[JBoss JIRA] (WFLY-40) AS7.1 jars run against as7.2/eap6.1 cause test case (eclipse) to hang
by RH Bugzilla Integration (JIRA)
[ https://issues.jboss.org/browse/WFLY-40?page=com.atlassian.jira.plugin.sy... ]
RH Bugzilla Integration commented on WFLY-40:
---------------------------------------------
Ladislav Thon <lthon(a)redhat.com> made a comment on [bug 927318|https://bugzilla.redhat.com/show_bug.cgi?id=927318]
Verified with EAP 6.1.0.ER5.
> AS7.1 jars run against as7.2/eap6.1 cause test case (eclipse) to hang
> ---------------------------------------------------------------------
>
> Key: WFLY-40
> URL: https://issues.jboss.org/browse/WFLY-40
> Project: WildFly
> Issue Type: Bug
> Components: Remoting
> Reporter: Rob Stryker
> Assignee: Emanuel Muckenhuber
> Priority: Blocker
> Fix For: 8.0.0.Alpha1
>
> Attachments: AS7-6779.out2.txt, AS7-6779.stack.txt, out.zip
>
>
> Somewhere along the line recently, EAP6.1 / AS 7.2 has a regression in remoting or xnio somehow, and this is causing blocking problems for tools.
> Recall:
> 1) JBossTools bundles a set of jars capable of communicating with a running application server
> 2) This single set of jars must be capable of communicating with all servers in the 7.x stream
> 3) Bundling a second set of jars is not easily accomplished, but could be possible
> And:
> 4) Using the as7.1.0 jars against eap6.1 in a specific unit test causes eclipse to hang
> The basic code of our test case (once we get rid of the jbt stuff wrapped around it) is basically as follows:
> {code}
> this.client = ModelControllerClient.Factory.create(details.getHost(), details.getManagementPort(),
> getCallbackHandler());
> this.manager = ServerDeploymentManager.Factory.create(client);
> DeploymentPlanBuilder builder = manager.newDeploymentPlan().replace(name, file);
> try {
> DeploymentAction action = builder.getLastAction();
> Future<ServerDeploymentPlanResult> planResult = manager.execute(builder.build());
> // FREEZE HERE
> // This is "DeploymentOperationResult.getStatus() in the stack trace
> ServerDeploymentActionResult actionResult = planResult.get().getDeploymentActionResult(action.getId());
> IStatus status = createStatus(action.getDeploymentUnitUniqueName(), action.getType().name(), actionResult);
> } catch (Exception e) {
> throw new JBoss7ManangerException(e);
> }
> {code}
> The actual stack trace is:
> {code}
> Thread [main] (Suspended)
> owns: RunnableLock (id=461)
> waiting for: ActiveOperationSupport$ActiveOperationImpl<T,A> (id=462)
> Object.wait(long) line: not available [native method]
> ActiveOperationSupport$ActiveOperationImpl<T,A>(Object).wait() line: 503
> ActiveOperationSupport$ActiveOperationImpl<T,A>(AsyncFutureTask<T>).await() line: 192
> ActiveOperationSupport$ActiveOperationImpl<T,A>(AsyncFutureTask<T>).get() line: 266
> AbstractModelControllerClient$DelegatingCancellableAsyncFuture.get() line: 363
> AbstractModelControllerClient$DelegatingCancellableAsyncFuture.get() line: 317
> ServerDeploymentPlanResultFuture.get() line: 76
> ServerDeploymentPlanResultFuture.get() line: 42
> DeploymentOperationResult.getStatus() line: 52
> AS7ManagerTestUtils.waitUntilFinished(IJBoss7DeploymentResult) line: 93
> AS7ManagerIntegrationTest.canReplaceWar() line: 129
> {code}
> The list of jars we are using to communicate with the server are:
> {code}
> jboss-as-controller-client-7.1.0.Final.jar
> jboss-as-protocol-7.1.0.Final.jar
> jboss-dmr-1.1.1.Final.jar
> jboss-logging-3.1.0.GA.jar
> jboss-marshalling-1.3.9.GA.jar
> jboss-remoting-3.2.7.GA.jar
> jboss-sasl-1.0.0.Final.jar
> jboss-threads-2.0.0.GA.jar
> xnio-api-3.0.3.GA.jar
> xnio-nio-3.0.3.GA.jar
> {code}
> It was suggested in irc by ctomc that we simply replace the following jars with new jars from jboss-as 7.2 / eap6.1:
> {code}
> jboss-remoting-3.2.7.GA.jar -> 3.2.15.GA
> xnio-api-3.0.3.GA.jar -> 3.0.7
> xnio-nio-3.0.3.GA.jar -> 3.0.7
> {code}
> This approach also failed, with the following stack trace:
> {code}
> java.lang.NoClassDefFoundError: org/xnio/Cancellable
> (sic)
> at java.lang.ClassLoader.loadClass(ClassLoader.java:356)
> at org.jboss.as.controller.client.impl.RemotingModelControllerClient.<init>(RemotingModelControllerClient.java:59)
> at org.jboss.as.controller.client.ModelControllerClient$Factory.create(ModelControllerClient.java:211)
> at org.jboss.as.controller.client.ModelControllerClient$Factory.create(ModelControllerClient.java:173)
> {code}
> So it is obvious that jboss-as-controller jar from as-7.1 requires classes that are no longer present in xnio. If this is the case, it would indicate that xnio has REMOVED classes from its jar during an incremental version from 3.0.3 to 3.0.7, which would constitute an API breakage.
> So clearly for us, updating only remoting and xnio does not work. The next possible solution would be to update ALL jars, however, this led to a similar situation when using AS7.2 / Eap 6.1 jars to communicate with an AS 7.0.0 server. Again, similar to what was mentioned above, all tests passed EXCEPT the test above. The test above failed with an almost identical stack trace:
> {code}
> Thread [main] (Suspended)
> owns: RunnableLock (id=160)
> waiting for: ActiveOperationSupport$ActiveOperationImpl<T,A> (id=161)
> Object.wait(long) line: not available [native method]
> ActiveOperationSupport$ActiveOperationImpl<T,A>(Object).wait() line: 503
> ActiveOperationSupport$ActiveOperationImpl<T,A>(AsyncFutureTask<T>).await() line: 192
> ActiveOperationSupport$ActiveOperationImpl<T,A>(AsyncFutureTask<T>).get() line: 266
> AbstractModelControllerClient$DelegatingCancellableAsyncFuture(AbstractDelegatingAsyncFuture<T>).get() line: 100
> ServerDeploymentPlanResultFuture.get() line: 76
> ServerDeploymentPlanResultFuture.get() line: 42
> DeploymentOperationResult.getStatus() line: 52
> AS7ManagerTestUtils.waitUntilFinished(IJBoss7DeploymentResult) line: 93
> AS7ManagerTestUtils.quietlyUndeploy(String, AS71Manager) line: 76
> AS7ManagerIntegrationTest.canReplaceWar() line: 135
> {code}
> No matter how you look at it, the contract has been broken. JBossTools now is looking for a specific set of jars which care capable of communicating with AS7.0, 7.1, 7.2, eap6.0, 6.1, jpp6.0, etc. Without a set of jars that can communicate with all of these servers, our tools will need to implement drastic workarounds or bundle several versions of the app server's jars within it.
> It's interesting to me that the as7.1 jars work against everything except eap6.1, and yet, the eap6.1 jars work against everything except as7.0.0.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
13 years, 2 months
[JBoss JIRA] (WFLY-40) AS7.1 jars run against as7.2/eap6.1 cause test case (eclipse) to hang
by RH Bugzilla Integration (JIRA)
[ https://issues.jboss.org/browse/WFLY-40?page=com.atlassian.jira.plugin.sy... ]
RH Bugzilla Integration commented on WFLY-40:
---------------------------------------------
Ladislav Thon <lthon(a)redhat.com> changed the Status of [bug 927318|https://bugzilla.redhat.com/show_bug.cgi?id=927318] from ON_QA to VERIFIED
> AS7.1 jars run against as7.2/eap6.1 cause test case (eclipse) to hang
> ---------------------------------------------------------------------
>
> Key: WFLY-40
> URL: https://issues.jboss.org/browse/WFLY-40
> Project: WildFly
> Issue Type: Bug
> Components: Remoting
> Reporter: Rob Stryker
> Assignee: Emanuel Muckenhuber
> Priority: Blocker
> Fix For: 8.0.0.Alpha1
>
> Attachments: AS7-6779.out2.txt, AS7-6779.stack.txt, out.zip
>
>
> Somewhere along the line recently, EAP6.1 / AS 7.2 has a regression in remoting or xnio somehow, and this is causing blocking problems for tools.
> Recall:
> 1) JBossTools bundles a set of jars capable of communicating with a running application server
> 2) This single set of jars must be capable of communicating with all servers in the 7.x stream
> 3) Bundling a second set of jars is not easily accomplished, but could be possible
> And:
> 4) Using the as7.1.0 jars against eap6.1 in a specific unit test causes eclipse to hang
> The basic code of our test case (once we get rid of the jbt stuff wrapped around it) is basically as follows:
> {code}
> this.client = ModelControllerClient.Factory.create(details.getHost(), details.getManagementPort(),
> getCallbackHandler());
> this.manager = ServerDeploymentManager.Factory.create(client);
> DeploymentPlanBuilder builder = manager.newDeploymentPlan().replace(name, file);
> try {
> DeploymentAction action = builder.getLastAction();
> Future<ServerDeploymentPlanResult> planResult = manager.execute(builder.build());
> // FREEZE HERE
> // This is "DeploymentOperationResult.getStatus() in the stack trace
> ServerDeploymentActionResult actionResult = planResult.get().getDeploymentActionResult(action.getId());
> IStatus status = createStatus(action.getDeploymentUnitUniqueName(), action.getType().name(), actionResult);
> } catch (Exception e) {
> throw new JBoss7ManangerException(e);
> }
> {code}
> The actual stack trace is:
> {code}
> Thread [main] (Suspended)
> owns: RunnableLock (id=461)
> waiting for: ActiveOperationSupport$ActiveOperationImpl<T,A> (id=462)
> Object.wait(long) line: not available [native method]
> ActiveOperationSupport$ActiveOperationImpl<T,A>(Object).wait() line: 503
> ActiveOperationSupport$ActiveOperationImpl<T,A>(AsyncFutureTask<T>).await() line: 192
> ActiveOperationSupport$ActiveOperationImpl<T,A>(AsyncFutureTask<T>).get() line: 266
> AbstractModelControllerClient$DelegatingCancellableAsyncFuture.get() line: 363
> AbstractModelControllerClient$DelegatingCancellableAsyncFuture.get() line: 317
> ServerDeploymentPlanResultFuture.get() line: 76
> ServerDeploymentPlanResultFuture.get() line: 42
> DeploymentOperationResult.getStatus() line: 52
> AS7ManagerTestUtils.waitUntilFinished(IJBoss7DeploymentResult) line: 93
> AS7ManagerIntegrationTest.canReplaceWar() line: 129
> {code}
> The list of jars we are using to communicate with the server are:
> {code}
> jboss-as-controller-client-7.1.0.Final.jar
> jboss-as-protocol-7.1.0.Final.jar
> jboss-dmr-1.1.1.Final.jar
> jboss-logging-3.1.0.GA.jar
> jboss-marshalling-1.3.9.GA.jar
> jboss-remoting-3.2.7.GA.jar
> jboss-sasl-1.0.0.Final.jar
> jboss-threads-2.0.0.GA.jar
> xnio-api-3.0.3.GA.jar
> xnio-nio-3.0.3.GA.jar
> {code}
> It was suggested in irc by ctomc that we simply replace the following jars with new jars from jboss-as 7.2 / eap6.1:
> {code}
> jboss-remoting-3.2.7.GA.jar -> 3.2.15.GA
> xnio-api-3.0.3.GA.jar -> 3.0.7
> xnio-nio-3.0.3.GA.jar -> 3.0.7
> {code}
> This approach also failed, with the following stack trace:
> {code}
> java.lang.NoClassDefFoundError: org/xnio/Cancellable
> (sic)
> at java.lang.ClassLoader.loadClass(ClassLoader.java:356)
> at org.jboss.as.controller.client.impl.RemotingModelControllerClient.<init>(RemotingModelControllerClient.java:59)
> at org.jboss.as.controller.client.ModelControllerClient$Factory.create(ModelControllerClient.java:211)
> at org.jboss.as.controller.client.ModelControllerClient$Factory.create(ModelControllerClient.java:173)
> {code}
> So it is obvious that jboss-as-controller jar from as-7.1 requires classes that are no longer present in xnio. If this is the case, it would indicate that xnio has REMOVED classes from its jar during an incremental version from 3.0.3 to 3.0.7, which would constitute an API breakage.
> So clearly for us, updating only remoting and xnio does not work. The next possible solution would be to update ALL jars, however, this led to a similar situation when using AS7.2 / Eap 6.1 jars to communicate with an AS 7.0.0 server. Again, similar to what was mentioned above, all tests passed EXCEPT the test above. The test above failed with an almost identical stack trace:
> {code}
> Thread [main] (Suspended)
> owns: RunnableLock (id=160)
> waiting for: ActiveOperationSupport$ActiveOperationImpl<T,A> (id=161)
> Object.wait(long) line: not available [native method]
> ActiveOperationSupport$ActiveOperationImpl<T,A>(Object).wait() line: 503
> ActiveOperationSupport$ActiveOperationImpl<T,A>(AsyncFutureTask<T>).await() line: 192
> ActiveOperationSupport$ActiveOperationImpl<T,A>(AsyncFutureTask<T>).get() line: 266
> AbstractModelControllerClient$DelegatingCancellableAsyncFuture(AbstractDelegatingAsyncFuture<T>).get() line: 100
> ServerDeploymentPlanResultFuture.get() line: 76
> ServerDeploymentPlanResultFuture.get() line: 42
> DeploymentOperationResult.getStatus() line: 52
> AS7ManagerTestUtils.waitUntilFinished(IJBoss7DeploymentResult) line: 93
> AS7ManagerTestUtils.quietlyUndeploy(String, AS71Manager) line: 76
> AS7ManagerIntegrationTest.canReplaceWar() line: 135
> {code}
> No matter how you look at it, the contract has been broken. JBossTools now is looking for a specific set of jars which care capable of communicating with AS7.0, 7.1, 7.2, eap6.0, 6.1, jpp6.0, etc. Without a set of jars that can communicate with all of these servers, our tools will need to implement drastic workarounds or bundle several versions of the app server's jars within it.
> It's interesting to me that the as7.1 jars work against everything except eap6.1, and yet, the eap6.1 jars work against everything except as7.0.0.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
13 years, 2 months
[JBoss JIRA] (WFLY-1289) Spurious warnings about invalid META-INF/services class name
by David Lloyd (JIRA)
[ https://issues.jboss.org/browse/WFLY-1289?page=com.atlassian.jira.plugin.... ]
David Lloyd commented on WFLY-1289:
-----------------------------------
In reality a class name has very few constraints; it is a series of "unqualified names" separated by periods. The JVM spec §4.2.2 describes unqualified names:
{quote}
An unqualified name must not contain any of the ASCII characters . ; [ / (that is, period or semicolon or left square bracket or forward slash).
{quote}
Other than these, a class name can contain *any* Unicode code point. So we clearly have some adjustment to do here. :)
> Spurious warnings about invalid META-INF/services class name
> ------------------------------------------------------------
>
> Key: WFLY-1289
> URL: https://issues.jboss.org/browse/WFLY-1289
> Project: WildFly
> Issue Type: Bug
> Components: Class Loading
> Environment: 7.2.0.Final
> java version "1.7.0_21"
> Ubuntu 12.04 LTS
> Reporter: Harald Wellmann
> Assignee: David Lloyd
>
> When an application contains a {{META-INF/services}} resource referencing an inner class, JBoss prints a spurious warning
> {noformat}
> JBAS015893: Encountered invalid class name 'com.acme.Foo$Bar' for service type 'com.acme.Service'
> {noformat}
> {{META-INF/services}} lists classes with their binary name, so dollar characters are likely to occur.
> {{org.jboss.as.server.deployment.ServiceLoaderProcessor}} scans all {{META-INF/services}} resources but does not accept dollar characters in its validation pattern.
> (Note: I can't select any 7.x release for "Affects Version".)
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
13 years, 2 months
[JBoss JIRA] (WFLY-133) ClusteredSingleSignOn ClassCastException with non-distributable apps
by RH Bugzilla Integration (JIRA)
[ https://issues.jboss.org/browse/WFLY-133?page=com.atlassian.jira.plugin.s... ]
RH Bugzilla Integration commented on WFLY-133:
----------------------------------------------
Jitka Kudrnacova <jkudrnac(a)redhat.com> changed the Status of [bug 920367|https://bugzilla.redhat.com/show_bug.cgi?id=920367] from ON_QA to VERIFIED
> ClusteredSingleSignOn ClassCastException with non-distributable apps
> --------------------------------------------------------------------
>
> Key: WFLY-133
> URL: https://issues.jboss.org/browse/WFLY-133
> Project: WildFly
> Issue Type: Bug
> Components: Web (JBoss Web)
> Reporter: Dennis Reed
> Assignee: Jesper Pedersen
> Fix For: 8.0.0.Alpha1
>
> Attachments: test.war
>
>
> When ClusteredSingleSignOn is enabled, and a user has an active session in a non-distributable web application, invalidating any session in the same SSO from a different AS instance in the cluster causes the following exception (logged numerous times in various ways):
> ERROR [org.infinispan.interceptors.InvocationContextInterceptor] (OOB-20,null) ISPN000136: Execution error: org.infinispan.CacheException: Caught exception [java.lang.ClassCastException] while invoking method [public void org.jboss.as.clustering.web.sso.infinispan.SSOClusterManager.cacheEntryRemoved(org.infinispan.notifications.cachelistener.event.CacheEntryRemovedEvent)] on listener instance: org.jboss.as.clustering.web.sso.infinispan.SSOClusterManager@28e13c84
> ...
> Caused by: java.lang.ClassCastException: org.apache.catalina.session.StandardSession cannot be cast to org.jboss.as.web.session.ClusteredSession
> at org.jboss.as.web.sso.ClusteredSingleSignOn.deregisterLocal(ClusteredSingleSignOn.java:523) [jboss-as-web-7.1.3.Final-redhat-4.jar:7.1.3.Final-redhat-4]
> ...
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
13 years, 2 months
[JBoss JIRA] (WFLY-133) ClusteredSingleSignOn ClassCastException with non-distributable apps
by RH Bugzilla Integration (JIRA)
[ https://issues.jboss.org/browse/WFLY-133?page=com.atlassian.jira.plugin.s... ]
RH Bugzilla Integration commented on WFLY-133:
----------------------------------------------
Jitka Kudrnacova <jkudrnac(a)redhat.com> made a comment on [bug 920367|https://bugzilla.redhat.com/show_bug.cgi?id=920367]
Verified.
> ClusteredSingleSignOn ClassCastException with non-distributable apps
> --------------------------------------------------------------------
>
> Key: WFLY-133
> URL: https://issues.jboss.org/browse/WFLY-133
> Project: WildFly
> Issue Type: Bug
> Components: Web (JBoss Web)
> Reporter: Dennis Reed
> Assignee: Jesper Pedersen
> Fix For: 8.0.0.Alpha1
>
> Attachments: test.war
>
>
> When ClusteredSingleSignOn is enabled, and a user has an active session in a non-distributable web application, invalidating any session in the same SSO from a different AS instance in the cluster causes the following exception (logged numerous times in various ways):
> ERROR [org.infinispan.interceptors.InvocationContextInterceptor] (OOB-20,null) ISPN000136: Execution error: org.infinispan.CacheException: Caught exception [java.lang.ClassCastException] while invoking method [public void org.jboss.as.clustering.web.sso.infinispan.SSOClusterManager.cacheEntryRemoved(org.infinispan.notifications.cachelistener.event.CacheEntryRemovedEvent)] on listener instance: org.jboss.as.clustering.web.sso.infinispan.SSOClusterManager@28e13c84
> ...
> Caused by: java.lang.ClassCastException: org.apache.catalina.session.StandardSession cannot be cast to org.jboss.as.web.session.ClusteredSession
> at org.jboss.as.web.sso.ClusteredSingleSignOn.deregisterLocal(ClusteredSingleSignOn.java:523) [jboss-as-web-7.1.3.Final-redhat-4.jar:7.1.3.Final-redhat-4]
> ...
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
13 years, 2 months
[JBoss JIRA] (WFLY-1298) Possible race when restoring persistent bundles
by Thomas Diesler (JIRA)
Thomas Diesler created WFLY-1298:
------------------------------------
Summary: Possible race when restoring persistent bundles
Key: WFLY-1298
URL: https://issues.jboss.org/browse/WFLY-1298
Project: WildFly
Issue Type: Bug
Components: OSGi
Reporter: Thomas Diesler
Assignee: Thomas Diesler
There is a possible race when restoring persistent bundles. The FrameworkActivateProcessor may want to activate the framework before FrameworkBootstrapService.start() completes.
{code}
[0m [31m14:33:49,853 ERROR [org.jboss.msc.service.fail] (MSC service thread 1-3) MSC00001: Failed to start service jboss.deployment.unit."de.xxx.yyy.data.services.jar".PARSE: org.jboss.msc.service.StartException in service jboss.deployment.unit."de.xxx.yyy.data.services.jar".PARSE: JBAS018733: Failed to process phase PARSE of deployment "de.xxx.yyy.data.services.jar"
at org.jboss.as.server.deployment.DeploymentUnitPhaseService.start(DeploymentUnitPhaseService.java:127) [jboss-as-server-7.2.0.Alpha1-redhat-4.jar:7.2.0.Alpha1-redhat-4]
at org.jboss.msc.service.ServiceControllerImpl$StartTask.startService(ServiceControllerImpl.java:1811) [jboss-msc-1.0.4.GA.jar:1.0.4.GA]
at org.jboss.msc.service.ServiceControllerImpl$StartTask.run(ServiceControllerImpl.java:1746) [jboss-msc-1.0.4.GA.jar:1.0.4.GA]
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1110) [rt.jar:1.7.0_05]
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:603) [rt.jar:1.7.0_05]
at java.lang.Thread.run(Thread.java:722) [rt.jar:1.7.0_05]
Caused by: java.lang.NullPointerException
at org.jboss.as.osgi.service.FrameworkActivator.activate(FrameworkActivator.java:76)
at org.jboss.as.osgi.deployment.FrameworkActivateProcessor.deploy(FrameworkActivateProcessor.java:72)
at org.jboss.as.server.deployment.DeploymentUnitPhaseService.start(DeploymentUnitPhaseService.java:120) [jboss-as-server-7.2.0.Alpha1-redhat-4.jar:7.2.0.Alpha1-redhat-4]
{code}
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
13 years, 2 months