Bug 778330 (SOA-815) - JBM tests failing
Summary: JBM tests failing
Keywords:
Status: CLOSED NEXTRELEASE
Alias: SOA-815
Product: JBoss Enterprise SOA Platform 4
Classification: JBoss
Component: JBoss Messaging
Version: 4.3 IR4
Hardware: Unspecified
OS: Unspecified
urgent
urgent
Target Milestone: ---
: 4.3 IR5
Assignee: Julian Coleman
QA Contact:
URL: http://jira.jboss.org/jira/browse/SOA...
Whiteboard:
Depends On: SOA-747
Blocks:
TreeView+ depends on / blocked
 
Reported: 2008-09-22 12:28 UTC by Aleksandar Kostadinov
Modified: 2008-09-30 09:35 UTC (History)
0 users

Fixed In Version:
Doc Type: Bug Fix
Doc Text:
Clone Of:
Environment:
Last Closed: 2008-09-30 09:35:28 UTC
Type: Bug


Attachments (Terms of Use)


Links
System ID Private Priority Status Summary Last Updated
Red Hat Issue Tracker SOA-815 0 None None None Never

Description Aleksandar Kostadinov 2008-09-22 12:28:07 UTC
Date of First Response: 2008-09-22 10:34:20
project_key: SOA

For full details, please look at http://hudson.qa.jboss.com/hudson/view/SOA-Release/job/soa-jbm/lastBuild/#showFailuresLink

Generally most tests fail stating queues are not in state they need to be. Exceptions like:
junit.framework.AssertionFailedError: Message count for queue sourceQueue on server is 229
java.lang.IllegalStateException: Cannot remove references while deliveries are in progress, there are 1

But also:
java.lang.NoSuchMethodError: org.jboss.remoting.transport.socket.ServerAddress.<init>(Ljava/lang/String;IZII)V
	at org.jboss.test.thirdparty.remoting.ServerAddressTest.testEquals(ServerAddressTest.java:40)

These errors are not seen when tests run against EAP.

Comment 1 Aleksandar Kostadinov 2008-09-22 12:32:47 UTC
(03:30:48 PM) fox: akostadinov: if you're getting  java.lang.NoSuchMethodError: org.jboss.remoting.transport.socket.ServerAddress.<init>(Ljava/lang/String;IZII)V 
(03:30:48 PM) fox: 	at org.jboss.test.thirdparty.remoting.ServerAddressTest.testEquals(ServerAddressTest.java:40)
(03:31:05 PM) fox: akostadinov: that almost certainly means you're running using an incorrect version of jboss remoting
(03:31:16 PM) fox: akostadinov: which would explain your problem
(03:32:15 PM) fox: akostadinov: you've compiled against one version and you're running tests using another (probably older) version
(03:32:26 PM) akostadinov: fox: cool, that can help, thanks, I'll check

Comment 2 Aleksandar Kostadinov 2008-09-22 14:11:47 UTC
I can't find any differences in jboss-remoting.jar between EAP and SOA-P. Will investigate further tomorrow. 

Comment 3 Len DiMaggio 2008-09-22 14:34:20 UTC
From IRC:

<jcoleman> ldimaggi_mtg: If I compile a trivial test against the client jar and run it against the server - boom - same exception ;-(

<ldimaggi_mtg> jcoleman, Same exception as in SOA-815?

<jcoleman> ldimaggi_mtg: It looks like the class definition is different between client and server.



Comment 4 Len DiMaggio 2008-09-22 14:45:16 UTC
Changed severity to blocker- this problem be the root cause of the repeated failures being seen by the automated regression test suites.



Comment 5 Julian Coleman 2008-09-22 14:51:27 UTC
The version we have in jboss-as/client/jbossall-client.jar is the version that we get from
EAP FP01.

The versions in jboss-as/server/*/soap.esb/jbossall-client.jar are different versions (but all
the same different version).  They appear to have an older remoting version within them,
that has a different constructor for ServerAddress, which is what causes the runtime
exception.  There are also other differences.


Comment 6 Julian Coleman 2008-09-22 15:00:55 UTC
Link: Added: This issue depends SOA-747


Comment 7 Aleksandar Kostadinov 2008-09-23 04:26:57 UTC
Highly likely SOA-708 is related to this one.

Comment 8 Aleksandar Kostadinov 2008-09-23 04:26:57 UTC
Link: Added: This issue related SOA-708


Comment 9 Jiri Pechanec 2008-09-23 08:00:19 UTC
It seems that the versions of JBoss Remoting is 
1) JBM is compiled against 2.2.2SP8 
2) It is run against 2.2.2SP9
3) In the wrong JAR file there is 2.2.0SP4

Comment 10 Mark Little 2008-09-27 19:30:58 UTC
Jiri, is this all coming from EAP FP01 then, as J mentioned?

Comment 11 Aleksandar Kostadinov 2008-09-28 13:32:47 UTC
All failures except one are gone. Actually ther e are some on every build but these are timing related and are not consistent between runs. 

org.jboss.test.messaging.jms.ManifestTest.testManifestEntries
junit.framework.ComparisonFailure: expected:<JBoss Messaging> but was:<UNKNOWN>

It is failing consistently. It is *not* present in EAP builds. But we've seen it before as well SOA-547 JBMESSAGING-1287

Comment 12 Aleksandar Kostadinov 2008-09-28 13:34:12 UTC
All failures except one are gone in IR5. Actually there are some on every build but these are timing related and are not consistent between runs.

org.jboss.test.messaging.jms.ManifestTest.testManifestEntries
junit.framework.ComparisonFailure: expected:<JBoss Messaging> but was:<UNKNOWN>

It is failing consistently. It is *not* present in EAP builds. But we've seen it before as well SOA-547 JBMESSAGING-1287 so I don't think it is so critical.

Comment 13 Aleksandar Kostadinov 2008-09-28 13:34:59 UTC
Link: Added: This issue depends JBMESSAGING-1287


Comment 14 Aleksandar Kostadinov 2008-09-28 13:35:33 UTC
Link: Removed: This issue related SOA-708 


Comment 15 Mark Little 2008-09-28 13:41:23 UTC
"3) In the wrong JAR file there is 2.2.0SP4"

So do we know where this version comes (came) from?

Comment 16 Jiri Pechanec 2008-09-29 06:53:10 UTC
The JAR with wrong version is no longer present in IR5, so it was probably removed as part of SOA-747


Note You need to log in before you can comment on or make changes to this bug.