[jboss-dev] Re: [Fwd: Re: Beta3 status] - please reply

[Dimitris Andreadis]

Ok, so we'll wait for the following runs to complete, which means another 5 hours or so?

After that we should be ready to release. Rajesh, will you take it from there so the 
binaries are posted tomorrow?

Alexey, last chance for your "harmless" template check-ins while the suites are running. :-)

http://hudson.qa.jboss.com/hudson/view/JBoss%20AS/job/JBoss-AS-5.0.x-TestSuite-sun15/
http://hudson.qa.jboss.com/hudson/job/tck5-jaxws/
http://hudson.qa.jboss.com/hudson/job/tck5-ejb30/

I'm thinking of possibly moving the Beta3 tag to include the 2 commits:
r68452 - ales - Recursive managed deployment check.
r68462 - adrian - [JBAS-5095] - Port IMCP duplicate check from 4.2.

Issues, anyone?

Rajesh Rajasekaran wrote:
> Dimitris Andreadis wrote:
>> Forwarding to jboss-dev as it seems most people missed it.
>>
>> =========================================================
>>
>> I've applied the tag but there are some things I want to clarify 
>> before actually releasing:
>> https://svn.jboss.org/repos/jbossas/tags/JBoss_5_0_0_Beta3/
>>
>> - Ales & Pete (after a long long night) are done with the Seam tests. 
>> They report however
>> that Seam is dog slow on AS5.
>> - Alexey had some minor issues with the templates (done now?)
>> - The testsuite looks pretty good
>>
>> I've triggered the tck5 runs last night and they show regressions in 
>> the following:
>>
> 
> Sorry, missed the email as it landed in the wrong place in my inbox.
> 
>> http://hudson.qa.jboss.com/hudson/view/TCK5/job/tck5-ejb30/
> 
> The last run is consistent with the latest status on ejb30 (~ 326 
> failures) + some persistence runs which have been crashing on recent 
> ejb30 runs in hudson.
> 
>> http://hudson.qa.jboss.com/hudson/view/TCK5/job/tck5-saaj/
>> http://hudson.qa.jboss.com/hudson/view/TCK5/job/tck5-j2eetools/
> 
> These 2 modules are back to normal.
> 
>> http://hudson.qa.jboss.com/hudson/view/TCK5/job/tck5-jaxws/
>>
> 
> jaxws and jws modules had the reverse test classes built with our tools 
> removed from tck svn.
> Now i have made building those as part of the hudson run.
> the recent run of jws is back to 100%
> I have forced a run on jaxws now with with the reverse classes built and 
> they should be at 100%
> 
>> I don't really understand why those would come up and whether those 
>> results are reliable so
>> I triggered more runs to verify. Can the respective owners take a look 
>> (EJB3 and WS).
>>
> 
> Overall, TCK5 will be at 98.75% at the end of the ongoing ejb30 and 
> jaxws runs. I see no regressions.
> 
>> I've also asked Rajesh to setup a JBossWS-2.0.2.GA-testsuite-AS5.0.x 
>> build but we couldn't
>> get this to work, the testsuite wouldn't compile since some of the 
>> jars are moved around.
>>
> 
> The run i had set up was on the JBossWS 2.0.2.GA release tag where the 
> testsuite no longer compiles as Dimitris pointed out.
> The error can be seen here.
> http://hudson.qa.jboss.com/hudson/view/JBoss%20AS/job/JBossWS-2.0.2.GA-testsuite-AS5.0.x/1/console 
> 
> 
> The only way to verify this is with the latest testsuite from JBossWS 
> trunk against latest AS5. Thomas, do you already have these verified on 
> JBWS hudson?
> 
>> Can this be fixed or what is the proper runs to look for, on jbossws' 
>> own hudson server?
>>
>> http://jbws.dyndns.org:8180/hudson/view/AS%20Tests/
>>
>> /Dimitris
>>
>> Scott M Stark wrote:
>>> The run that picks up mc 2.0.0.beta9, the juddi rollback, etc is done
>>> and looks good. Time to tag.
>>> http://hudson.qa.jboss.com/hudson/job/JBoss-AS-5.0.x-TestSuite-sun15/256/ 
>>>
>>
>>
>