[00:02:55] *** ldimaggi has quit IRC [00:07:07] *** jose_freitas has quit IRC [00:26:46] *** mhuniewicz has quit IRC [01:19:34] *** ldimaggi has joined #jbosstesting [02:22:17] *** jose_freitas has joined #jbosstesting [02:51:16] *** johnament has joined #jbosstesting [02:52:10] *** ianbrandt has quit IRC [02:59:18] *** ldimaggi has quit IRC [02:59:19] *** OndraZizka has quit IRC [02:59:46] *** ldimaggi has joined #jbosstesting [03:00:08] *** OndraZizka has joined #jbosstesting [03:34:39] *** johnament has quit IRC [04:00:16] *** jose_freitas has quit IRC [04:22:24] *** ldimaggi has quit IRC [04:22:45] *** ldimaggi has joined #jbosstesting [05:10:26] *** bobmcw has quit IRC [05:12:09] *** bobmcw has joined #jbosstesting [05:15:26] *** ldimaggi has quit IRC [06:23:40] *** ALR has joined #jbosstesting [06:28:30] *** pilhuhn has joined #jbosstesting [08:21:26] *** rruss has quit IRC [08:26:46] *** mgoldmann has joined #jbosstesting [08:38:24] *** milestone has joined #jbosstesting [09:01:49] *** jhuska has joined #jbosstesting [09:16:18] *** aslak has joined #jbosstesting [09:21:36] *** tdiesler has joined #jbosstesting [09:26:07] *** maeste has joined #jbosstesting [09:47:58] *** maeste has quit IRC [10:00:24] *** Jaikiran has joined #jbosstesting [10:04:55] *** maeste has joined #jbosstesting [10:18:55] *** aslak has quit IRC [10:57:37] *** maschmid has joined #jbosstesting [11:03:03] *** mgoldmann has quit IRC [11:03:39] *** aslak has joined #jbosstesting [11:04:03] *** vtunka has joined #jbosstesting [11:21:09] *** milestone has quit IRC [11:39:20] *** milestone has joined #jbosstesting [11:57:02] *** OndraZizka has quit IRC [11:57:17] *** OndraZizka has joined #jbosstesting [12:10:29] *** mgoldmann has joined #jbosstesting [12:33:41] *** tdiesler is now known as tdiesler_afk [12:33:58] *** jose_freitas has joined #jbosstesting [12:37:25] *** ldimaggi has joined #jbosstesting [12:59:59] *** jose_freitas has quit IRC [13:02:35] <aslak> ALR, https://github.com/aslakknutsen/jboss-as/commits/ARQ_CONFIG [13:02:37] <aslak> :) [13:31:52] *** jose_freitas has joined #jbosstesting [13:47:43] *** milestone has quit IRC [14:03:18] <jose_freitas> aslak: ping/good morning [14:03:37] <aslak> jose_freitas, good morning [14:05:37] <jose_freitas> do you know what could be causing this error? Caused by: com.sun.jersey.api.client.UniformInterfaceException: PUT http://null:null@localhost:8080/manager/deploy?path=/71f6f830-196d-41eb-88a8-432891f1d9fb returned a response status of 401 [14:08:00] *** milestone has joined #jbosstesting [14:11:34] <jose_freitas> aslak ^ [14:12:07] <aslak> jose_freitas, which container ? glassfish ? [14:13:00] <aslak> glassfish remote [14:13:21] <jose_freitas> tomcat remote [14:15:15] *** ldimaggi has quit IRC [14:15:23] <aslak> jose_freitas, a obvious bad handling of arq.xml user/pass it seems [14:16:08] <aslak> jose_freitas, set those options in your arq.xml and see if that fixes it [14:16:26] <aslak> i'm pretty sure the null:null auth is the issue [14:17:26] <jose_freitas> so it could be that my jmx options were ignored? [14:17:38] <aslak> jose_freitas, ,no both are required [14:17:50] <aslak> jose_freitas, http via manager app to deploy, and jmx to introspect the deployment [14:18:00] <jose_freitas> ok [14:18:38] <aslak> jose_freitas, you probably need to activate some users in the tomcat.users / roles files under catalina_home somewhere [14:19:51] <aslak> jose_freitas, if that fixes it, please file a jira on the tomcat containers, not handling user/pass properly [14:20:56] <jose_freitas> ok [14:24:40] <jose_freitas> I think you're right, but do I have to add an specific role to the user? [14:24:47] <jose_freitas> it's now returning 403 [14:24:57] <aslak> i think it's manager [14:25:01] <jose_freitas> which is access denied [14:25:30] <aslak> i have this in mine [14:25:32] <aslak> <role rolename="manager" /> [14:25:32] <aslak> <user username="admin" password="admin" roles="manager" /> [14:25:46] <jose_freitas> thanks [14:31:36] <jose_freitas> nice [14:31:38] <jose_freitas> it's working [14:34:27] <jose_freitas> I'll fire the jira today [14:34:28] <aslak> :) [14:34:35] <aslak> great, thanks [14:41:05] *** alesj has joined #jbosstesting [14:50:46] *** rruss has joined #jbosstesting [14:50:50] *** rruss has joined #jbosstesting [14:53:22] *** tdiesler_afk is now known as tdiesler [14:58:10] *** tdiesler has quit IRC [15:03:55] <jose_freitas> aslak: I suspect that the way we're setting a system.property with maven don't reach remote containers (because they're on different vm). could that be? [15:04:34] <aslak> jose_freitas, sounds about right [15:26:17] *** ldimaggi has joined #jbosstesting [16:18:33] <ALR> aslak: Know what I'm missing here? [16:18:34] <ALR> arquillianBeforeClass(org.jboss.as.arquillian.testng.BasicTestNGIntegrationTestCase) Time elapsed: 0 sec <<< FAILURE! [16:18:34] <ALR> java.lang.NoClassDefFoundError: com/google/inject/Inject [16:18:34] <ALR> at org.jboss.arquillian.testng.container.TestNGDeploymentAppender.createAuxiliaryArchive(TestNGDeploymentAppender.java:55) [16:25:37] <ALR> aslak: For instance, why is TestNGDeploymentAppender doing this? https://gist.github.com/1149193 [16:26:17] <ALR> And also why is Guice not brought in my these dependencies if it's required? https://gist.github.com/1149194 [16:39:32] <jose_freitas> aslak: with servlet 2.5 we must add a filter manually on the main web.xml right? adding a web-fragment to a jar won't do it. Am I mistaken? [16:43:23] *** bgeorges has joined #jbosstesting [17:02:26] *** maschmid has quit IRC [17:10:04] <aslak> ALR, TestNG require it, but they don't depend on it with TestNG > 6 or so [17:10:32] <ALR> aslak: So you think that ARQ should not export the dependency? (I do) [17:10:34] <aslak> ALR, they used to bundle it, but due to ppl wanting to use other versions and what not Cedric removed it. [17:10:41] <ALR> Because for runtime it needs to be declared. [17:13:41] <aslak> jose_freitas, no, we add it dynamically for 2.5 as well.. [17:14:32] <aslak> jose_freitas, JSFUnitProtocolArchiveProcessor [17:15:30] <aslak> jose_freitas, ALR hmmm have a issue there tho, since WebAppDescriptor is not stable yet.. [17:16:39] <aslak> ALR, i'm a bit uncertain atm if the Guice dep from TestNG side was only there if you use the Guice TestNG extension, or is TestNG need guice to run [17:17:31] <ALR> aslak: OKee. [17:17:33] <aslak> is/if [17:17:40] <ALR> For the time being I'm manually adding the dependency [17:17:50] <ALR> But I think that's something we should know. [17:23:48] *** jhuska has quit IRC [17:30:06] <ALR> aslak: I fixed AS7-1303 [17:30:07] <jbossbot> jira [AS7-1303] TestNG not supported by Arquillian Service [Open (Unresolved) Bug, Critical, Andrew Rubinger] https://issues.jboss.org/browse/AS7-1303 [17:30:44] <aslak> ALR, great! not adding a check for "extends Arquillian" i hope.. :) [17:31:23] <ALR> Um, actually yes [17:31:26] <ALR> Why not? [17:31:43] *** bgeorges has quit IRC [17:32:24] <aslak> ALR, and as far as i can understand cedric in a testng-users post here, testng does not use guice internally any more, but has/had a dep on it for a @Guice annotation that required a member with a Guice Module class. so we should be able to remove that part now that ShrinkWrap no longer require us to load all found classes etc as well.. [17:32:29] <ALR> aslak: https://github.com/ALRubinger/jboss-as/commit/f1a8384b337faae3beb58ff74d93f931b222bfa9 [17:32:30] <jbossbot> git [jboss-as] f1a8384.. Andrew Lee Rubinger [AS7-1303] Add failing case showing TestNG not working in ARQ AS7; Make ArquillianConfigBuilder support both JUnit and TestNG [17:32:31] <jbossbot> jira [AS7-1303] TestNG not supported by Arquillian Service [Open (Unresolved) Bug, Critical, Andrew Rubinger] https://issues.jboss.org/browse/AS7-1303 [17:33:19] <ALR> aslak: Specifically: https://github.com/ALRubinger/jboss-as/commit/f1a8384b337faae3beb58ff74d93f931b222bfa9#L2R77 [17:33:20] <jbossbot> git [jboss-as] f1a8384.. Andrew Lee Rubinger [AS7-1303] Add failing case showing TestNG not working in ARQ AS7; Make ArquillianConfigBuilder support both JUnit and TestNG [17:33:20] <jbossbot> jira [AS7-1303] TestNG not supported by Arquillian Service [Open (Unresolved) Bug, Critical, Andrew Rubinger] https://issues.jboss.org/browse/AS7-1303 [17:33:36] <ALR> aslak: Please approve that before I do AS7 pull request [17:34:26] <aslak> ALR, you do know you just hardcoded the service to another testframework right ? [17:34:40] <ALR> aslak: No. [17:34:56] <ALR> I added another test framework to it, though. [17:35:07] <ALR> Look at the patch [17:35:10] <aslak> ALR, want me to add some new issues, Arq Service does not work with ScalaTest, Spock [17:35:21] <ALR> Ah, we have others? [17:35:35] <ALR> Interesting, I thought those ran off JUnit or TestNG. [17:35:40] <aslak> in the works / arq is meant to work with yea [17:36:09] <ALR> Any ideas for a solution here then? How is the builder to know what the test classes are? [17:36:18] <aslak> well.. hmm.. they kinda do.. not sure your tests there will match tho, but [17:36:56] <ALR> I'm open to suggestions. [17:37:07] <ALR> But as this is written, it needs to inspect the deployment to find test classes [17:37:24] <aslak> yea.. why? [17:37:31] <aslak> to fetch the Module cl ? [17:37:45] <ALR> I've no ide. [17:37:47] <ALR> *idea. [17:38:01] <ALR> I just made it work, didn't do a design overview on the thing really. Let me peek deeper. [17:38:30] <ALR> Though honestly for 7.0.1 I'd rather just push this change [17:38:33] <ALR> Get TestNG working [17:38:44] <ALR> Then do more extensive refactoring if we need to in the next AS7 releas. [17:38:47] <ALR> *release [17:38:51] <ALR> Because tag is coming today [17:38:57] <ALR> WDYT? [17:39:03] <aslak> ALR, it's about time some one get a overview.. ;) [17:39:12] <ALR> Yep, I agree. [17:39:13] <aslak> ALR, as a quick fix sure [17:39:16] <ALR> Cool. [17:39:29] <aslak> you need me to push CR4 tho ? [17:39:35] <ALR> It's no hackier than it was, I suppose :) [17:39:47] <ALR> I need you to push CR4? Why? [17:39:51] <aslak> well, twice as hacky isn't it ? hehe [17:40:03] <aslak> i mean the Guice stuff [17:40:41] *** milestone has quit IRC [17:42:26] <ALR> Nope. [17:42:33] <ALR> Because I took out the direct class references. [17:42:42] <ALR> So this needs redesign. [17:42:48] <ALR> I'll put in this fix now though [17:42:49] <aslak> ? [17:42:59] <ALR> aslak: No more "import org.junit..." [17:43:33] <ALR> Thanks for raising that point though. [17:43:35] <aslak> but what does that have to do with the TestNG ArchiveAppender ? [17:43:43] <ALR> Nothing. [17:43:53] <ALR> "Twice as hacky?" Nope. [17:44:11] <ALR> I'm not understanding what you're asking about CR4 though? [17:44:16] <aslak> ALR, before you checked for one, now you check for two.. twice as hacky. hehe [17:44:56] <ALR> Yeah, um *I* didn't check for anything. [17:45:13] <ALR> Not to mention passing around jandex types for some reason. [17:45:26] <aslak> ALR, well, if you want the update of the TestNG ArchiveAppender to not force guice on cp, we need a new core release.. (CR4) [17:45:51] <ALR> aslak: Ah yes. [17:45:59] <ALR> THhat'd be nice, yes. [17:46:05] <aslak> right [17:46:07] <ALR> The choice for release criteria of ARQ is to you though :) [17:46:07] <aslak> 2 sec [17:46:42] <aslak> CR3 was based on another fix for the as7 release, i'm sure i can push a CR4 for it as well.. :) [17:54:29] <jbossbot> git [arquillian-core] push master b73cfcc.. Aslak Knutsen ARQ-549 Remove direct dependency on Guice, TestNG no longer use it internally. Only add it if found on ClassPath, we assume it is related to TestNG. [17:54:30] <jbossbot> jira [ARQ-549] Guice is no longer a required dependency of TestNG, Arquillian TestNG Archive Appender should not force it [Open (Unresolved) Bug, Major, Aslak Knutsen] https://issues.jboss.org/browse/ARQ-549 [17:54:30] <jbossbot> git [arquillian-core] push master URL: http://github.com/arquillian/arquillian-core/compare/37b0883...b73cfcc [17:54:42] <aslak> ALR, you do the v update ? [17:55:00] <ALR> aslak: What version update? [17:55:10] <aslak> ALR, as7 arq v [17:55:23] <ALR> I hadn't, no. I will. [17:55:30] *** ldimaggi has quit IRC [17:55:35] <ALR> What should it be? [17:55:38] <aslak> let me push it first.. :) [17:55:48] <aslak> arquillian_core 1.0.0.CR4 [17:56:09] <ALR> Yeah OK [17:56:13] <ALR> I have it ready right here. [17:56:40] <ALR> So then maybe I can remove the dep on Google Guice [17:56:44] <ALR> And re-add the exclusion [17:56:57] <aslak> yea [17:57:20] <aslak> this is the fix: https://github.com/arquillian/arquillian-core/compare/37b0883...b73cfcc [17:57:37] <aslak> it's added if there [17:57:45] <aslak> but no direct class dep [17:58:49] <ALR> aslak: 2 issues: [17:58:51] <ALR> AS7-1549 [17:58:53] <jbossbot> jira [AS7-1549] Upgrade to Arquillian 1.0.0.CR4 [Open (Unresolved) Component Upgrade, Major, Andrew Rubinger] https://issues.jboss.org/browse/AS7-1549 [17:58:55] <ALR> AS7-1550 [17:58:56] <jbossbot> jira [AS7-1550] Remove Dependency upon Guice in arquillian/testng-integration [Open (Unresolved) Task, Major, Andrew Rubinger] https://issues.jboss.org/browse/AS7-1550 [17:59:08] <aslak> ok [17:59:29] <ALR> aslak: You feel certain the upgrade will work? [17:59:33] <aslak> the reason for 1550 is already pushed? [17:59:44] <ALR> Not upstream [17:59:48] <ALR> The reason for 1550 is 1303 [18:00:27] <aslak> i would have changed 1303 to not expose them to begin with instead.. but same same.. :) [18:01:23] <aslak> ALR, the upgrade should work.. if the testng stuff does is another case [18:01:33] <ALR> The testng stuff works. [18:01:40] <ALR> I just want it to work w/out a Guice dep [18:01:41] <ALR> :) [18:01:42] <aslak> i mean testng without guice on cp [18:02:00] <ALR> Because I feel bad about removing the exclusion from the AS7 root POM [18:02:18] <ALR> Then again, they only put that exclusion in the enforcer because it's not needed, not because it's problematic. [18:02:26] <aslak> hehe [18:14:54] <aslak> ALR, give it a spin, https://repository.jboss.org/nexus/content/repositories/jboss_releases_staging_profile-106/ [18:15:07] <ALR> THx [18:19:13] <ALR> aslak: Thanks a blank link for me [18:20:08] <aslak> ALR, jboss nexus has turned of directory listing [18:20:10] <ALR> s/Thanks/that's [18:20:18] <ALR> Ah, that's what that is. [18:20:27] <ALR> Yeah, I've been complaining to them about that for a few weeks now [18:20:47] <ALR> So how do I use your staging profile? [18:20:53] <ALR> To build? [18:21:56] <aslak> add a reposiotry in your settings or similar [18:22:04] <aslak> a / the staging [18:24:13] <aslak> ALR, brb [18:24:35] <ALR> Oh right [18:24:36] <ALR> Sure [18:26:51] *** lightguard_jp has joined #jbosstesting [18:27:43] *** maeste has quit IRC [18:33:14] *** pilhuhn is now known as pil-dinner-bbl [18:42:36] *** mgoldmann has quit IRC [18:45:24] *** ianbrandt has joined #jbosstesting [18:48:39] *** alesj has quit IRC [19:03:21] <aslak> ALR, how is it looking? [19:04:13] <ALR> aslak: Building [19:04:29] <ALR> Did the upgrade in one commit, now the 2nd build for removing the dep is running [19:05:36] <aslak> mm [19:13:52] *** Jaikiran has quit IRC [19:21:35] <aslak> ALR, i see a pull request.. meaning all ok? [19:22:41] <ALR> Yep [19:22:47] <aslak> i'll push then [19:22:53] <ALR> https://github.com/jbossas/jboss-as/pull/176 [19:23:04] <ALR> Oh yeah, release that Nexus thing [19:23:05] <ALR> Sorry [19:23:20] <ALR> You need much time, because I already triggered the pull request w/ Jas [19:23:47] <aslak> done [19:24:16] <ALR> Thanks bud [19:24:23] <ALR> Your fix works great [19:24:58] <aslak> then we're good to go for testng cdi tck [19:25:02] <aslak> cool [19:26:54] <ALR> Yup. [19:40:10] <jose_freitas> Hey aslak, I'm back. I think that web-fragments are not accepted with servlet 2.5, are you sure of that? [19:42:30] *** mhuniewicz has joined #jbosstesting [19:42:37] <mhuniewicz> ALR, how's it going? [19:44:33] <ALR> mhuniewicz: Good. [19:45:32] <mhuniewicz> ALR, anna see that stacktrace? [19:45:37] <mhuniewicz> *wanna [19:45:40] <ALR> mhuniewicz: OK [19:45:57] <mhuniewicz> http://pastebin.com/SNkcAsKj [19:46:49] <ALR> I thought this was an EJB [19:46:55] <ALR> Why is the trace coming from org.jboss.as.weld.ejb.SessionObjectReferenceImpl.getBusinessObject(SessionObjectReferenceImpl.java:73) ? [19:47:12] <ALR> mhuniewicz: Recommend raising the question on the AS7 User Forum [19:48:10] <mhuniewicz> ALR, I see. [19:48:18] <mhuniewicz> Also, I got something like that: JBREM00200: Remote connection failed: java.io.IOException: Window open for non-existent channel [19:48:29] <mhuniewicz> And it didn't cause the test to fail, [19:48:43] <mhuniewicz> but the test didn't give any server output. I googled but found no answers. [19:48:44] <aslak> jose_freitas, that is correct, but when there is a web.xml in version 2.5 the jsfunitarchiveprocessor will add the filter [19:50:17] <jose_freitas> it should do that? [19:50:23] <jose_freitas> I think it's not doing [19:50:26] <jose_freitas> lemme check [19:50:54] <ALR> mhuniewicz: Same thing I always tell you. :) [19:50:58] <ALR> 1) Context is important. [19:51:06] <ALR> Both knowing the context of the error. [19:51:12] <aslak> jose_freitas, sorry, JSFUnitProtocolArchiveProcessor [19:51:22] <ALR> And raising the issue in the right forum so that people in the know can answer. [19:51:43] <mhuniewicz> ALR, it was raised on the Arquillian forum but no one knew the answer. [19:51:51] <ALR> Because it's not an Arquillian error. [19:52:06] <jose_freitas> aslak: yeah, just checked, it's doing that, but it's not removing, so that bug/feature prevents overwritting the file :) [19:52:30] <mhuniewicz> On context, it's an integration test, no more errors or anything. [19:52:35] <aslak> jose_freitas, aa of course.. :) [19:52:37] <jose_freitas> thanks [19:52:41] <mhuniewicz> I was hoping you saw that somewhere. [19:52:45] <mhuniewicz> I will look elsewhere then. :) [19:52:52] <ALR> mhuniewicz: Hehe, that's not enough context ("it's an integration test") [19:52:59] <ALR> mhuniewicz: Trying to help you here. [19:53:09] <ALR> I think first goal is to determine the subsystem throwing the error [19:53:26] <ALR> Then on that subsystem's forum, describe fully what you're doing and supply a tract [19:53:28] <ALR> *trace [19:53:35] <ALR> Generally speaking that should help out a lot. [19:53:39] <ALR> Both for us and for you. [19:54:21] <ALR> So for instance an Arquillian error would be something where you can't deploy, or the container doesn't start from an ARQ-based test [19:54:31] <ALR> But once you enter the container runtime, it's out of ARQ's hands [19:54:59] <mhuniewicz> I'm not sure it is, as there is no server stuff logged. But then the test seems to pass (but once it failed because of that). [19:54:59] *** dmlloyd is now known as dmvoid [19:55:21] *** dmvoid is now known as dmlloyd [20:19:15] *** jose_freitas has quit IRC [20:24:17] *** jose_freitas has joined #jbosstesting [20:30:14] *** mhuniewicz has quit IRC [20:32:51] *** ldimaggi has joined #jbosstesting [20:33:14] *** jose_freitas has quit IRC [20:34:43] *** pil-dinner-bbl is now known as pilhuhn [20:36:19] *** toepi has joined #jbosstesting [20:39:39] *** jose_freitas has joined #jbosstesting [20:46:19] *** jose_freitas has quit IRC [21:02:11] *** jose_freitas has joined #jbosstesting [21:21:45] *** ldimaggi has quit IRC [21:29:37] *** alesj has joined #jbosstesting [21:35:37] *** toepi has quit IRC [21:37:55] <jose_freitas> aslak: what could provoke a remoteLoadableExtension to not load? [21:41:30] *** ALR has left #jbosstesting [21:49:10] <aslak> jose_freitas, container ? [21:51:09] <jose_freitas> tomcat remote [21:51:13] <aslak> jose_freitas, missing SPI file is one [21:51:19] <jose_freitas> it's there [21:51:34] <aslak> wrong classloader [21:52:14] <jose_freitas> yeah, I was thinking about that, mixing maven build and eclipse build (junit), it might have screw it up a little bit [21:52:26] <aslak> check LoadableExtensionLoader, in container it should load a 'incontainer' v of ExtensionLoader [21:52:33] *** ALR has joined #jbosstesting [21:54:09] *** ALR has left #jbosstesting [21:56:23] *** ldimaggi has joined #jbosstesting [22:04:02] <jose_freitas> LoadableExtensionLoader should be called once in client and once in server, right? it seems that it has been called just once, in client. [22:05:28] <jose_freitas> I saw that "load" method is an observer, who is firing the event? [22:05:32] <jose_freitas> ManagerProcessing event [22:05:41] <jose_freitas> aslak ^ [22:08:03] <aslak> that's fired by Manager, during startup. so when the Servlet request is coming in [22:08:10] <aslak> via TestableAdaptorBuilder [22:08:37] <jose_freitas> so, this load should be called twice, right? [22:08:43] <aslak> sorry, TestRunnerAdaptorBuilder [22:09:11] <aslak> jose_freitas, right, but that happens in contianer. in the remote vm etc [22:09:36] <jose_freitas> let me debug with jboss to check where it should pass [22:12:44] *** ALR has joined #jbosstesting [22:17:41] <jose_freitas> damn, jboss stop working too. [22:28:00] <jose_freitas> uff, rebuilding it make it work, it's definetely something with classloader [22:31:24] <aslak> hmm [22:35:58] *** pilhuhn has quit IRC [22:59:55] *** bleathem has quit IRC [23:11:26] *** rruss has quit IRC [23:26:05] *** rruss has joined #jbosstesting [23:28:21] *** ldimaggi has quit IRC