den 26 november 2002 kl 22.55 skrev Jeremy Boynes: I'm trying to get PetStore 1.3.1 running with 3.2 but its DAO code uses ResultSet.absolute() which is not supported by hsqldb 1.6 but is by 1.7.1 Will the changes to support 1.7.1 be backported to 3.2 before its release? Is that a request or an order ? ... ! ... The schedule is/was set for 3.2RC ... tisdagen den 26 november 2002 kl 23.41 skrev Jason Dillon: I am not sure, but I believe the changes to back port might be small... you could post a patch... getting you closer to a RW... if you want it that is. It is really small ... just copy paste ... hehe .. or do a *ci copy* of the head imp ... then test ... test again ... double check ... $£twice´£$ ... will take **time** ... as Jason points out - imp is not hard making sure it works ***is*** ! ... Please DO it ( for me ) ! ... Fred has expressed a will to back ... U(s) up ... if and when ... and Fred == hsqldb team source ... ice ... / U Know Who --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
[JBoss-dev] [ jboss-Bugs-644439 ] Deadlock in ClassLoader
Bugs item #644439, was opened at 2002-11-27 00:40 You can respond by visiting: https://sourceforge.net/tracker/?func=detailatid=376685aid=644439group_id=22866 Category: JBossMX Group: v3.2 Status: Open Resolution: None Priority: 5 Submitted By: Bill Alexander (balexander) Assigned to: Nobody/Anonymous (nobody) Summary: Deadlock in ClassLoader Initial Comment: We are intermittently seeing deadlocks between two threads in classloader code. The scenario goes something like this: Thread 1 org.jboss.mx.loading.HeirarchicalLoaderRepo sitory2.loadClass(java.lang.String, boolean, java.lang.ClassLoader) line: 88 org.jboss.mx.loading.UnifiedClassLoader.load Class(String, boolean) line: 283 java.lang.ClassLoader.loadClass(String) line: 262 java.lang.ClassLoader.loadClassInternal (String) line: 322 java.lang.Class.forName0(String, boolean, ClassLoader) java.lang.Class.forName(String, boolean, ClassLoader) ... remaining, irrelevant stuff deleted. Thread 2 java.lang.ClassLoader.loadClass(String, boolean) line: 288 org.jboss.mx.loading.UnifiedClassLoader.load ClassLocally(String, boolean) line: 234 org.jboss.mx.loading.UnifiedLoaderRepository 2)loadClassFromClassLoader(String, boolean, ClassLoader) line: 234 org.jboss.mx.loading.HeirarchicalLoaderRepo sitory2.loadClass(String, boolean, ClassLoader) line: 94 org.jboss.mx.loading.UnifiedClassLoader.load Class(String, boolean) line: 283 java.lang.ClassLoader.loadClass(String) line: 262 Thread 1 has the lock on the class, and waits for the lock on the repository, while Thread 2 has the lock on the repository, and waits for the lock on the class. We are running JBoss-3.2.0beta on Windows 2000sp2. JDK is 1.4.0_01. Is this bug the same as this forum entry? http://www.jboss.org/forums/thread.jsp? forum=66thread=12454message=3706157q=thread+ deadlock+classloader#3706157 -- You can respond by visiting: https://sourceforge.net/tracker/?func=detailatid=376685aid=644439group_id=22866 --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
RE: [JBoss-dev] new direction for JBoss AOP
wow... This will take AOP to the next level (interceptors on fields and and constructors? WOW.). I'll checkout your stuff tonight. Regards, Hiram -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]On Behalf Of Bill Burke Sent: Tuesday, November 26, 2002 3:39 PM To: Jboss-Dev Cc: Bob Lee; Jboss-Group@Jboss. Org Subject: [JBoss-dev] new direction for JBoss AOP Code is under jboss-head/aop It compiles but that's about it. Its incomplete. The new direction is shown in the below xml if you can follow along. Don't have much time for explanation before Thanksgiving, but the idea is POJOs and DynamicProxies married into one framework. Field access, Constructors and methods interceptable on POJOs. More funtionality can be inferred if you read the following XML. !-- Interceptor definition. This is just a template name is the string reference to the interceptor classname is its classname scope defines how the interceptor is instantiated instance - one interceptor instance per object instance class - one instance per class global - one instance for entire VM config - defines any arbitrary xml to pass to component for initialization -- interceptor name=service1 classname= scope=instance,class,global config/ /interceptor-service !-- Extensions are classes that can be attached to an object to provide extended APIs. name is the string reference to the extention classname is its classname scope defines how the extension is instantiated instance - one extension instance per object instance class - one instance per class global - one instance for entire VM interfaces - are a list of interfaces to extend the aspected object. Any methods invoked on the aspect that pertain to these interfaces will be delegated directly to the extension class config - defines any arbitrary xml to pass to component for initialization i.e. Class MyClass {} interface api { void doSometing(); } Class MyExtension implements api { public void doSomething(){ ... } } You could do in code { MyClass instance = new MyClass(); api a = (api)instance; a.doSomething(); } Even though MyClass doesn't implement the api interface, or has the code for it you can cast and invoke on it anyways. This is the extension. -- extension nameextension1 classname= scope=instance,class,global interfaces/ config/ /extension !-- chain of interceptor definition -- stack name=stack1 interceptor-ref name=service1/ interceptor name=int1 classname= config/ /interceptor /stack !-- interceptor pointcut defines what classes, methods,fields must be intercepted and by what -- interceptor-pointcut name= class=.*pattern* method-pointcut expr=*set*/ !-- Don't intercept method defined in these interfaces -- method-pointcut excluded-interfaces/ /method-pointcut !-- Intercept methods defined in all these interfaces -- method-pointcut included-interfaces/ /method-pointcut field-pointcut expr=*variable*/ field-pointcut expr={PUBLIC,PRIVATE,PROTECTED}/ interceptors stack-ref name=stack1/ interceptor-ref name= config/ /interceptor-ref interceptors/ /interceptor-pointcut !-- Extension pointcut defines what classes should be extended -- extension-pointcut name= class=.* extensions extension-ref name=extension1/ extension/ /extensions /extension-pointcut !-- A proxy is a template for creating Dynamic Proxies. -- proxy name= interfaces/ interceptor-pointcuts/ extensions/ /proxy !-- Add a pointcut to created proxy -- interceptor-pointcut name= proxy= method-pointcut included-interfaces/ /method-pointcut interceptors stack-ref name=stack1/ interceptor-ref name= config/ /interceptor-ref interceptors/ /interceptor-pointcut --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
[JBoss-dev] [ jboss-Bugs-644439 ] Deadlock in ClassLoader
Bugs item #644439, was opened at 2002-11-26 16:40 You can respond by visiting: https://sourceforge.net/tracker/?func=detailatid=376685aid=644439group_id=22866 Category: JBossMX Group: v3.2 Status: Closed Resolution: Out of Date Priority: 5 Submitted By: Bill Alexander (balexander) Assigned to: Nobody/Anonymous (nobody) Summary: Deadlock in ClassLoader Initial Comment: We are intermittently seeing deadlocks between two threads in classloader code. The scenario goes something like this: Thread 1 org.jboss.mx.loading.HeirarchicalLoaderRepo sitory2.loadClass(java.lang.String, boolean, java.lang.ClassLoader) line: 88 org.jboss.mx.loading.UnifiedClassLoader.load Class(String, boolean) line: 283 java.lang.ClassLoader.loadClass(String) line: 262 java.lang.ClassLoader.loadClassInternal (String) line: 322 java.lang.Class.forName0(String, boolean, ClassLoader) java.lang.Class.forName(String, boolean, ClassLoader) ... remaining, irrelevant stuff deleted. Thread 2 java.lang.ClassLoader.loadClass(String, boolean) line: 288 org.jboss.mx.loading.UnifiedClassLoader.load ClassLocally(String, boolean) line: 234 org.jboss.mx.loading.UnifiedLoaderRepository 2)loadClassFromClassLoader(String, boolean, ClassLoader) line: 234 org.jboss.mx.loading.HeirarchicalLoaderRepo sitory2.loadClass(String, boolean, ClassLoader) line: 94 org.jboss.mx.loading.UnifiedClassLoader.load Class(String, boolean) line: 283 java.lang.ClassLoader.loadClass(String) line: 262 Thread 1 has the lock on the class, and waits for the lock on the repository, while Thread 2 has the lock on the repository, and waits for the lock on the class. We are running JBoss-3.2.0beta on Windows 2000sp2. JDK is 1.4.0_01. Is this bug the same as this forum entry? http://www.jboss.org/forums/thread.jsp? forum=66thread=12454message=3706157q=thread+ deadlock+classloader#3706157 -- Comment By: Scott M Stark (starksm) Date: 2002-11-26 17:00 Message: Logged In: YES user_id=175228 This was fixed with a new class loading model introduced in the 3.2beta2. -- You can respond by visiting: https://sourceforge.net/tracker/?func=detailatid=376685aid=644439group_id=22866 --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
[JBoss-dev] Re: jboss-development@lists.sourceforge.net
onsdagen den 27 november 2002 kl 01.36 skrev Peter Fagerlund: The schedule is/was set for 3.2RC ... So in short ... i am not doing it ... show yourselfs ... ( hint ; Jason do not know how to ) ... /thesame --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
Re: [JBoss-dev] ClassLoader issue with protected access and manifest refs
There is really nothing incorrect about the behavior of the current class loading. Its our use of multiple class loaders that is conflicting with the type system. Scott Stark Chief Technology Officer JBoss Group, LLC - Original Message - From: Dain Sundstrom [EMAIL PROTECTED] To: [EMAIL PROTECTED] Sent: Tuesday, November 26, 2002 12:36 PM Subject: Re: [JBoss-dev] ClassLoader issue with protected access and manifest refs Could we just write our own URLClassloader that handles this stuff correctly? We have the URL and Zip classes available. -dain --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
Re: [JBoss-dev] ClassLoader issue with protected access and manifest refs
- Original Message - From: David Jencks [EMAIL PROTECTED] To: [EMAIL PROTECTED] Sent: Tuesday, November 26, 2002 12:50 PM Subject: Re: [JBoss-dev] ClassLoader issue with protected access and manifest refs Won't this eliminate the possiblity of redeploying a subpackage of say an unpacked .ear deployment? In fact, I have been able to redeploy subpackage of a packaged .ear using emacs ability to modify a .jar file's contents and the main deployers ability to redeploy packages explicitly. Not necessarily. I'm thinking that the shared class loader only handles manifest and sar library jars and is the parent class loader of the current DeploymentInfo UCLs. The problem redeployment of a nested component introduces in this scheme is that you have to be able to update the parent component. I don't see this happening now in 3.2 or 3.0, is it being done in 4.0? I take it this behavior occurs even though the jar referenced on the manifest classpath has been loaded in its own ucl, so the only other approach would be to disable the manifest-classpath tracking behavior of URLClassloader? It occurs because the referenced jar is loaded by each referencing UCL. This is what causes the two classes in the same package name to not really be the same package. I suppose one possiblity would be to rewrite the manifest to omit the classpath when we make the local copy. This can't be done in general as the jar could be sealed and have security permissions that include the jar signer. --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
Re: [JBoss-dev] Jetty NIO SL in JBoss 3.0.4 no worky
Curious, but does the ajp (or whatever the apache connector is called) use many threads? If it does, then would it make sense to use nio there too? --jason On Thursday, November 21, 2002, at 08:33 AM, Greg Wilkins wrote: It's important to remember that the jetty nio implementation is not so much aimed at increasing socket performance - rather it's aim is to reduce the number of threads required by avoiding allocating threads to idle connections. This is more to avoid problems on operating systems that become undstable or non linear with large numbers of threads allocated. To date, the few organizations that have tried nio have eventually gone back to using the normal listener. It has helped while they had other threading issues with their systems. Once the OS/JVM have been patched properly, most big irons don't have too much trouble with large numbers of threads. Do selects over large number of sockets can prove to be more difficult for many OS's - even those with /dev/poll. Note that the nature of the servlet/filter API means that the true promise of NIO will never be realized until the spec is radically changed. Some servers may be able to use NIO to assist with serving static content, but if static content is to be made a special case, then you may as well serve it from a kernal server rather than java. While the servlet spec is in it's current revision, you will never get data streaming directly from a direct memory mapped buffer, out the http socket without touching user space. cheers Jules Gosnell wrote: Anything is possible. Anyway, the standard Listener works fine in 1.4 and Greg has chosen to keep it on as the default and make a special case of the NIO Listener which, whilst giving better throughput at high volumes may actually be slower at low ones. I would rather stick with standard Jetty's approach, since I have enough branches to maintain already :-( If anyone wants to go this way badly, they can donate a PortableListener which wrap-n-delegates to a dynamically chosen Jetty Listener and I will alter the config to use this instead. Jules James Higginbotham wrote: Jules, Um... I hate to say this, but I think Weblogic works around this - it defaults to nio and backs off to its older non-nio.. They used to call it their performance pack before 1.4, and have since just swapped theirs for nio I think.. You said it was a static config, but I don't see why you can't abstract out the strategy, and back off to the default strategy if nio isn't available. Just a thought, and may sound simpler than the impl, not sure.. I haven't looked a the Jetty codebase in a while.. James -Original Message- From: Jules Gosnell [mailto:[EMAIL PROTECTED]] Sent: Thursday, November 21, 2002 6:34 AM To: [EMAIL PROTECTED] Subject: Re: [JBoss-dev] Jetty NIO SL in JBoss 3.0.4 no worky Peter Fagerlund wrote: torsdagen den 21 november 2002 kl 11.46 skrev Jason Dillon: Fuck... how can we make one build work for both? Use a runtime switch ... 1. Jetty's config is static - I could work around it 2. You would need to build to the LCD (1.3) and then do another build for optional 1.4 components... Or - we just live with it until we can force a move to 1.4, of course then we will have the same problem with 1.5... Software huh ! Jules --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development __ __ This email has been scanned for all viruses by the MessageLabs SkyScan service. For more information on a proactive anti-virus service working around the clock, around the globe, visit http://www.messagelabs.com __ __ --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development __ __ This email has been scanned for all viruses by the MessageLabs SkyScan service. For more information on a proactive anti-virus service working around the clock, around the globe, visit http://www.messagelabs.com
[JBoss-dev] [ jboss-Bugs-644496 ] a problem of JCA's connection pool
Bugs item #644496, was opened at 2002-11-27 02:48 You can respond by visiting: https://sourceforge.net/tracker/?func=detailatid=376685aid=644496group_id=22866 Category: None Group: None Status: Open Resolution: None Priority: 5 Submitted By: longjianglei (catherine_long) Assigned to: Nobody/Anonymous (nobody) Summary: a problem of JCA's connection pool Initial Comment: Operating System :windows2000 professional JDK Version :v1.3.1 Server trace from the console :none Steps to reproduce the bug : 1.Enviroment: *a jboss server *a MSSQL database server They are not in the same computer. *we use LocalTX *jboss version is 3.0.2 2.Write a EJB on jboss server which has a test method that access the database continuously. 3.At beginning,the EJB's client call this test method,we can see accessing database is OK. 4.Then we disconnect the network between the jboss server and database,we found accessing database fail. 5.When we reconnect the network, we found accessing database always fail and can not success again. Conclusion: 1.We have traced the src code of JCA.We found that when a SQLException ocurred during accessing database,JCA does not kill the bad connection.So,when this connection is put back to connection pool,user will get an invalid connection when they re-get a connection.In the worst case,all of the connections in connection pool are invalid,user can not get any connection again. -- You can respond by visiting: https://sourceforge.net/tracker/?func=detailatid=376685aid=644496group_id=22866 --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
Re: [JBoss-dev] ClassLoader issue with protected access and manifest refs
On 2002.11.26 20:13:14 -0500 Scott M Stark wrote: - Original Message - From: David Jencks [EMAIL PROTECTED] To: [EMAIL PROTECTED] Sent: Tuesday, November 26, 2002 12:50 PM Subject: Re: [JBoss-dev] ClassLoader issue with protected access and manifest refs Won't this eliminate the possiblity of redeploying a subpackage of say an unpacked .ear deployment? In fact, I have been able to redeploy subpackage of a packaged .ear using emacs ability to modify a .jar file's contents and the main deployers ability to redeploy packages explicitly. Not necessarily. I'm thinking that the shared class loader only handles manifest and sar library jars and is the parent class loader of the current DeploymentInfo UCLs. The problem redeployment of a nested component introduces in this scheme is that you have to be able to update the parent component. I don't see this happening now in 3.2 or 3.0, is it being done in 4.0? I don't think so. I don't understand what you mean here. If I have an [expanded] .ear with entities.jar sessions.jar webstuff.war I think I should be able to redeploy: webstuff.war or webstuff.war and sessions.jar or all three. In all cases, I don't see why the .ear should be redeployed (maybe there's something else in it that all of these depend on). Also, I don't see what will happen in your scheme if ejbsA.jar and ejbsB.jar both reference lib.jar in their manifest classpaths. Am I missing the point? I really don't see how to make this work without disabling the manifest classpath stuff, which as you point out, can't simply be ignored. thanks david jencks I take it this behavior occurs even though the jar referenced on the manifest classpath has been loaded in its own ucl, so the only other approach would be to disable the manifest-classpath tracking behavior of URLClassloader? It occurs because the referenced jar is loaded by each referencing UCL. This is what causes the two classes in the same package name to not really be the same package. I suppose one possiblity would be to rewrite the manifest to omit the classpath when we make the local copy. This can't be done in general as the jar could be sealed and have security permissions that include the jar signer. --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
[JBoss-dev] Automated JBoss(JBoss_3_2_0_beta2 WonderLand) Testsuite Results: 26-November-2002
Number of tests run: 1038 Successful tests: 1024 Errors:12 Failures: 2 [time of test: 2002-11-27.02-28 GMT] [java.version: 1.3.1_05] [java.vendor: Sun Microsystems Inc.] [java.vm.version: 1.3.1_05-b02] [java.vm.name: Java HotSpot(TM) Client VM] [java.vm.info: mixed mode] [os.name: Windows 2000] [os.arch: x86] [os.version: 5.0] Useful resources: - http://users.jboss.org/~starksm/Branch_3_2/2002-11-27.02-28 for the junit report of this test. - http://users.jboss.org/~starksm/Branch_3_2/2002-11-27.02-28/logs/ for the logs for this test. NOTE: If there are any errors shown above - this mail is only highlighting them - it is NOT indicating that they are being looked at by anyone. Remember - if a test becomes broken after your changes - fix it or fix the test! Oh dear - still got some errors! Thanks for all your effort - we really do love you! --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
[JBoss-dev] [ jboss-Bugs-644496 ] a problem of JCA's connection pool
Bugs item #644496, was opened at 2002-11-27 02:48 You can respond by visiting: https://sourceforge.net/tracker/?func=detailatid=376685aid=644496group_id=22866 Category: JBossServer Group: None Status: Open Resolution: None Priority: 5 Submitted By: longjianglei (catherine_long) Assigned to: David Jencks (d_jencks) Summary: a problem of JCA's connection pool Initial Comment: Operating System :windows2000 professional JDK Version :v1.3.1 Server trace from the console :none Steps to reproduce the bug : 1.Enviroment: *a jboss server *a MSSQL database server They are not in the same computer. *we use LocalTX *jboss version is 3.0.2 2.Write a EJB on jboss server which has a test method that access the database continuously. 3.At beginning,the EJB's client call this test method,we can see accessing database is OK. 4.Then we disconnect the network between the jboss server and database,we found accessing database fail. 5.When we reconnect the network, we found accessing database always fail and can not success again. Conclusion: 1.We have traced the src code of JCA.We found that when a SQLException ocurred during accessing database,JCA does not kill the bad connection.So,when this connection is put back to connection pool,user will get an invalid connection when they re-get a connection.In the worst case,all of the connections in connection pool are invalid,user can not get any connection again. -- You can respond by visiting: https://sourceforge.net/tracker/?func=detailatid=376685aid=644496group_id=22866 --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
[JBoss-dev] 3 Transactions?
OK we have a transaction associated with the thread, each invocation and the each context has a transaction (very confusing). Once we get by the tx interceptor should all of these be the same thing? If we are suspending the transaction for the invocation (Not Supported) should the ctx and invocation tx be null or just a suspended transaction? -dain --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
Re: [JBoss-dev] ClassLoader issue with protected access and manifest refs
this happening now in 3.2 or 3.0, is it being done in 4.0? I don't think so. I don't understand what you mean here. If I have an [expanded] .ear with entities.jar sessions.jar webstuff.war I think I should be able to redeploy: webstuff.war or webstuff.war and sessions.jar or all three. In all cases, I don't see why the .ear should be redeployed (maybe there's something else in it that all of these depend on). The ear DeploymentInfo has references to the subdeployment DeploymentInfo objects. How are these being updated when a subdeployment is redeployed? That is what I'm referring to. Also, I don't see what will happen in your scheme if ejbsA.jar and ejbsB.jar both reference lib.jar in their manifest classpaths. Am I missing the point? I really don't see how to make this work without disabling the manifest classpath stuff, which as you point out, can't simply be ignored. There will be a UCL_ejbsA.jar, UCL_ejbsB.jar and a UCL_shared(lib.jar) that is the parent class loader for both UCL_ejbsA.jar and UCL_ejbsB.jar. Since the parent delegation model takes precedence over the manifests this resolves the type problem as UCL_shared(lib.jar) contains all referenced jars. By virtue of there only being one class loader package protected access problems disappear. --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
Re: [JBoss-dev] 3 Transactions?
On 2002.11.26 22:26:30 -0500 Dain Sundstrom wrote: OK we have a transaction associated with the thread, each invocation and the each context has a transaction (very confusing). Once we get by the tx interceptor should all of these be the same thing? yes (at least the thread and invocation, I don't know about the context, isn't that a UserTx, which is almost completely unrelated?) If we are suspending the transaction for the invocation (Not Supported) should the ctx and invocation tx be null or just a suspended transaction? Invocation should have null. You shouldn't be able to tell if there was a tx associated upstream. IMO the invocation is the master and the threadlocal is there in case you can't get to the invocation (such as in the jca stuff). So the tx interceptor needs to figure out what the current tx actually is, and make sure the txManager knows it so it can supply it per thread. david jencks -dain --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
[JBoss-dev] Exploded format in JBoss
Hello all, Where can I put the ejb components and the web components in the JBoss heirarchy in an exploded format. Thanks in advance Rosh Nair __ Do you Yahoo!? Yahoo! Mail Plus - Powerful. Affordable. Sign up now. http://mailplus.yahoo.com --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
Re: [JBoss-dev] 3 Transactions?
Is that the way is is supposed to work, because it is not the way it is today. When we have a Not Supported method the invocation in the context is not reset to null. I can make the change, but I want to be certain that this is the desired behavior. Also does anyone know the correct handling of the enterprise context transaction. Is it only for BMP beans? Should it be null otherwise? I would say yes to both. -dain On Tuesday, November 26, 2002, at 10:23 PM, David Jencks wrote: On 2002.11.26 22:26:30 -0500 Dain Sundstrom wrote: OK we have a transaction associated with the thread, each invocation and the each context has a transaction (very confusing). Once we get by the tx interceptor should all of these be the same thing? yes (at least the thread and invocation, I don't know about the context, isn't that a UserTx, which is almost completely unrelated?) If we are suspending the transaction for the invocation (Not Supported) should the ctx and invocation tx be null or just a suspended transaction? Invocation should have null. You shouldn't be able to tell if there was a tx associated upstream. IMO the invocation is the master and the threadlocal is there in case you can't get to the invocation (such as in the jca stuff). So the tx interceptor needs to figure out what the current tx actually is, and make sure the txManager knows it so it can supply it per thread. david jencks -dain --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
Re: [JBoss-dev] 3 Transactions?
Is it legal to call getTransaction on the TransactionManager in afterComplete? It seems to work, but is it legal? -dain On Tuesday, November 26, 2002, at 11:55 PM, Dain Sundstrom wrote: Is that the way is is supposed to work, because it is not the way it is today. When we have a Not Supported method the invocation in the context is not reset to null. I can make the change, but I want to be certain that this is the desired behavior. Also does anyone know the correct handling of the enterprise context transaction. Is it only for BMP beans? Should it be null otherwise? I would say yes to both. -dain On Tuesday, November 26, 2002, at 10:23 PM, David Jencks wrote: On 2002.11.26 22:26:30 -0500 Dain Sundstrom wrote: OK we have a transaction associated with the thread, each invocation and the each context has a transaction (very confusing). Once we get by the tx interceptor should all of these be the same thing? yes (at least the thread and invocation, I don't know about the context, isn't that a UserTx, which is almost completely unrelated?) If we are suspending the transaction for the invocation (Not Supported) should the ctx and invocation tx be null or just a suspended transaction? Invocation should have null. You shouldn't be able to tell if there was a tx associated upstream. IMO the invocation is the master and the threadlocal is there in case you can't get to the invocation (such as in the jca stuff). So the tx interceptor needs to figure out what the current tx actually is, and make sure the txManager knows it so it can supply it per thread. david jencks -dain --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
RE: [JBoss-dev] 3 Transactions?
David, Is a ThreadLocal lookup more expensive than a hashmap lookup?? Regards, Hiram -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]On Behalf Of David Jencks Sent: Tuesday, November 26, 2002 8:23 PM To: [EMAIL PROTECTED] Subject: Re: [JBoss-dev] 3 Transactions? On 2002.11.26 22:26:30 -0500 Dain Sundstrom wrote: OK we have a transaction associated with the thread, each invocation and the each context has a transaction (very confusing). Once we get by the tx interceptor should all of these be the same thing? yes (at least the thread and invocation, I don't know about the context, isn't that a UserTx, which is almost completely unrelated?) If we are suspending the transaction for the invocation (Not Supported) should the ctx and invocation tx be null or just a suspended transaction? Invocation should have null. You shouldn't be able to tell if there was a tx associated upstream. IMO the invocation is the master and the threadlocal is there in case you can't get to the invocation (such as in the jca stuff). So the tx interceptor needs to figure out what the current tx actually is, and make sure the txManager knows it so it can supply it per thread. david jencks -dain --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
RE: [JBoss-dev] Exploded format in JBoss
I assume you can define where that is by enabling the following MBean in jboss-service.xml: !-- An alternate hot deployer mbean. Use URLDirectoryScanner if you need to specify which urls are directories to scan, and which are urls to be deployed directly. URLDeploymentScanner assumes that all directories are to be scanned, which can cause problems if directory referred to in the deploy attribute is an exploded archive. -- mbean code=org.jboss.deployment.scanner.URLDirectoryScanner name=jboss.deployment:type=DeploymentScanner,flavor=URL Cheers, Fred -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]] On Behalf Of Rosh Nair Sent: Tuesday, November 26, 2002 11:28 PM To: [EMAIL PROTECTED] Subject: [JBoss-dev] Exploded format in JBoss Hello all, Where can I put the ejb components and the web components in the JBoss heirarchy in an exploded format. Thanks in advance Rosh Nair __ Do you Yahoo!? Yahoo! Mail Plus - Powerful. Affordable. Sign up now. http://mailplus.yahoo.com --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development --- This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. Power Color in a compact size! http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
Re: [JBoss-dev] Jetty NIO SL in JBoss 3.0.4 no worky
Greg is in Geneva for the week, but when he gets back, I'll ask him about this, Jules Jason Dillon wrote: Curious, but does the ajp (or whatever the apache connector is called) use many threads? If it does, then would it make sense to use nio there too? --jason On Thursday, November 21, 2002, at 08:33 AM, Greg Wilkins wrote: It's important to remember that the jetty nio implementation is not so much aimed at increasing socket performance - rather it's aim is to reduce the number of threads required by avoiding allocating threads to idle connections. This is more to avoid problems on operating systems that become undstable or non linear with large numbers of threads allocated. To date, the few organizations that have tried nio have eventually gone back to using the normal listener. It has helped while they had other threading issues with their systems. Once the OS/JVM have been patched properly, most big irons don't have too much trouble with large numbers of threads. Do selects over large number of sockets can prove to be more difficult for many OS's - even those with /dev/poll. Note that the nature of the servlet/filter API means that the true promise of NIO will never be realized until the spec is radically changed. Some servers may be able to use NIO to assist with serving static content, but if static content is to be made a special case, then you may as well serve it from a kernal server rather than java. While the servlet spec is in it's current revision, you will never get data streaming directly from a direct memory mapped buffer, out the http socket without touching user space. cheers Jules Gosnell wrote: Anything is possible. Anyway, the standard Listener works fine in 1.4 and Greg has chosen to keep it on as the default and make a special case of the NIO Listener which, whilst giving better throughput at high volumes may actually be slower at low ones. I would rather stick with standard Jetty's approach, since I have enough branches to maintain already :-( If anyone wants to go this way badly, they can donate a PortableListener which wrap-n-delegates to a dynamically chosen Jetty Listener and I will alter the config to use this instead. Jules James Higginbotham wrote: Jules, Um... I hate to say this, but I think Weblogic works around this - it defaults to nio and backs off to its older non-nio.. They used to call it their performance pack before 1.4, and have since just swapped theirs for nio I think.. You said it was a static config, but I don't see why you can't abstract out the strategy, and back off to the default strategy if nio isn't available. Just a thought, and may sound simpler than the impl, not sure.. I haven't looked a the Jetty codebase in a while.. James -Original Message- From: Jules Gosnell [mailto:[EMAIL PROTECTED]] Sent: Thursday, November 21, 2002 6:34 AM To: [EMAIL PROTECTED] Subject: Re: [JBoss-dev] Jetty NIO SL in JBoss 3.0.4 no worky Peter Fagerlund wrote: torsdagen den 21 november 2002 kl 11.46 skrev Jason Dillon: Fuck... how can we make one build work for both? Use a runtime switch ... 1. Jetty's config is static - I could work around it 2. You would need to build to the LCD (1.3) and then do another build for optional 1.4 components... Or - we just live with it until we can force a move to 1.4, of course then we will have the same problem with 1.5... Software huh ! Jules --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development __ __ This email has been scanned for all viruses by the MessageLabs SkyScan service. For more information on a proactive anti-virus service working around the clock, around the globe, visit http://www.messagelabs.com __ __ --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Jboss-development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development __ __ This email has been scanned for all viruses by the MessageLabs SkyScan service. For more information on a proactive anti-virus service working around the clock, around