May 17, 2011 
< | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | >
May 17, 2011 NOTICE: This channel is no longer actively logged.
[00:05:49] *** epbernard has joined #jboss-as7[00:05:50] *** epbernard is now known as emmanuel[00:05:50] *** ChanServ sets mode: +v emmanuel[00:13:38] *** misty has quit IRC[00:14:46] *** misty has joined #jboss-as7[00:14:47] *** bstansberry is now known as bstans_afk[00:15:16] <dmlloyd> stuartdouglas: so what do you think about merging ejb[00:15:48] <stuartdouglas> I think we should do it[00:16:08] <dmlloyd> any outstanding issues or patches from your end?[00:16:32] <dmlloyd> I think I'll tag modules CR2 beforehand but other than that I'm ready...[00:16:48] <stuartdouglas> not really, I have a half done JSF injection one, and a little bit of CDI stuff, but nothing that is ready to merge[00:17:08] <stuartdouglas> the JSF managed bean injection is not very nice, cause of the way their SPI works[00:17:13] <dmlloyd> okay then, I'll get working on this[00:17:17] <dmlloyd> yeah the SPI is dumb[00:17:35] <dmlloyd> pretty much forces a global registry of classes[00:18:15] <stuartdouglas> yea, and because they create the actual instance, I can't just use Component.createInstance() and have it just work[00:19:29] *** pferraro has joined #jboss-as7[00:19:29] *** ChanServ sets mode: +v pferraro[00:21:09] *** sannegrinovero has quit IRC[00:24:26] <dmlloyd> yeah[00:24:39] <dmlloyd> not sure why they need to construct the instances[00:24:58] <stuartdouglas> they don't, it's just the way they have designed the SPI[00:28:01] *** hbraun has joined #jboss-as7[00:33:32] *** hbraun has quit IRC[00:39:12] *** tcrawley_away has joined #jboss-as7[00:40:11] *** smcgowan has joined #jboss-as7[00:40:11] *** ChanServ sets mode: +v smcgowan[00:41:32] *** smarlow has joined #jboss-as7[00:41:32] *** ChanServ sets mode: +v smarlow[00:43:58] <dmlloyd> okay folks, I'm just about to merge the EJB branch![00:44:19] <dmlloyd> well, once I rebase and do a little more testing.[00:44:27] <dmlloyd> get ready for meyhem![00:44:36] <dmlloyd> or may ham, as appropriate[00:44:42] *** mbg is now known as mbg|away[00:46:11] <dmlloyd> anyone know what this might be all about: "java.lang.ClassNotFoundException: javax.management.ObjectName from [Module "org.jboss.as.standalone:main" from local module loader @2d20cc56 (roots: /home/david/src/java/as7/jboss-as/arquillian2/container-managed/target/jbossas/modules)] (no security manager: RMI class loader disabled)"[00:47:01] <dmlloyd> http://fpaste.org/9MOV/ <- the whole thing[00:48:01] <stuartdouglas> looks like a ClassNotFoundException due to a missing dep?[00:48:07] <smcgowan> dmlloyd: did sguilhen's jacorb/rmi iiop update get pushed[00:48:30] <dmlloyd> stuartdouglas: yeah I guess I"ll add javax.api to as.standalone, just kinda weird it's only happening now...[00:48:35] <dmlloyd> smcgowan: I'm not sure[00:48:50] <dmlloyd> Nihility: did you merge the jacorb patch?[00:49:42] <dmlloyd> stuartdouglas: weird thing is, as.standalone is an empty module, just re-exports as.server[00:49:49] <dmlloyd> oh, shit, I know the prob[00:49:50] <dmlloyd> TCCL[00:50:10] <dmlloyd> so, something behaves differently if we have a TCCL set[00:50:12] <dmlloyd> on boot I mean[00:50:18] <dmlloyd> that's not a good thing :)[00:51:14] <smcgowan> stuartdouglas: you see this: http://hudson.qa.jboss.com/hudson/view/TCK6-AS7-WEB/job/tck6-as7-cdi/lastCompletedBuild/testReport/ ?[00:52:06] <stuartdouglas> smcgowan: not yet, I will look into it[00:52:37] <stuartdouglas> locally on the ejb branch I am down to 10 fails (probably 9 now)[00:53:21] <smcgowan> stuartdouglas: we can wait for that, just a bit surprised - i thought you were down around there so didn't expect those results[00:55:18] <smcgowan> they're deployment failures[00:56:18] *** fnasser has quit IRC[01:02:00] <dmlloyd> hmm okay weird, I added javax.api to the problem module and it's still blowing up[01:02:08] <dmlloyd> the arq tests must be doing something wacky[01:02:09] <dmlloyd> AGAIN[01:03:07] <smcgowan> stuartdouglas: speaking of wacky, the jaxrs tests are all failing too, this is with jbossas master, let me dig into this[01:04:03] *** fnasser has joined #jboss-as7[01:04:08] <stuartdouglas> dmlloyd: is this the embedded tests?[01:04:42] <dmlloyd> Arquillian Container Managed (arq2)[01:05:03] <dmlloyd> arquillian2/container-managed in other words[01:05:24] <stuartdouglas> that should be the same as just starting the server normally[01:06:36] *** fnasser has quit IRC[01:07:25] <dmlloyd> hm had a bunch of zombie ASes from the failed test runs[01:07:30] <dmlloyd> might be that I fixed it after all[01:07:33] * dmlloyd tries again[01:10:39] <smcgowan> stuartdouglas: the jaxrs tests are OK locally[01:11:48] <stuartdouglas> damn[01:12:11] <stuartdouglas> Is it possible to get the server output for those runs? Rather than just the test console output?[01:13:54] <dmlloyd> http://github.com/dmlloyd/jboss-as/commit/2731caa[01:14:05] <dmlloyd> this is where I'm at, have one more test failure which I'll check after dinner[01:14:31] <dmlloyd> hm, crap, non-ff[01:14:33] <dmlloyd> what did I do[01:14:39] <dmlloyd> oh well, will fix it later[01:19:45] <smcgowan> stuartdouglas: was that question for me, RE: server output for test runs?[01:19:53] <stuartdouglas> yes[01:19:58] <smcgowan> for CDI?[01:20:29] <stuartdouglas> yes, as usually it is much more useful in telling you what went wrong[01:20:44] <smcgowan> http://hudson.qa.jboss.com/hudson/view/TCK6-AS7-WEB/job/tck6-as7-cdi/ws/jbossas7/standalone/log/[01:21:16] <smcgowan> i'll archive them so they are readily available[01:25:43] *** alexsmirnov has quit IRC[01:29:31] *** pferraro has quit IRC[01:33:03] <stuartdouglas> smcgowan: I can reproduce the weld failures locally, I should have a fix fairly soon[01:33:38] <dmlloyd> http://github.com/dmlloyd/jboss-as/commit/9d10281[01:33:38] <jbossbot> git [jboss-as] 9d10281.. David M. Lloyd Update to Modules 1.0.0.CR2[01:33:40] <dmlloyd> that's the real one[01:35:43] <smcgowan> stuartdouglas: ok, the JAXRS errors are due to the server not starting due to a parse error in the standalone configuration, they are passing for me locally, may have to sync with some updates[01:36:00] *** pferraro has joined #jboss-as7[01:36:00] *** ChanServ sets mode: +v pferraro[01:36:12] <stuartdouglas> ok, the weld problem is due to the content parameter of the add operation being null[01:36:15] <smcgowan> specifically: Message: Unexpected attribute 'verify-client' encountered[01:36:26] <stuartdouglas> but there is also another issue where this message is swallowed by the lgs[01:36:47] <stuartdouglas> logs even[01:38:29] *** pferraro has left #jboss-as7[01:39:40] <dmlloyd> testing rebase[01:43:59] <stuartdouglas> dmlloyd: jboss-logmanager seems to be loosing the Throwables message in some cases[01:44:02] <stuartdouglas> e.g. http://fpaste.org/nNZo/[01:44:44] <stuartdouglas> The OperationFailedException's message is ("paramerter content must not be null") is not displayed[01:46:06] <stuartdouglas> actually I am wrong[01:46:24] <stuartdouglas> it is just a bad toString on the OperationFailedException[02:00:02] *** dmison has joined #jboss-as7[02:01:08] *** aslak has quit IRC[02:02:02] <jamezp> dmlloyd: I think I have the txn converted https://github.com/jamezp/jboss-as/commit/504c91c24b0c8a2844216ba1498281ba45bcc333[02:02:03] <jbossbot> git [jboss-as] 504c91c.. James Perkins Use logging tool to create messages and loggers. Convert throwables and logs to use the new interfaces.[02:03:31] <dmlloyd> cool, I"ll check it out shortly.[02:04:11] <jamezp> No rush, a bunch of import changes because IDEA keeps reorganizing them. Looking now to see if I can stop that from happening.[02:05:39] <dmlloyd> [INFO] BUILD SUCCESS[02:05:39] <dmlloyd> [INFO] Total time: 5:35.684s[02:05:44] <dmlloyd> somehow I don't feel like a winner[02:06:05] <dmlloyd> jamezp: yeah, that's a good idea. In particular, * static imports are fine (however * regular imports are not)[02:06:20] <dmlloyd> okay I got a clean build/run with ejb rebased on upstream.[02:06:37] <dmlloyd> if nobody objects really really soon I'm pushing all 9,000,000 commits of it upstream[02:06:57] <jamezp> dmlloyd: It keeps annoying me because it makes it hard to see the changes.[02:07:03] <jamezp> I've no objections.[02:07:10] * stuartdouglas thinks jbossbot is about to be kicked for flooding[02:07:18] <jamezp> Haha[02:07:52] <dmlloyd> okay 114 changes[02:08:01] <dmlloyd> all right folks here it comes[02:08:18] <jbossbot> git [jboss-as] push master e90dbdc.. David M. Lloyd [JBAS-9223] Framework for new view description classes, binding, injection, and interceptors[02:08:19] <jbossbot> jira [JBAS-9223] Redirected to: https://issues.jboss.org/si/jira.issueviews:issue-xml/AS7-387/AS7-387.xml[02:08:20] <jbossbot> jira [AS7-387] Separate view description and config from component description and config [Open (Unresolved) Task, Blocker, David Lloyd] https://issues.jboss.org/browse/AS7-387[02:08:20] <jbossbot> git [jboss-as] push master 5e250bf.. John E. Bailey Update annotation based injection processors to follow the new injection impl.[02:08:20] <jbossbot> git [jboss-as] push master eb72dbf.. John E. Bailey Update life-cycle annotation processing to use class based descriptions[02:08:20] <jbossbot> git [jboss-as] push master 7802df1.. jaikiran Port EnvEntryProcessor to the new EE framework[02:08:21] <jbossbot> git [jboss-as] push master 73eac23.. John E. Bailey Update interceptor annotation processing to use new class based framework[02:08:21] <jbossbot> git [jboss-as] push master 8a1b8c3.. jaikiran Migrate more resource reference processors to new EE framework[02:08:21] <jbossbot> git [jboss-as] push master 9e828e3.. David M. Lloyd Use name resolution context instead of component config[02:08:22] <jbossbot> git [jboss-as] push master c364da1.. John E. Bailey Add processors to fire configurators for classes and components.[02:08:22] <jbossbot> git [jboss-as] push master 73abc5d.. jaikiran Fix compilation error[02:08:23] <jbossbot> git [jboss-as] push master 1606b23.. jaikiran Fix compilation errors[02:08:23] <jbossbot> git [jboss-as] push master 70e1e4f.. jaikiran Make checkstyle happy[02:08:24] <jbossbot> git [jboss-as] push master 706b796.. jaikiran Introduce a way to add and process module level jndi bindings[02:08:24] <jbossbot> git [jboss-as] push master 6ac3ba3.. David M. Lloyd Additional support for jboss namespace[02:08:25] <jbossbot> git [jboss-as] push master 8f0a89a.. jaikiran Migrate the DataSourceDefinitionDeployer to the new EE framework[02:08:25] <jbossbot> git [jboss-as] push master 7a01d5b.. jaikiran Port the JPA annotation processor to the new EE framework[02:08:25] <jbossbot> git [jboss-as] push master 77a442d.. jaikiran Initial changes to the EJB3 component to move to the new EE framework (WIP)[02:08:26] <jbossbot> git [jboss-as] push master 503a18c.. jaikiran More changes to migrate EJB3 component to new EE framework (WIP)[02:08:26] *** jbossbot has quit IRC[02:08:50] *** jbossbot has joined #jboss-as7[02:08:50] *** ChanServ sets mode: +v jbossbot[02:09:07] <dmlloyd> heh[02:09:39] <stuartdouglas> smcgowan: I have updated the TCK runner for weld, all it needed was a version bump in the pom[02:10:31] <smcgowan> stuartdouglas: was that it, i can kick a build now[02:10:55] <dmlloyd> well, that was anti-climactic[02:12:42] <dmlloyd> smcgowan: the ejb branch is deleted now, it's all upstream[02:13:56] <smcgowan> dmlloyd: nice - i can start those builds too after I resolve the standalone config for TCK, the conflict seems to be with nihilitys commit for the welcome page[02:14:17] <smcgowan> there was a change in the jboss-web.xsd which no longer likes the verify-client attribute[02:14:52] <smcgowan> for my secure http definition[02:19:52] *** misty has quit IRC[02:26:17] <smcgowan> i have that fixed, now run some ejb3 lite tests - :)[02:29:25] *** jamezp is now known as jamezp_afk[02:34:08] *** stansilvert has joined #jboss-as7[02:38:54] *** misty has joined #jboss-as7[02:43:13] <stuartdouglas> why were the integration tests changed so they no longer fail the build?[02:43:56] *** stansilvert has left #jboss-as7[02:44:46] <smcgowan> you guys are good: http://hudson.qa.jboss.com/hudson/view/TCK6-AS7-WEB/job/tck6-as7-cdi/56/[02:45:06] <stuartdouglas> that is what I have locally as well[02:45:28] <smcgowan> that picked up the job after dmlloyd commit[02:45:36] <stuartdouglas> most of them are actually EJB fails, rather than CDI ones[02:45:48] <smcgowan> i mean push[02:46:10] <dmlloyd> neat[02:46:15] <dmlloyd> that's our best result thus far isn't it?[02:46:26] <smcgowan> dmlloyd: yes it is[02:46:27] <dmlloyd> at least, best in a while[02:46:44] <dmlloyd> yay for refactors! \o/[02:46:47] <stuartdouglas> it's the best so far, I fixed a heap of ejb integration stuff last week[02:47:24] <stuartdouglas> I did have it almost at that point locally before the big refactor, but the refactor changed so much stuff that most of it had to be re-done :-([02:47:49] <dmlloyd> sshhh!![02:49:14] *** lgao has joined #jboss-as7[02:51:50] *** pferraro has joined #jboss-as7[02:51:50] *** ChanServ sets mode: +v pferraro[02:51:56] *** misty has left #jboss-as7[02:52:59] <bbrowning> I'm seeing an error in the AS7 logs that confuses me - 20:51:19,770 ERROR [org.apache.catalina.session.ManagerBase] (MSC service thread 1-2) ClassNotFoundException while loading persisted sessions: java.lang.ClassNotFoundException: org.jboss.weld.context.conversation.ConversationIdGenerator from [Module "org.jboss.as.web:main"[02:53:22] <bbrowning> I'm interpreting that error to be that org.jboss.as.web module needs a dependency on org.jboss.weld.core - does that sound right?[02:53:49] <stuartdouglas> maybe, but I think there is something else going on[02:54:05] <stuartdouglas> is this reproducable?[02:54:16] <bbrowning> Yes, although not in an easy-to-consume test-case[02:54:27] <bbrowning> I'm just running TorqueBox's integration tests against AS7 and seeing these errors in server.log[02:54:41] <dmlloyd> seems possible that it could be a TCCL issue[02:55:08] <stuartdouglas> yea, it looks like the wrong CL is being used during deserialization[02:56:07] <bbrowning> so nothing I can fix on my end? :D[02:56:25] <stuartdouglas> you can try adding the dep and see what happens[02:56:35] <stuartdouglas> it will get rid of that error for now[02:58:10] <stuartdouglas> where is the torquebox source with AS7 integration github?[02:58:43] *** pferraro has left #jboss-as7[02:59:22] <bbrowning> it's at https://github.com/torquebox/torquebox/tree/as7[03:00:11] <bbrowning> quite a large source tree so if you're poking for a specific piece I can guide you :)[03:00:25] <stuartdouglas> how do I run the integration tests?[03:01:20] <bbrowning> "mvn -Pinteg test" from the project root, wait a long time[03:02:33] <bbrowning> well, I usually do "mvn install" from root and then "cd integration-tests && mvn test"[03:03:04] <bbrowning> use the latter way since the first command may have been made-up[03:03:11] <stuartdouglas> bbrowning: was there a stack trace to go with that exception?[03:03:18] <bbrowning> sure, one sec[03:03:41] *** jhalliday has joined #jboss-as7[03:03:50] <bbrowning> stuartdouglas: https://gist.github.com/06e99827cda0568339ee[03:04:38] <jamezp_afk> stuartdouglas: I think it's ./build.sh install -Pintegration-tests[03:04:39] <dmlloyd> oaky I'm going to check out sguilhen's IIOP stuff now...[03:05:23] <stuartdouglas> jamezp_afk: The torquebox integrations tests :-)[03:05:30] <jhalliday> oo, there is IIOP stuff now? cool.[03:06:30] <jamezp_afk> stuartdouglas: I guess I should read from the top down :-)[03:07:42] <stuartdouglas> actually adding the dep on weld to the web module is probably the correct solution[03:08:40] <bbrowning> stuartdouglas: is this a more general problem of having persisted sessions that reference external classes?[03:08:43] <stuartdouglas> although I can' t help thinking that the deployments CL should be being used in the deserialization[03:09:11] <stuartdouglas> yea, when it is de-serializing the persistent session it is trying to use the web modules class loader[03:09:21] <stuartdouglas> instead of the class loader for the deployed application[03:13:00] <stuartdouglas> bbrowning: which project do I need to build to get org.torquebox:gems-parent?[03:13:21] <bbrowning> should get built as part of torquebox itself[03:13:59] <stuartdouglas> oops, I mean rubyrems:rspec , but don't worry, I think I wound it[03:14:05] <bbrowning> k[03:15:30] <stuartdouglas> or maybe not[03:17:53] <bbrowning> 'mvn install' from the root should build and run unit tests then 'mvn test' from integration-tests. If you're running the first command I gave you, I just tried it and it's not working at the moment[03:20:04] <stuartdouglas> bbrowning: I think I am missing the ruby gems repo[03:20:23] <bbrowning> hrm[03:20:52] <bbrowning> ahh yes we have a custom settings file that I've copied to mvn home so I always forget about it[03:20:59] <bbrowning> mvn -s build-support/settings.xml install[03:22:48] <bbrowning> we have a source guide at http://torquebox.org/development/source-guide/ but have been lazy about updating it for the as7 branch[03:25:34] <bbrowning> stuartdouglas: as expected, adding weld to the org.jboss.as.web.main module did get rid of those errors[03:27:23] <stuartdouglas> ok, I'm more interested in why it is using that CL to start with, because if there are any classes from the deployment persisted in the session then it is going to fail[03:27:46] <bbrowning> yep[03:35:01] *** mbg|away is now known as mbg[03:38:54] *** mdobozy has joined #jboss-as7[03:38:55] *** Nihility has quit IRC[03:39:56] <smcgowan> stuartdouglas: can i ask you about your thoughts on this issue: https://issues.jboss.org/browse/JBCTS-1101[03:39:57] <jbossbot> jira [JBCTS-1101] Redirected to: https://issues.jboss.org/login.jsp?permissionViolation=true&os_destination=%2Fsi%2Fjira.issueviews%3Aissue-xml%2FJBCTS-1101%2FJBCTS-1101.xml[03:41:59] * stuartdouglas looks[03:44:08] *** sgilda has quit IRC[03:45:03] <stuartdouglas> I am not really sure, what exception are you getting in the failed tests?[03:52:20] <smcgowan> stuartdouglas: that's about it: http://pastebin.test.redhat.com/49732[03:52:48] *** clebert has quit IRC[03:53:09] <stuartdouglas> Is that with debry deployed in WEB-INF/lib as mentioned in the JIRA?[03:53:11] *** mbg is now known as mbg|away[03:53:16] <smcgowan> no[03:53:33] <stuartdouglas> how is derby set up?[03:53:36] <smcgowan> i was just trying that to see, as some suggest that is how it needs to be done[03:53:41] <smcgowan> derby is a module[03:53:47] <smcgowan> org.apache.derby[03:53:55] <smcgowan> containing derbyclient.jar[03:55:17] <stuartdouglas> is there a module dep set up in the deployments manifest?[03:55:26] <smcgowan> no[03:55:48] <stuartdouglas> that should fix it[03:56:10] <smcgowan> add the dep to the MANIFEST of the war?[03:56:17] *** mdobozy has left #jboss-as7[03:56:18] <stuartdouglas> yes[03:56:31] <stuartdouglas> Dependencies: org.apache.derby[03:58:02] <smcgowan> really?[03:58:36] <stuartdouglas> just because derby is a module in the server does not give deployments access to it[04:06:45] <jhalliday> is testMissingService known to be broken, or is a problem with my environment?[04:09:54] <stuartdouglas> dmlloyd: Is there any plan to add some way of making modules / jars globally accessible to all deployments? sort of like what you get in AS6 when you drop a jar in the lib folder?[04:10:06] <dmlloyd> no, not really[04:11:55] <stuartdouglas> https://issues.jboss.org/browse/JBCTS-1101 seems to be a problem with the derby drivers not being availible, and we are not allowed to modify the archive to add Dependencies: entries[04:11:56] <jbossbot> jira [JBCTS-1101] Redirected to: https://issues.jboss.org/login.jsp?permissionViolation=true&os_destination=%2Fsi%2Fjira.issueviews%3Aissue-xml%2FJBCTS-1101%2FJBCTS-1101.xml[04:13:47] <stuartdouglas> imho it would be a useful feature, especially for cases where third party apps expect the container to make things available, as opening up a third part jar to modify the manifest is not an ideal solution[04:14:18] <smcgowan> stuartdouglas: is this a different issue than what bbrowning describes above?[04:14:33] <dmlloyd> can't we have the JSP data source thing create the dep like @DSD does?[04:14:40] <dmlloyd> or is it happening too late[04:14:42] <stuartdouglas> smcgowan: yes, that issue was to do with de-serializing a persistent sesion[04:15:38] <stuartdouglas> I think it is happening too late[04:16:32] <dmlloyd> there's nothing in the servlet spec that says that JDBC drivers have to magically be on the class path, is there?[04:16:57] <smcgowan> dmlloyd: let me check the JSTL section on that[04:17:35] <dmlloyd> if so, that could get interesting, we'd have to relink every web deployment every time a JDBC driver was deployed or undeployed :)[04:22:30] *** JimMa has joined #jboss-as7[04:23:09] <jhalliday> could someone spare a moment to look at ServerModelControllerImplUnitTestCase.testMissingService please. I'm seeing failure which I think is a timing issue - the ServerControllerImpl$ServerStateMonitorListener tick is not happening in my logs, whereas it does in the (passing) hudson run.[04:23:25] <dmlloyd> what platform, jhalliday?[04:23:46] <jhalliday> fedora, jdk 6_24[04:24:18] <stuartdouglas> jhalliday: does it fail every time?[04:24:22] <dmlloyd> hmm O dpm[04:24:26] <dmlloyd> I don't see that problem locally[04:24:30] <jhalliday> yup, so far anyhow[04:24:37] <stuartdouglas> ServerControllerImpl$ServerStateMonitorListener has race conditions[04:24:49] <stuartdouglas> where tick and untick can come in the wrong order[04:25:10] <stuartdouglas> or at least it did, I am not 100% sure if they have been fixed or not[04:26:28] <jhalliday> well my results differ from hudson's for the same git checksum, so it's either my environment or a timing issue in the code.[04:27:03] <jhalliday> I shall consider the matter further after lunch.[04:27:10] *** jhalliday is now known as jhalli_lunch[04:30:36] <dmlloyd> okay so here's what we can do for the SQL thing[04:30:55] <dmlloyd> the spec talks specifically about DriverManager.[04:31:00] <dmlloyd> so, let's give them DriverManager.[04:31:21] <dmlloyd> when we deploy a driver, we add it via registerDriver, and when we remove it we call deregisterDriver[04:31:54] <dmlloyd> then this thing can use DriverManager, which it honestly probably already is doing[04:32:03] * dmlloyd works up a quick patch[04:32:37] <smcgowan> dmlloyd: did you see this comment by remy: https://issues.jboss.org/browse/JBCTS-1101?focusedCommentId=12600923&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-12600923[04:32:38] <jbossbot> jira [JBCTS-1101] Redirected to: https://issues.jboss.org/login.jsp?permissionViolation=true&os_destination=%2Fsi%2Fjira.issueviews%3Aissue-xml%2FJBCTS-1101%2FJBCTS-1101.xml[04:33:07] <dmlloyd> yeah[04:37:56] <dmlloyd> http://github.com/dmlloyd/jboss-as/commit/79384fe[04:37:56] <jbossbot> git [jboss-as] 79384fe.. David M. Lloyd Make Drivers work from JSTL[04:38:00] <dmlloyd> give that a shot[04:38:09] <dmlloyd> stuartdouglas: wdyt?[04:38:51] <dmlloyd> (given servlet 10.1.1 and the _probable_ implementation of the JSTL crap, I'm hopeful)[04:39:03] <smcgowan> dmlloyd: ok[04:39:13] <stuartdouglas> where do you inject the driver registry? Is there[04:39:18] <stuartdouglas> any more to the patch?[04:39:30] <stuartdouglas> oh, that was already there[04:39:31] <dmlloyd> no, that's the whole thing[04:39:32] <stuartdouglas> never mind[04:39:32] <dmlloyd> yeah[04:39:39] <dmlloyd> yeah there was no copyright :([04:39:57] <stuartdouglas> I think that should fix the issue[04:41:34] *** smarlow has quit IRC[04:41:37] <stuartdouglas> jhalli_lunch: I can't reproduce your issue, I have been running it repeatedly on my laptop in a few different situations (normally, using taskset to limit it to a single cpu and under heavy load)[04:46:16] *** sebersole has quit IRC[04:47:28] *** bbrowning is now known as bbrowning_away[04:59:01] <smcgowan> dmlloyd: not having success with that patch[04:59:25] <dmlloyd> hmmm, so the JSTL impl must not be using DriverManager[04:59:36] <dmlloyd> will have to look and see what it's doing...[05:00:08] <smcgowan> dmlloyd: we can talk to remy in the morning[05:00:24] <dmlloyd> sure. I think this gets us some of the way there though[05:00:32] <dmlloyd> if we can get JSTL to use DriverManager, we're gtg[05:00:45] <smcgowan> i think so too[05:04:16] *** stliu has joined #jboss-as7[05:04:17] *** ChanServ sets mode: +v stliu[05:04:35] *** jhalli_lunch is now known as jhalliday[05:07:20] *** Binbin has quit IRC[05:09:55] *** Binbin has joined #jboss-as7[05:14:15] <jhalliday> oh great, now I can't even get as far as the failing test because a snapshot dependency has vanished from the repo.[05:14:21] * jhalliday is not a maven fan[05:15:37] <dmlloyd> stuartdouglas, smcgowan: small problem. I'm reading the source for java.sql.DriverManager and it tests the caller's class loader to see if it can "see" the JDBC driver class, for some stupid reason.[05:16:03] <dmlloyd> it uses the caller's CL directly, rather than using the TCCL[05:16:32] <dmlloyd> it dates back to the old, old days when classloader isolation was considered "security[05:17:03] <dmlloyd> so we'll have to have our own[05:17:13] <dmlloyd> the actual logic here is simple enough to replicate[05:17:22] <dmlloyd> but it does mean a tomcat enhancement :([05:17:52] <smcgowan> this is 'old'[05:18:27] <jhalliday> who does org.jboss.as:jboss-as-welcome:war:7.0.0.Beta4-SNAPSHOT belong to?[05:18:54] <smcgowan> jhalliday: that should be in as7 dist[05:19:09] <dmlloyd> honestly we could use org.jboss.as.connector.registry.DriverRegistry directly[05:19:09] <jhalliday> hmm[05:19:45] <dmlloyd> though I think I should keep this patch; it will make silly deployments which use DriverManager work *if* they add the dependency on the JDBC driver[05:20:35] <smcgowan> jhalliday: there was a change earlier today: https://github.com/jbossas/jboss-as/commit/52150c00c6c47f7dad3d0815a2fad64b34038bbf[05:20:36] <jbossbot> git [jboss-as] 52150c0.. Jason T. Greene Add support for a static welcome root[05:21:16] <dmlloyd> no[05:21:21] <dmlloyd> actually that change *removed* jboss-as-welcome[05:21:44] <smcgowan> i just saw that[05:21:48] <dmlloyd> you probably need a clean build, jhalliday, and you probably also have garbage in your tree: try a "git clean"[05:22:14] <dmlloyd> "git status" to verify that there is/is not garbage[05:22:53] <jhalliday> there isn't, I didn't have the git tree on this box at all until an hour ago so I'm pretty sure it's fresh.[05:23:24] <dmlloyd> aha, well that's the problem[05:23:31] <dmlloyd> the opposite rather[05:23:42] <dmlloyd> jboss-as-welcome was removed, but not removed enough[05:25:24] <dmlloyd> http://github.com/dmlloyd/jboss-as/commit/af808e8[05:25:24] <jbossbot> git [jboss-as] af808e8.. David M. Lloyd Get rid of leftover welcome stuff[05:25:38] <dmlloyd> that should fix it[05:25:47] <dmlloyd> guess I'll push those both up[05:26:29] <jbossbot> git [jboss-as] push master 19ec2ba.. David M. Lloyd Get rid of leftover welcome stuff[05:26:29] <jbossbot> git [jboss-as] push master 589e366.. David M. Lloyd Make Drivers work from DriverManager, for those who want it[05:26:30] <jbossbot> git [jboss-as] push master URL: http://github.com/jbossas/jboss-as/compare/016c909...589e366[05:26:34] <dmlloyd> try that on for size[05:27:31] *** smcgowan has quit IRC[05:28:55] <dmlloyd> damn, the TCK is seriously expecting the class to be on the class path[05:28:58] <dmlloyd> er[05:28:59] <dmlloyd> *cough*[05:29:09] <dmlloyd> what I meant to say was, it's a lovely night[05:29:31] <dmlloyd> heh[05:32:36] *** magesh has joined #jboss-as7[05:36:03] <stuartdouglas> I just went to lunch, did that patch fix the JDBC thing?[05:36:36] <stuartdouglas> reading the scrollback it does not sound like it[05:37:13] <dmlloyd> nah DriverManager is pretty dumb[05:37:26] <dmlloyd> it makes DriverManager work though if you do add the dep[05:37:56] <dmlloyd> the, uh, code in question seems to assume that all JDBC drivers are on the classpath[05:38:04] <dmlloyd> which seems like something it should not be doing, by spec[05:40:02] <jhalliday> anyone know if Shelly's changeset for AS7-795 is stashed somewhere public, or do I need to start from scratch?[05:40:03] <jbossbot> jira [AS7-795] Upgrade to JBossTS 4.15 [Open (Unresolved) Task, Major, Jason Greene] https://issues.jboss.org/browse/AS7-795[05:50:28] *** bstans_afk has quit IRC[05:57:10] *** liweinan has joined #jboss-as7[05:59:26] *** mbg|away is now known as mbg[06:13:17] *** frainone has quit IRC[06:14:29] *** ccrouch has joined #jboss-as7[06:14:30] *** ChanServ sets mode: +v ccrouch[06:45:35] *** magesh1 has joined #jboss-as7[06:46:41] *** magesh has quit IRC[06:57:43] <dmlloyd> jhalliday: I think the diff in the issue is the whole thing...[07:02:15] <jhalliday> dmlloyd: yeah, I'm further down the road already thanks, breaking the ejb container :-)[07:04:10] <jhalliday> is there an easy way to a) tell the build to grab all the source.jar from mvn repo and b) run a single unit test under the debugger?[07:04:33] <dmlloyd> I use IDEA to do (a)[07:04:36] <stuartdouglas> a remote test?[07:04:43] <stuartdouglas> sorry, and integration test?[07:04:51] <stuartdouglas> or a smoke test?[07:04:58] <jhalliday> smoke[07:05:07] <stuartdouglas> the smoke tests are a pita to debug[07:05:16] <stuartdouglas> if there is a corresponding demo I usually just use that[07:05:23] *** ALR1 has joined #jboss-as7[07:05:32] *** ALR1 has quit IRC[07:05:41] <jhalliday> the tx upgrade killed the ejb container, so now I don't have a working AS build to run a demo in.[07:06:04] <jhalliday> unless I disable the tests and get a 'working' build that way I guess[07:06:14] <dmlloyd> I"ve done that[07:06:18] <dmlloyd> there's no shame in it! :)[07:06:21] *** ALR has quit IRC[07:06:22] <stuartdouglas> the tests run after the build is finished[07:06:42] <stuartdouglas> you can just kill the build when the smoke tests start and it will still run[07:19:33] *** liweinan has quit IRC[07:21:23] *** ccrouch has quit IRC[07:53:08] *** miclorb has quit IRC[08:00:32] *** jfclere has joined #jboss-as7[08:00:32] *** ChanServ sets mode: +v jfclere[08:05:46] *** tdiesler has joined #jboss-as7[08:05:46] *** ChanServ sets mode: +v tdiesler[08:15:59] *** kkhan has joined #jboss-as7[08:15:59] *** ChanServ sets mode: +v kkhan[08:17:16] *** Binbin has quit IRC[08:20:19] *** irooskov has joined #jboss-as7[08:20:19] *** ChanServ sets mode: +v irooskov[08:33:37] *** Binbin has joined #jboss-as7[08:33:37] *** jamezp_afk has quit IRC[08:34:38] *** opalka has joined #jboss-as7[08:34:38] *** ChanServ sets mode: +v opalka[08:40:08] *** mgoldmann has joined #jboss-as7[08:40:08] *** ChanServ sets mode: +v mgoldmann[08:46:33] *** pferraro has joined #jboss-as7[08:46:33] *** ChanServ sets mode: +v pferraro[08:48:30] *** mlinhard has joined #jboss-as7[08:48:30] *** ChanServ sets mode: +v mlinhard[08:49:29] *** pferraro has left #jboss-as7[09:02:27] *** adietisheim has joined #jboss-as7[09:02:30] *** ChanServ sets mode: +v adietisheim[09:11:24] *** alesj has joined #jboss-as7[09:13:07] *** jhalliday has quit IRC[09:15:44] *** rawbdor has joined #jboss-as7[09:15:44] *** ChanServ sets mode: +v rawbdor[09:25:22] *** mbg has quit IRC[09:36:12] *** dmison has quit IRC[09:46:12] *** aslak has joined #jboss-as7[09:46:12] *** ChanServ sets mode: +v aslak[09:52:04] <opalka> kkhan, ping[09:52:13] <kkhan> opalka: hey[09:52:18] <opalka> kkhan, Hi Kabir[09:52:30] <opalka> kkhan, a DMR question[09:52:36] <opalka> kkhan, I'm investigating how to do it right[09:52:58] <opalka> kkhan, now we have list-endpoints operation for webservices address on model[09:53:14] <opalka> kkhan, I'd like to make all endpoints to be visible using ls command in CLI[09:53:23] <opalka> kkhan, I'm wondering how to achieve it?[09:54:18] <kkhan> Hmm, I'm not sure what happens behind the scenes when 'ls' is used[09:54:41] *** rmaucher has joined #jboss-as7[09:55:01] <kkhan> Does it do a :read-resource?[09:55:16] <opalka> kkhan, it finds and prints children elements AFAIK[09:55:49] * kkhan is taking a quick look at the model[09:56:29] <opalka> kkhan, from my investigation I need to introduce webservices/endpoint element[09:56:42] <opalka> kkhan, that will have WS endpoints childrens[09:57:25] <kkhan> Yeah, so when you add/remove an endpoint you need to add/remove that information to the model[09:57:32] *** wolfc has joined #jboss-as7[09:57:32] *** ChanServ sets mode: +v wolfc[09:59:05] <opalka> kkhan, so I should introduce add & remove operations associated with webservice/endpoint address[09:59:19] <opalka> kkhan, and these operations will just update the model accordingly ?[10:00:05] <kkhan> I guess you have some ops pushing those into msc at the moment? Normally you use the same op to update msc and the model[10:01:16] <kkhan> Ah, I see in WSEndpointysListOperationHandler you are using an EndpointRegistry[10:01:55] <opalka> kkhan, The MSC endpoint registration is going to happen soon[10:02:10] <opalka> kkhan, I have also another question[10:02:16] <opalka> kkhan, Let's say I have the following endpoint in the model /subsystem=webservices/endpoint=jboss.ws:context=jaxws-samples-jaxbintros,endpoint=EndpointBean[10:02:32] <opalka> kkhan, how can I remove it using DMR API?[10:04:11] <kkhan> Normally you register an ADD handler for the address you want populated, e.g. [subsystem, webservices, endpoint, testthing][10:04:30] <kkhan> The add operation handler then populates a model node under that address[10:04:44] <kkhan> Similarly[10:05:01] <kkhan> you register a REMOVE handler for the address you want populated[10:05:24] <kkhan> This then deletes the node under the address[10:05:26] <kkhan> Sorry[10:05:39] <kkhan> Address should be subsystem, webservices, endpoint, *][10:05:57] <kkhan> So if you call ADD for [ss, ws, ep, newthing[[10:06:01] <kkhan> you get a new node there[10:07:35] <kkhan> Just search for classes called *RemoveHandler and *AddHandler, they implent ModelRemoveOperationHandler and ModelUpdateOperationHandler[10:07:40] <opalka> kkhan, let's say I have registered both ADD & REMOVE handlers for [ss, ws, ep, *][10:07:48] <opalka> kkhan, what place you would suggest to call them?[10:08:05] <opalka> kkhan, Introduce DUP processor that will use DMR client API to call these handlers ?[10:08:36] *** torben has joined #jboss-as7[10:08:36] *** ChanServ sets mode: +v torben[10:08:51] <kkhan> Oh, these will be part of deployments (obviously)[10:09:37] <kkhan> I believe that is the right way to do it, although I have been focussing on the bootstrap config files[10:10:03] *** hardy__ has joined #jboss-as7[10:11:28] <opalka> kkhan, U mean to register these handlers in ServerControllerModelUtil like DeplymentAddHandler is?[10:11:54] <kkhan> Yes, although you would need to do it in your subsystem[10:12:00] <kkhan> Looking for a good exampl[10:12:03] <kkhan> e[10:12:15] <opalka> kkhan, yep, example would help a lot[10:12:36] <kkhan> Look at ConnectorExtension[10:13:12] <kkhan> The ExtensionContext already contains the subsystem=connector part[10:13:48] <kkhan> So when you register ConnectorSubsystemAdd.INSTANCE under "add"[10:14:49] <kkhan> The operation belongs to that subsystem only[10:15:08] <opalka> kkhan, yes, we have this ATM - https://github.com/ropalka/jboss-as/blob/wsintegration/webservices/server-integration/src/main/java/org/jboss/as/webservices/dmr/WSExtension.java[10:15:15] *** AndyTaylor has joined #jboss-as7[10:15:15] *** ChanServ sets mode: +v AndyTaylor[10:15:22] <opalka> kkhan, this is the place where we register all our handlers[10:16:06] <kkhan> Actually, my example wasn't very good[10:16:13] <opalka> kkhan, :)[10:16:15] <kkhan> It just shows how to add/remove the subsystem[10:17:06] <kkhan> ok, ThreadsExtension[10:17:27] <kkhan> So that contains ADD for the subsystem[10:17:45] <kkhan> i.e. susbsystem=threads[10:18:05] <kkhan> Then we create a child registration under thread-factory[10:18:22] <kkhan> Since the PathElement only contains one thing[10:18:38] <kkhan> this becomes thread-factory=*[10:18:50] <kkhan> And against that we register ADD/REMOVE etc.[10:19:23] <kkhan> So when an operation comes called 'add' for subsystem=threads,thread-factory=myfactoryt[10:19:30] <kkhan> that add handler is called[10:19:57] <kkhan> So the address + op name decide what is called, if that makes sense[10:20:13] <opalka> kkhan, sure, makes perfect sense.[10:20:18] <opalka> kkhan, thanks for the pointer. This is the example I was looking for :)[10:20:23] <kkhan> Cool :-)[10:20:40] <kkhan> And yeah, the interfaces of the operation handlers matter[10:20:51] <kkhan> as I mentioned above[10:21:30] <opalka> kkhan, last question[10:22:07] <opalka> kkhan, https://github.com/ropalka/jboss-as/blob/wsintegration/webservices/server-integration/src/main/java/org/jboss/as/webservices/dmr/WSExtension.java#L58[10:22:19] <opalka> kkhan, OperationEntry.EntryType.PRIVATE[10:22:40] <kkhan> DESCRIBE is an internal operation which users cannot see[10:22:42] <opalka> kkhan, if I specify this parameter ^^^, does it mean only subystem can see this operation and it's not visible in the model ?[10:23:01] <kkhan> It is invisible in the model, yes[10:23:02] <opalka> kkhan, well I want to register both ADD & REMOVE handlers that would be private[10:23:10] <opalka> kkhan, thanks for the answer ;)[10:23:10] *** galderz has joined #jboss-as7[10:23:11] *** ChanServ sets mode: +v galderz[10:23:18] <kkhan> I think your should be public[10:23:28] <kkhan> since DESCRIBE is kind of a special operation[10:23:37] <kkhan> In domain mode[10:23:48] <opalka> kkhan, well we don't want our uses to allow add / remove WS endpoints[10:23:58] <kkhan> opalka: ok, maybe[10:24:13] <opalka> kkhan, can U elaborate more on DESCRIBE in domain mode ?[10:24:30] <kkhan> It is used at bootstrap[10:24:35] <opalka> kkhan, I have assigned the following issue - AS7-740[10:24:37] <jbossbot> jira [AS7-740] WS Endpoint doesn't appear in domain mode [Open (Unresolved) Bug, Major, Richard Opalka] https://issues.jboss.org/browse/AS7-740[10:24:59] <kkhan> we parse domain.xml and get the domain model built up, with all the subsystems[10:25:12] <kkhan> Then when we start the servers[10:25:28] *** aslak has quit IRC[10:25:30] <kkhan> we call DESCRIBE on the active profile in the domain model[10:25:45] <kkhan> which then calls DESCRIBE on each of its subsystems[10:25:56] <kkhan> the job of the DESCRIBE op[10:26:02] *** aslak has joined #jboss-as7[10:26:02] *** ChanServ sets mode: +v aslak[10:26:10] <kkhan> is to translate the contents of its portion of the model[10:26:27] <kkhan> into operations to be able to recreate that content[10:26:27] <opalka> kkhan, makes sense, something like proxy for other nodes[10:26:38] <opalka> s/nodes/cluster nodes/[10:26:51] <kkhan> These operations are then what are sent across to the server instances[10:27:01] <kkhan> since all changes to the model are done via ops[10:27:12] <kkhan> ok?[10:27:21] <opalka> kkhan, thx for explanation[10:27:25] <kkhan> np[10:28:01] *** aslak has quit IRC[10:28:21] *** aslak has joined #jboss-as7[10:28:21] *** ChanServ sets mode: +v aslak[10:39:21] *** aslak has quit IRC[10:39:39] *** aslak has joined #jboss-as7[10:39:39] *** ChanServ sets mode: +v aslak[10:40:52] *** jlocker has joined #jboss-as7[10:40:52] *** ChanServ sets mode: +v jlocker[10:42:14] *** maeste has joined #jboss-as7[10:42:14] *** ChanServ sets mode: +v maeste[10:44:12] *** aslak has quit IRC[10:50:38] *** hbraun has joined #jboss-as7[10:52:56] *** jcosta has joined #jboss-as7[10:52:57] *** ChanServ sets mode: +v jcosta[10:56:16] *** aslak has joined #jboss-as7[10:56:16] *** aslak has joined #jboss-as7[10:56:16] *** ChanServ sets mode: +v aslak[11:04:04] <nickarls> g'day[11:04:21] <nickarls> where to start looking when a failed deployment just says "undefined"?[11:04:26] <nickarls> stack trace is http://pastebin.com/iPPaUSqk[11:04:41] <nickarls> this is upstream master built a few hours ago[11:05:30] <nickarls> standalone war with EJB 3.1, CDI, JPA, JSF 2 etc. Just some unsupported stuff?[11:15:18] <maeste> hbraun: ping[11:15:27] <hbraun> maeste: 5 min[11:15:34] <maeste> may I re-close AS7-726?[11:15:35] <jbossbot> jira [AS7-726] create a driver repository service [Reopened (Unresolved) Task, Major, Stefano Maestri] https://issues.jboss.org/browse/AS7-726[11:15:40] <maeste> hbraun: oki[11:27:56] *** alesj has quit IRC[11:28:08] *** sannegrinovero has joined #jboss-as7[11:28:08] *** sannegrinovero has quit IRC[11:28:08] *** sannegrinovero has joined #jboss-as7[11:28:08] *** ChanServ sets mode: +v sannegrinovero[11:31:09] *** magesh1 has quit IRC[11:31:42] *** vtunka has joined #jboss-as7[11:40:52] *** emmanuel has quit IRC[11:41:45] *** pmuir has joined #jboss-as7[11:41:46] *** pmuir has joined #jboss-as7[11:41:46] *** ChanServ sets mode: +v pmuir[11:42:08] *** vtunka has quit IRC[11:47:16] <hbraun> maeste: just assign it to me[11:47:25] <hbraun> i'll close once i've verified that it works[11:47:33] <hbraun> maeste: currently working on that[11:47:39] <maeste> hbraun: oki[11:47:53] <hbraun> but so far it looks good[11:48:02] <maeste> cool[11:48:14] <hbraun> i'll be able to finish that task today[11:48:32] <hbraun> maeste: currently ading the driver selection to the UI[11:48:39] <hbraun> it's really much better[11:48:42] <maeste> hbraun: I've just seen I haven't removed drivers type from AS7 xsd, I'm fixing right now[11:49:05] <maeste> hbraun: yup I think it's much better too[11:55:14] *** magesh has joined #jboss-as7[11:55:58] *** Binbin has quit IRC[11:56:29] *** davidbos has joined #jboss-as7[12:04:18] *** pmuir has quit IRC[12:04:58] *** pmuir has joined #jboss-as7[12:04:58] *** pmuir has joined #jboss-as7[12:04:58] *** ChanServ sets mode: +v pmuir[12:09:21] *** jwulf has quit IRC[12:28:41] *** tdiesler has quit IRC[12:30:34] <hbraun> maeste: in the data source description there is still a "driver-class"[12:30:49] <hbraun> shouldn't that be part of the driver resource instead[12:30:51] <hbraun> ?[12:35:29] *** tdiesler has joined #jboss-as7[12:35:29] *** ChanServ sets mode: +v tdiesler[12:41:59] *** pferraro has joined #jboss-as7[12:41:59] *** ChanServ sets mode: +v pferraro[12:45:36] *** kcbabo has joined #jboss-as7[12:45:36] *** ChanServ sets mode: +v kcbabo[12:54:02] *** darranl has joined #jboss-as7[12:54:02] *** darranl has joined #jboss-as7[12:54:02] *** ChanServ sets mode: +v darranl[12:57:47] *** pil-dinner is now known as pilhuhn[12:59:37] *** JimMa has quit IRC[13:07:47] *** darranl is now known as darranl_afk[13:09:46] *** tcrawle__ has joined #jboss-as7[13:11:57] *** tcrawley_away has quit IRC[13:23:31] *** sgilda has joined #jboss-as7[13:24:26] *** epbernard has joined #jboss-as7[13:24:26] *** epbernard is now known as emmanuel[13:24:26] *** ChanServ sets mode: +v emmanuel[13:24:55] *** bbrowning_away is now known as bbrowning[13:25:19] *** Binbin_ has joined #jboss-as7[13:30:28] *** bbrowning has quit IRC[13:31:57] <maeste> hbraun: "driver-class in the data source description it's not mandatory, and it is overriding the driver's one. It is needed also when IronJacamar is used standalone and drivers are loaded as librararies[13:33:50] <pilhuhn> Hey[13:34:04] <pilhuhn> anyone else seeing a compile error in shouldBeAbleToInject(org.jboss.as.arquillian.container.managed.IntegrationTestCase) ?[13:37:48] *** AndyTaylor has quit IRC[13:39:06] <pilhuhn> s/compile/build/[13:44:32] *** Jaikiran has joined #jboss-as7[13:44:33] *** ChanServ sets mode: +v Jaikiran[13:45:01] *** darranl_afk is now known as darranl[13:47:45] *** bbrowning has joined #jboss-as7[13:47:45] *** ChanServ sets mode: +v bbrowning[13:48:59] *** lgao has quit IRC[13:49:35] <hbraun> maeste: yes, but I mean the operation description, not the schema[13:54:31] <hbraun> maeste: on your branch, this one still exists: https://issues.jboss.org/browse/AS7-719[13:54:33] <jbossbot> jira [AS7-719] En/Disable datasource returns error although state change is applied [Open (Unresolved) Bug, Critical, Stefano Maestri] https://issues.jboss.org/browse/AS7-719[13:56:23] *** magesh has quit IRC[13:56:28] *** magesh has joined #jboss-as7[14:00:40] *** ccrouch has joined #jboss-as7[14:00:40] *** ChanServ sets mode: +v ccrouch[14:12:15] *** vtunka has joined #jboss-as7[14:13:51] *** bstansberry has joined #jboss-as7[14:13:52] *** ChanServ sets mode: +v bstansberry[14:14:48] <pilhuhn> is the general idea to have multiple domains running on one physical OS? Could there be 2 host controllers? Or 2 domain controllers?[14:15:27] <pilhuhn> In theory with all ports being remapped, I guess yes. But is this realistic? Would the HC still know their managed AS ?[14:15:40] *** adietisheim has quit IRC[14:16:18] *** adietisheim has joined #jboss-as7[14:16:19] *** ChanServ sets mode: +v adietisheim[14:17:42] <bbrowning> stuartdouglas: I created AS7-799 from our conversation about 12 hours ago[14:17:43] <jbossbot> jira [AS7-799] Wrong Classloader Used When Deserializing Persisted Sessions [Open (Unresolved) Bug, Major, Jason Greene] https://issues.jboss.org/browse/AS7-799[14:17:46] *** AndyTaylor has joined #jboss-as7[14:17:47] *** ChanServ sets mode: +v AndyTaylor[14:18:48] <hbraun> maeste: fyi https://issues.jboss.org/browse/AS7-800[14:18:49] <jbossbot> jira [AS7-800] Index out of bounds when creating datasource [Open (Unresolved) Bug, Blocker, Stefano Maestri] https://issues.jboss.org/browse/AS7-800[14:19:58] *** pgier has quit IRC[14:22:34] *** pgier has joined #jboss-as7[14:22:34] *** ChanServ sets mode: +v pgier[14:24:03] <hbraun> maeste: something seems to be severely broken with the datasource operations: https://issues.jboss.org/browse/AS7-801[14:24:05] <jbossbot> jira [AS7-801] Subsystem "datasource": Management operations result in inconsistent state [Open (Unresolved) Bug, Blocker, Jason Greene] https://issues.jboss.org/browse/AS7-801[14:29:46] *** jpederse has joined #jboss-as7[14:33:48] *** jpederse has quit IRC[14:33:48] *** jpederse has joined #jboss-as7[14:34:13] *** jpederse has left #jboss-as7[14:34:36] *** jpederse has joined #jboss-as7[14:34:36] *** ChanServ sets mode: +v jpederse[14:35:48] *** lgao has joined #jboss-as7[14:36:25] *** lgao has quit IRC[14:41:15] *** aloubyansky has joined #jboss-as7[14:45:29] <hbraun> maeste: ping[14:46:29] *** tcrawley_away has joined #jboss-as7[14:49:15] *** tcrawle__ has quit IRC[14:50:05] *** stliu has quit IRC[14:52:09] *** tcrawle__ has joined #jboss-as7[14:52:45] *** pferraro has left #jboss-as7[14:52:45] *** tcrawley_away has quit IRC[14:54:05] <hbraun> maeste: ping pong[14:55:22] *** jlocker1 has joined #jboss-as7[14:55:22] *** ChanServ sets mode: +v jlocker1[14:55:29] *** jlocker has quit IRC[14:56:23] *** tcrawle__ is now known as tcrawley[14:56:24] *** tcrawley has joined #jboss-as7[15:00:08] *** pilhuhn is now known as pil-meeting[15:00:39] *** jlocker has joined #jboss-as7[15:00:39] *** ChanServ sets mode: +v jlocker[15:01:07] *** jlocker has left #jboss-as7[15:03:35] *** jlocker1 has quit IRC[15:11:04] *** tcrawley_ has joined #jboss-as7[15:11:04] *** ChanServ sets mode: +v tcrawley_[15:12:21] *** tcrawley has quit IRC[15:12:42] *** mmoyses has joined #jboss-as7[15:12:42] *** ChanServ sets mode: +v mmoyses[15:13:49] *** Nihility has joined #jboss-as7[15:13:49] *** ChanServ sets mode: +v Nihility[15:18:38] *** smarlow has joined #jboss-as7[15:18:38] *** ChanServ sets mode: +v smarlow[15:22:09] *** tcrawley has joined #jboss-as7[15:22:09] *** ChanServ sets mode: +v tcrawley[15:24:04] *** tcrawle__ has joined #jboss-as7[15:24:15] *** tcrawley_ has quit IRC[15:26:42] *** tcrawley has quit IRC[15:26:45] <rmaucher> pull requests:[15:27:22] <rmaucher> Adding the JPA valve (assuming there are tests in the testsuite which validate this): https://github.com/rmaucher/jboss-as/commit/910de734e86793bcd62685e0cea2eaf9580b1fca[15:27:24] <jbossbot> git [jboss-as] 910de73.. Rémy Maucherat Add the JPA valve metadata.[15:29:30] <rmaucher> Add the URL stream handler factory for the web module so that funny URL manipulations work (not tested with Hudson, but new URL(ServletContext.getResource().toString()) works, which should be an accurate test): https://github.com/rmaucher/jboss-as/commit/491701567e7d86bcf2c5b7bcc9e147a68e474332[15:29:31] <jbossbot> git [jboss-as] 4917015.. Rémy Maucherat Add stream handler factory for web.[15:29:46] <rmaucher> the location where this is added may not be the best[15:31:33] *** tcrawle__ is now known as tcrawley[15:31:33] *** tcrawley has joined #jboss-as7[15:33:21] *** fnasser has joined #jboss-as7[15:33:46] *** frainone has joined #jboss-as7[15:33:46] *** ChanServ sets mode: +v frainone[15:35:27] *** bobmcw has quit IRC[15:37:44] *** tcrawley_ has joined #jboss-as7[15:37:44] *** ChanServ sets mode: +v tcrawley_[15:38:15] *** tcrawley has quit IRC[15:38:53] <smarlow> rmaucher: I started a NonTransactionalEmTestCase unit test that needs to be completed. NonTransactionalEmTestCase should cause the valve to be invoked, if it doesn't blow up, that is a good start.[15:39:59] <wolfc> rmaucher: ever seen this http://fpaste.org/tIdp/ ?[15:40:20] <wolfc> I have a .. in my jboss.home (for good reason :-) )[15:41:10] *** pil-meeting is now known as pilhuhn[15:44:29] *** bobmcw has joined #jboss-as7[15:44:29] *** ChanServ sets mode: +v bobmcw[15:44:34] <rmaucher> either fix your path or use an absolute path then[15:45:34] <hbraun> darranl: ping[15:45:43] <darranl> hi hbraun[15:46:02] <hbraun> what#s the authentication status?[15:46:17] <hbraun> can I use it for domain authentication?[15:47:25] *** tcrawley has joined #jboss-as7[15:47:27] *** tcrawley has quit IRC[15:47:28] *** tcrawley has joined #jboss-as7[15:47:28] *** ChanServ sets mode: +v tcrawley[15:47:56] <hbraun> darranl: is it implemented and ready to be used?[15:47:57] <darranl> just resubmitted the pull request after the comments from bstansberry at the end of last week but yes it is at the point you can start to enable BASIC auth, DIGEST auth and SSL (not CLIENT-CERT) yet[15:48:03] *** tcrawley_ has quit IRC[15:48:19] <hbraun> you mean when the pull request is accepted?[15:48:37] <hbraun> i guess it will be enabled by default, no?[15:48:59] <hbraun> i mean non-authenticated access is not possible[15:49:00] <darranl> hbraun, no I mean bstansberry has made some comments on the first pull so I have applied those changes and submitted again[15:49:31] <bstansberry> that will be my next task :)[15:50:05] <hbraun> will it be enforced once that patch is applied?[15:50:15] <darranl> for EAP yes that always starts with some part enabled for AS we have never shipped secured by default[15:50:19] <hbraun> or does it need to be configured?[15:50:21] <darranl> no the patch is just the implementation[15:50:30] <hbraun> how do I condigure it then?[15:50:34] <hbraun> configure[15:51:13] <darranl> let me send you some config I sent to bstansberry earlier (for LDAP you will need to be on VPN)[15:51:20] <hbraun> ok[15:54:59] <darranl> hbraun, sent - even if there are additional changes in the pull request the actual behaviour from the console side will be the same i.e. it will still be BASIC and DIGEST even if there is something else to move around internally[15:55:59] <hbraun> darranl: it doesn't contain domain.xml. is it the same?[15:56:40] <darranl> hbraun, yes for the moment just define it in the host.xml - I am adding it to the domain.xml shortly but the host.xml definition will still be valid[15:56:41] *** JimMa has joined #jboss-as7[15:57:04] <hbraun> but why host.xml?[15:57:21] <hbraun> so we support both?[15:57:28] <hbraun> on domain and host level?[15:57:56] <darranl> it will allow the host.xml to override the domain.xml setting e.g. local keystore location or local user properties definition[15:58:26] <hbraun> but the domain level will be the one for the console, right?[15:59:12] <darranl> the management interface security will be based on both[15:59:18] <darranl> certain configuration is host specific[15:59:29] <darranl> e.g. the keystore has to match the hostname of the host running the interface[15:59:37] <darranl> the Kerberos identity has to be matched to the actual host[15:59:49] <darranl> those kind of settings don't make sense stored centrally[16:00:12] <darranl> on the other hand certain configuration such as how to query LDAP do make sense centrally without duplication on each host[16:00:45] *** tdiesler has quit IRC[16:01:21] *** pilhuhn is now known as pil-meeting[16:08:52] *** mbg has joined #jboss-as7[16:08:52] *** ChanServ sets mode: +v mbg[16:09:57] *** mbg has quit IRC[16:09:58] *** mbg has joined #jboss-as7[16:09:58] *** barjavel.freenode.net sets mode: +v mbg[16:10:34] *** asoldano has joined #jboss-as7[16:10:34] *** ChanServ sets mode: +v asoldano[16:17:36] *** aloubyansky has quit IRC[16:18:43] <hbraun> baileyje: can you give me an estimate on https://issues.jboss.org/browse/AS7-771 ?[16:18:45] <jbossbot> jira [AS7-771] Provide extended logging management operations [Open (Unresolved) Feature Request, Major, John Bailey] https://issues.jboss.org/browse/AS7-771[16:19:03] *** mlinhard1 has joined #jboss-as7[16:19:03] *** ChanServ sets mode: +v mlinhard1[16:19:39] <baileyje> hbraun: Still working on it. In the process of adding the ops, I realized there was a bit of cleanup/refactor needed for the logging subsystem. I should have it today or early tomorrow morning.[16:19:49] <hbraun> great[16:19:51] <hbraun> tnx[16:21:26] <hbraun> dmlloyd: could you comment on https://issues.jboss.org/browse/AS7-793 ?[16:21:28] <jbossbot> jira [AS7-793] Expose thread queue usage/metrics [Open (Unresolved) Feature Request, Major, Brian Stansberry] https://issues.jboss.org/browse/AS7-793[16:21:28] *** mlinhard has quit IRC[16:21:36] *** opalka has quit IRC[16:32:31] *** clebert has joined #jboss-as7[16:32:32] *** ChanServ sets mode: +v clebert[16:32:41] *** tcrawley_ has joined #jboss-as7[16:32:41] *** ChanServ sets mode: +v tcrawley_[16:33:08] <maeste> hbraun: pong[16:33:20] <maeste> late pong[16:33:23] <hbraun> maeste: ;)[16:33:27] <hbraun> maeste: np[16:33:30] <maeste> hbraun: I've seen your mail[16:33:33] *** tcrawley has quit IRC[16:33:46] <hbraun> maeste: yes, I decided to write an email instead[16:33:59] <maeste> hbraun: I'm working on AS7-726 and AS7-803[16:34:00] <jbossbot> jira [AS7-726] create a driver repository service [Reopened (Unresolved) Task, Major, Stefano Maestri] https://issues.jboss.org/browse/AS7-726[16:34:01] <jbossbot> jira [AS7-803] Subsystem "datasource": xa datasource created but not persisted [Open (Unresolved) Bug, Blocker, Stefano Maestri] https://issues.jboss.org/browse/AS7-803[16:34:14] <maeste> hbraun: AS7-801 is reassigned[16:34:15] <jbossbot> jira [AS7-801] Subsystem "datasource": Management operations result in inconsistent state [Open (Unresolved) Bug, Blocker, Brian Stansberry] https://issues.jboss.org/browse/AS7-801[16:34:21] *** tcrawley_ is now known as tcrawley[16:34:24] <hbraun> yes, brian takes care of it[16:34:28] <hbraun> seems to be ageneral issue[16:34:31] <hbraun> general[16:34:52] <maeste> hbraun: and AS7-719 depends on 801[16:34:53] <jbossbot> jira [AS7-719] En/Disable datasource returns error although state change is applied [Open (Unresolved) Bug, Blocker, Stefano Maestri] https://issues.jboss.org/browse/AS7-719[16:34:57] <hbraun> what I am more concerned about is the interim IJ release[16:35:16] <maeste> hbraun: I have 2 question on the other 2 issues[16:35:27] <maeste> hbraun: AS7-723[16:35:28] <jbossbot> jira [AS7-723] Creating XA datasource requires driver name and version [Open (Unresolved) Bug, Major, Heiko Braun] https://issues.jboss.org/browse/AS7-723[16:35:45] <wolfc> maeste: have you seen this before http://fpaste.org/tvw3/ ?[16:36:10] <maeste> wolfc: never[16:36:25] <hbraun> maeste: just a sec[16:36:30] <hbraun> maeste: got a call[16:36:31] <wolfc> maeste: shall I create a jira out of it?[16:37:00] <maeste> wolfc: yup create it and assign it to me...if is possible with instruction to reproduce it[16:37:33] <maeste> wolfc: it seems related to service, that is changed a bit in my current branch, maybe it's already no more actual, but I'd be happy to verify[16:38:08] <maeste> hbraun: oki ping me when you will be free to chat[16:38:16] <hbraun> maeste: back[16:38:45] <maeste> hbraun: on 723 you are argueing that we should not need driver?[16:39:24] *** tcrawley_ has joined #jboss-as7[16:39:24] *** ChanServ sets mode: +v tcrawley_[16:40:55] *** emmanuel has quit IRC[16:41:55] *** tcrawley has quit IRC[16:45:05] *** vtunka has quit IRC[16:45:07] <asoldano> mmoyses, ping[16:45:18] *** mlinhard has joined #jboss-as7[16:45:18] *** ChanServ sets mode: +v mlinhard[16:45:24] <mmoyses> asoldano: hi[16:45:37] <asoldano> mmoyses, hi :-) Do you have time for a question?[16:46:17] <mmoyses> asoldano: sure[16:46:24] <asoldano> mmoyses, I'm currently trying to restore some ws-security UT functionalities in AS7[16:46:39] <asoldano> mmoyses, those includes authentication / authorization[16:46:49] <asoldano> mmoyses, using a given security domain[16:47:03] <asoldano> mmoyses, I have some code that used to rely on the SecurityAssociation[16:47:19] <asoldano> mmoyses, SecurityContextAssociation, sorry[16:47:32] *** mlinhard1 has quit IRC[16:47:46] *** epbernard has joined #jboss-as7[16:47:46] *** epbernard is now known as emmanuel[16:47:46] *** ChanServ sets mode: +v emmanuel[16:47:55] <asoldano> mmoyses, but that is not retrieving the current security context on AS7[16:48:09] <asoldano> mmoyses, so I wonder if I'm supposed to do things differently[16:48:31] *** irooskov has quit IRC[16:49:02] <asoldano> mmoyses, http://fpaste.org/XDVq/[16:49:30] <asoldano> mmoyses, line 16 internally tries retrieving the security context[16:50:03] *** tcrawley has joined #jboss-as7[16:50:03] *** ChanServ sets mode: +v tcrawley[16:50:06] <asoldano> mmoyses, sdc is the SecurityDomainContext that I'm retrieving using service injection[16:50:41] <asoldano> mmoyses, I can go past the sdc.isValid, so that's fine[16:50:56] <asoldano> mmoyses, but then my stuff blows up in the last line[16:51:45] *** tcrawley_ has quit IRC[16:51:50] <asoldano> mmoyses, http://fpaste.org/SUnH/ is the method of my securityAdaptor[16:52:39] <asoldano> mmoyses, any idea / suggestion? perhaps this needs to be done completely differently?[16:53:17] * mmoyses looks at code[16:54:10] *** jfclere has quit IRC[16:55:08] *** pferraro has joined #jboss-as7[16:55:08] *** ChanServ sets mode: +v pferraro[16:55:33] *** spagop has joined #jboss-as7[16:55:33] *** ChanServ sets mode: +v spagop[16:55:35] *** adietisheim has quit IRC[16:56:10] <mmoyses> asoldano: so security context is always null is that it?[16:56:42] <asoldano> mmoyses, yep, that's the problem[16:57:00] <asoldano> mmoyses, or at least, that's the reason of the exception[16:58:02] <mmoyses> asoldano: ok, so that's probably expected. you need to create a new security context. an example can be found in JBossWebRealm[16:58:12] <mmoyses> asoldano: see the authenticate(String, String) method[16:58:30] *** tcrawley has quit IRC[16:59:58] <asoldano> mmoyses, ok, thanks. So basically, I'm missing that "SecurityContext sc = SecurityActions.createSecurityContext(authenticationManager.getSecurityDomain());" ?[17:00:13] <darranl> bstansberry, do you have a sec? just found an inconsistency in the relative-to handling of the Logging subsystem and want to check the correct/expected behaviour[17:00:14] <darranl> https://issues.jboss.org/browse/AS7-805[17:00:16] <jbossbot> jira [AS7-805] relative-to on Logging file handlers defined with default which is not honoured. [Open (Unresolved) Bug, Major, Darran Lofthouse] https://issues.jboss.org/browse/AS7-805[17:00:18] <mmoyses> asoldano: and later on SecurityActions.setSecurityContextOnAssociation(sc);[17:00:52] <bstansberry> darranl: sure[17:01:26] <darranl> bstansberry, just want to check should a default really be specified in the xsd as this to me means it will always be relative to the default if no other is specified[17:01:30] <asoldano> mmoyses, ok, I see... good. Btw, this asssociation used to be done "automatically" in AS6 ?[17:01:58] *** pil-meeting is now known as pilhuhn[17:02:05] <bstansberry> darranl: good question[17:02:29] <darranl> trying to make sure I copy the correct behaviour myself ;-)[17:02:50] <bstansberry> baileyje: fyi, ^^^ since you are working in this area[17:03:03] <bstansberry> I think a default like that in the xsd is wrong[17:03:15] <mmoyses> asoldano: i think we had a valve in AS6 that always created an empty SC[17:03:21] <bstansberry> since it precludes absolute paths[17:03:21] <darranl> I think so as well - then omitting the relative-to allows you to be absolute[17:03:42] <bstansberry> we should just include the value in our standard configs[17:03:58] <darranl> yes the value is in the standard config already[17:04:11] <asoldano> mmoyses, ok, that would explain it. Thanks, let me see if/how I can create that SC similarly to what you do in JBossWebRealm[17:04:39] <darranl> I will submit a pull with the optional dropped so the xsd will match the current behaviour[17:04:48] <mmoyses> asoldano: ok. ping me if you need further help[17:04:48] <bstansberry> darranl: thanks[17:05:00] <asoldano> mmoyses, sure, thanks for now :)[17:05:11] *** stansilvert has joined #jboss-as7[17:05:55] <darranl> I will just check any other use of relative-to in the xsds and see if any others have the same issue[17:13:21] *** Jaikiran is now known as Jaikiran|Dinner[17:16:44] <bstansberry> Nihility: the patch from darranl raises the issue of the default security config[17:23:30] *** maeste has quit IRC[17:23:32] <dmlloyd> someone ping me if you see jamezp[17:24:07] *** AndyTaylor is now known as andy_brb[17:24:43] *** galderz has quit IRC[17:25:50] *** jamezp has joined #jboss-as7[17:27:50] *** hardy__ has quit IRC[17:28:46] *** mlinhard has quit IRC[17:29:03] *** smcgowan has joined #jboss-as7[17:29:03] *** ChanServ sets mode: +v smcgowan[17:30:35] *** mmoyses is now known as mmoyses_[17:32:57] <bobmcw> jamezp: dmlloyd is looking for you[17:33:04] <dmlloyd> oh hey[17:33:09] <jamezp> Hello[17:33:25] <dmlloyd> jamezp, would you mind doing a quick side task for me...[17:33:40] <jamezp> dmlloyd: Not at all.[17:33:52] <dmlloyd> AS is using a bunch of deprecated methods from modules that I want to kill off... could you find them and replace them with their recommended replacements[17:34:00] <dmlloyd> the javadoc says what those are[17:34:13] <jamezp> Sure thing.[17:35:07] <jamezp> I keep losing my VPN connections and it's kicking which ends up messing up my DNS.[17:37:07] <dmlloyd> thanks[17:37:34] <jamezp> dmlloyd: Do I need to create a JIRA for it? Also do I need to create one for the logging replacement stuff?[17:37:42] <hbraun> boilerplate boilerplate boilerplate[17:37:46] <dmlloyd> no and yes[17:38:08] <hbraun> i wish I had taken the time to write a code generator ...[17:38:51] <jamezp> Perfect, thanks![17:39:02] *** maeste has joined #jboss-as7[17:39:02] *** ChanServ sets mode: +v maeste[17:39:19] <jamezp> hbraun: What kind of code are you generating?[17:39:34] <hbraun> jamezp: i would, but i dion't[17:39:47] <hbraun> UI code to query the detyped model[17:40:20] <jamezp> Ah, that I don't have an answer for then :-)[17:40:22] *** frainone has quit IRC[17:40:34] <hbraun> ;)[17:41:54] *** tcrawley has joined #jboss-as7[17:41:54] *** ChanServ sets mode: +v tcrawley[17:43:37] *** JimMa has quit IRC[17:44:35] *** jfclere has joined #jboss-as7[17:47:35] <jbossbot> git [jboss-as] push master bbbaeee.. Darran Lofthouse [AS7-805] Removed 'optional' value from xsd as it is not used within code,...[17:47:36] <jbossbot> jira [AS7-805] relative-to on Logging file handlers defined with default which is not honoured. [Coding In Progress (Unresolved) Bug, Major, Darran Lofthouse] https://issues.jboss.org/browse/AS7-805[17:47:36] <jbossbot> git [jboss-as] push master URL: http://github.com/jbossas/jboss-as/compare/589e366...bbbaeee[17:51:08] *** andy_brb is now known as AndyTaylor[17:53:14] <pilhuhn> bstansberry Hi, can I assume that :add needs the same items that :read-resource / :read-resource-description emits as configuration properties[17:54:12] <bstansberry> an :add operation should provide a description of its parameters[17:54:19] <pilhuhn> I know[17:54:24] <pilhuhn> that was not the question :)[17:54:40] <kkhan> It depends on if the DescriptionProvider has been implemented properly :-[17:54:42] <kkhan> )[17:54:58] <bstansberry> there's a saying about assume making an ass of you and me[17:55:17] <kkhan> lol[17:55:29] <bstansberry> it's a reasonable assumption, but it's not a contract[17:55:33] <pilhuhn> Let me explain why I am asking: in rhq we have a thing called "resource config" which usually shows the output of :read-resource[17:55:49] <pilhuhn> now to create new resources that set of properties is also taken[17:55:59] <pilhuhn> bstansberry did not get that[17:56:10] <pilhuhn> Did not want to insult you or such[17:56:14] *** jcosta has quit IRC[17:56:15] *** alexsmirnov has joined #jboss-as7[17:56:23] <bstansberry> no, not at all![17:56:57] <pilhuhn> when the set is different, I need to see how to work around[17:57:00] *** miclorb has joined #jboss-as7[17:57:07] <bstansberry> that's just a saying; sorry I was just saying don't count on something that's not an official contract[17:57:27] <pilhuhn> ok[17:58:53] <pilhuhn> could have been that the :add params were created from the same source internally than :read-resource[17:59:47] <bstansberry> no, it doesn't work that way. it's not for example generated from a schema[18:00:16] <pilhuhn> okay - was just asking :)[18:00:26] <bstansberry> i'm trying to think of cases where an add doesn't work that way[18:01:23] <Jaikiran|Dinner> dmlloyd: can you please merge these Ejb3 related commits upstream https://github.com/jaikiran/jboss-as/commits/new-ee-framework[18:02:06] <dmlloyd> checking now[18:02:25] <bstansberry> somethings that involve list attributes likely don't; e.g. if a resource has ops like add-system-property, remove-system-property the :add operation may not take the list as a param[18:04:01] <asoldano> mmoyses_, I think the JBossCachedAuthenticationManager should be added a guard for null principal, unless you already did that (BoundedConcurrentHashMap throws NPE for null key)[18:05:36] *** tcrawley has quit IRC[18:07:01] *** tcrawley has joined #jboss-as7[18:07:01] *** ChanServ sets mode: +v tcrawley[18:07:56] *** pilhuhn is now known as pil-dinner-bbl[18:09:51] <hbraun> bstansberry: where did the domain level proeprties go?[18:09:55] <hbraun> properties[18:10:04] <bstansberry> system properties?[18:10:10] <hbraun> have they been removed or can I just not see them?[18:10:12] <hbraun> yes[18:10:19] <bstansberry> should be there[18:10:36] <bstansberry> let me boot a domain[18:10:50] <hbraun> maybe they are not in the CLI?[18:11:26] <hbraun> they exist in XML[18:12:05] *** rawbdor has quit IRC[18:12:24] <bstansberry> hbraun: http://fpaste.org/khLI/[18:12:46] <hbraun> ah, they are an attribute of the root resource[18:12:50] <hbraun> well hidden[18:13:07] <hbraun> why not /system-properties?[18:13:18] <hbraun> ah, i know[18:13:23] <hbraun> they are not addressable[18:13:25] <hbraun> tnx[18:14:21] *** tcrawley has quit IRC[18:15:34] <bstansberry> hbraun: I've considered a few times making each system property a resource instead of part of a attribute[18:15:50] *** tcrawley has joined #jboss-as7[18:15:50] *** ChanServ sets mode: +v tcrawley[18:16:06] <hbraun> yes, for me it#s the first time I see that[18:16:21] <hbraun> it#s just different then any other common operation[18:16:49] <hbraun> it#s rguable why path, and interfces are resource but system properties not[18:16:59] <hbraun> especially when they are subresource on other elements[18:17:16] <bstansberry> they've always been attributes, although there was some cruft around where resources had extraneous stuff that made it look like system-property was a valid child type[18:17:47] <hbraun> ah, hence the extra operations to add,remove props[18:17:57] <hbraun> i.e. on server-grpup[18:17:59] <bstansberry> with the addition of this "boot-time" notion, they look a whole lot more like resources than they did before[18:18:32] <bstansberry> there should be add-system-property, remove-system-property ops anywhere system properties are valid[18:19:07] <bstansberry> we can make them resources if it won't break your or pilhuhn[18:19:16] <bstansberry> s/your/you[18:19:46] <hbraun> nevermind[18:19:57] <hbraun> too late in the game and very little benefit[18:20:03] <bstansberry> ok, but this is your last chance :-)[18:20:13] <hbraun> hehe[18:20:18] *** tcrawley has quit IRC[18:20:24] <hbraun> no, thanks[18:20:52] <bstansberry> k[18:20:54] <hbraun> I can hardly catch up with all the last minute changes[18:23:45] *** tcrawley has joined #jboss-as7[18:23:45] *** ChanServ sets mode: +v tcrawley[18:23:57] <Jaikiran|Dinner> dmlloyd: i'll be off now. if some of those changes can't be committed, then let's discuss it tomorrow[18:24:00] <bstansberry> dmlloyd: seeing this on shutdown: http://fpaste.org/8jtY/[18:24:26] <bstansberry> you committed something about DriverManager, so I'm asking you first[18:24:26] <jpederse> bstansberry: AS7-804[18:24:27] <dmlloyd> hmmm[18:24:28] <jbossbot> jira [AS7-804] SecurityException when DriverService stops [Open (Unresolved) Bug, Major, Stefano Maestri] https://issues.jboss.org/browse/AS7-804[18:24:31] <bstansberry> k[18:24:32] <rmaucher> yes, me too[18:24:33] <dmlloyd> yeah that'd be me[18:24:33] <bstansberry> thanks[18:25:32] <dmlloyd> damn, DriverManager is an incredibly stupid class[18:25:39] <dmlloyd> just revert my patch, it's not worth it[18:25:53] <dmlloyd> I shall never touch DriverManager again, my lesson has been well-learned[18:26:05] <jpederse> dmlloyd: maeste has changes to that class anyway in our next update[18:26:23] <jbossbot> git [jboss-as] push master ca8800e.. jaikiran AS7-708 Add logger statements to EE module class configuration processor[18:26:24] <jbossbot> jira [AS7-708] Ability to specify classloading exceptions to ignore similar to scanning-deployers-jboss-beans.xml [Open (Unresolved) Bug, Major, David Lloyd] https://issues.jboss.org/browse/AS7-708[18:26:24] <jbossbot> git [jboss-as] push master e47a625.. jaikiran Remove dead commented out code[18:26:24] <jbossbot> git [jboss-as] push master 226bbcb.. jaikiran Fix duplicate bindings for a component[18:26:24] <jbossbot> git [jboss-as] push master 1006e14.. jaikiran Implement EJBContext.lookup[18:26:24] <jbossbot> git [jboss-as] push master c24a038.. jaikiran Implement getInvokedBusinessInterface on session beans[18:26:24] <jbossbot> git [jboss-as] push master ae1cb80.. jaikiran Fix @DataSourceDefinitions processing[18:26:25] <jbossbot> git [jboss-as] push master a3891d1.. jaikiran Fix EJBComponent.lookup() method[18:26:25] <dmlloyd> jpederse: okay great, can he revert it then?[18:26:25] <jbossbot> git [jboss-as] push master ce826eb.. jaikiran Fix @PersistenceContext processing on methods[18:26:26] <jbossbot> git [jboss-as] push master c249eb4.. jaikiran (1) Fix transaction attribute processing on bean methods (2) Introduce a toString() interceptor on bean's views for handling toString() invocation[18:26:26] <jbossbot> git [jboss-as] push master URL: http://github.com/jbossas/jboss-as/compare/bbbaeee...c249eb4[18:26:27] <jpederse> dmlloyd: yeah, it really sucks - we just hack it too[18:26:30] <Nihility> hahaha[18:26:41] <Jaikiran|Dinner> dmlloyd: thanks![18:26:43] <jpederse> dmlloyd: yup[18:26:51] <dmlloyd> jpederse, rmaucher while I have you both in the room...[18:26:52] <Jaikiran|Dinner> this should improve the tck runs for ejb30/lite by some number[18:26:56] <dmlloyd> this JSTL driver thing[18:27:06] <Nihility> Jaikiran|Dinner: a big number?[18:27:07] <Nihility> :)[18:27:18] <dmlloyd> I think it is using DriverManager to locate the JDBC driver[18:27:29] <dmlloyd> can we change it to use DriverRegistry (from the AS connector subsys) instead?[18:27:40] <rmaucher> yes, the failing tests all use that DM thing[18:27:47] <dmlloyd> it can do the same logic that DriverManager does to find the driver, without the stupid "security" checks[18:28:14] <rmaucher> don't know, that would be in the jstl impl I think[18:28:38] <jpederse> dmlloyd: those are part of the WP tests - so the JCA services won't be active[18:28:59] <jbossbot> git [jboss-as] push master 2b9c7cb.. Jason T. Greene Set transport class[18:28:59] <jbossbot> git [jboss-as] push master URL: http://github.com/jbossas/jboss-as/compare/c249eb4...2b9c7cb[18:29:11] <rmaucher> so I'll look at it a bit ;)[18:29:15] <dmlloyd> then the only other option is to have the driver be imported into the class path of the deployment[18:29:22] <dmlloyd> or to challenge the test[18:29:29] <dmlloyd> or both[18:29:37] *** Jaikiran|Dinner has quit IRC[18:29:44] <dmlloyd> unfortunately we cannot replace classes in java.sql[18:29:52] <jpederse> smcgowan: thoughts ^[18:30:12] *** torben has quit IRC[18:30:45] <dmlloyd> in principle I'd say challenge, just because nowhere in the spec does it say that DriverManager must be supported, and nowhere does it say that JDBC drivers should automatically be on the class path[18:30:47] <smcgowan> jpederse: i know about it[18:30:59] <smcgowan> jpederse: that's what i was asking you about yesterday[18:31:00] <dmlloyd> however speaking pragmatically I'd rather just put in a workaround feature[18:31:12] <dmlloyd> if we can, someho[18:31:13] <dmlloyd> w[18:31:26] <dmlloyd> like "add this dependency to deployments" or something ugly like that[18:32:02] <dmlloyd> afaict it would work fine if that driver were on the class path[18:32:07] <dmlloyd> though we should verify that[18:32:13] *** asaldhan has joined #jboss-as7[18:32:13] *** ChanServ sets mode: +v asaldhan[18:32:20] <smcgowan> dmlloyd: i've tested quite a few combinations without any luck[18:32:27] <smcgowan> but its' been a while[18:32:42] <dmlloyd> I wasn't sure if you didn't bother putting it in the MANIFEST because that wouldn't be allowed anyway[18:32:47] <dmlloyd> but it would make for a good verification[18:32:56] <asaldhan> org.jboss.as.test.surefire.servermodule.* tests fail for me in AS7 clean build[18:33:01] <smcgowan> i didn't try that one, actually but let me do that now quickly[18:33:49] <smcgowan> rmaucher: at what point does the driver class need to be available, during the deployment?[18:35:17] <rmaucher> just running the test I think, but I didn't look at the impl yet[18:35:41] <rmaucher> I only looked at the generation of all the parameters using the EL (which was looking ok)[18:39:51] *** torben has joined #jboss-as7[18:39:51] *** ChanServ sets mode: +v torben[18:42:51] *** maeste has quit IRC[18:44:27] <rmaucher> ok, so in JSTL, which has been moved to our specs project (thx Shelly), the data source business is all in a single DataSourceWrapper class, which we can rewrite as much as we like[18:51:07] <dmlloyd> cool.[18:51:18] <dmlloyd> I'm a lot less worried now :)[18:52:23] <bstansberry> hbraun: I see now that I have a commit for AS7-701 that I thought went in a week ago, but never did[18:52:24] <jbossbot> jira [AS7-701] server-config model carries both "system-property" and "system-properties" subresource [Open (Unresolved) Bug, Major, Brian Stansberry] https://issues.jboss.org/browse/AS7-701[18:52:43] <hbraun> yes, but that's just cosmetics[18:52:45] <bstansberry> wrote it while travelling and then travelled again before pushing[18:52:49] <hbraun> ;)[18:53:01] <hbraun> what about the remaining 99 issues?[18:53:03] <hbraun> ;)[18:53:28] <bstansberry> well, it generally cleans stuff up re: system properties, so you get one more chance at the "resource vs attribute" question ;)[18:54:03] <hbraun> bstansberry: maybe you can shed some light on this: https://issues.jboss.org/browse/AS7-793[18:54:05] <jbossbot> jira [AS7-793] Expose thread queue usage/metrics [Open (Unresolved) Feature Request, Major, Brian Stansberry] https://issues.jboss.org/browse/AS7-793[18:54:49] <hbraun> bstansberry: or this one? https://issues.jboss.org/browse/AS7-702[18:54:50] <jbossbot> jira [AS7-702] Expose server VM metrics through domain model [Open (Unresolved) Feature Request, Major, Brian Stansberry] https://issues.jboss.org/browse/AS7-702[18:54:52] *** asoldano is now known as asoldano_away[18:55:03] <hbraun> i would just need to know if we manage to include it or not[18:55:44] <hbraun> maybe 793 is laready possible, but I just don't know how?[18:55:58] <jamezp> dmlloyd: Maybe I'm missing some, but I only found 4 source files that were using deprecated methods https://github.com/jamezp/jboss-as/commit/5a30035ee2683cd9139466e46757974593c84bc1[18:55:58] <jbossbot> git [jboss-as] 5a30035.. James Perkins Remove usage of deprecated methods in jboss-modules usages.[18:56:04] <dmlloyd> okay cool[18:56:05] <bstansberry> I'd say 702 is a better choice[18:56:34] <hbraun> bstansberry: you mean higher priority? or better chance to get done in time?[18:57:18] <jbossbot> git [jboss-as] push master 23a4e52.. James Perkins Remove usage of deprecated methods in jboss-modules usages.[18:57:18] <jbossbot> git [jboss-as] push master URL: http://github.com/jbossas/jboss-as/compare/2b9c7cb...23a4e52[18:57:22] <bstansberry> hbraun: there's a discussion around moving thread configs into the subsystems themselves and just making the threading services a part of the core, not a separate subsystem[18:57:36] <hbraun> i c[18:57:46] <jamezp> dmlloyd: Thanks! Did you get a chance to look at the logging/message changes I made?[18:57:54] <bstansberry> so I'd say me knocking out a 702 impl is a faster way to get a stable way to look at threads[18:58:15] <dmlloyd> jamezp: yeah, we'll have some changes for sure but I have to get a minute to chat with Nihility about it[18:58:28] <hbraun> bstansberry: ahreed[18:58:33] <hbraun> bstansberry: agreed[18:58:40] <jamezp> dmlloyd: No problem. Anything I should work on in the mean time?[18:59:00] <jamezp> I could work on writing a wiki for the logging tools or something like that.[18:59:00] <hbraun> bstansberry: it's probably more important to the average user as well[18:59:16] *** spagop has quit IRC[18:59:41] *** kkhan has quit IRC[19:01:03] *** torben has quit IRC[19:01:18] *** davidbos has quit IRC[19:01:22] <bstansberry> hbraun: a deytped representation of java.lang.management.ThreadInfo is going to be verbose :)[19:03:19] <hbraun> bstansberry: system-properties on the host level?[19:03:25] <hbraun> bstansberry: same story as before?[19:03:38] *** jlocker has joined #jboss-as7[19:03:38] *** ChanServ sets mode: +v jlocker[19:03:48] *** jlocker has left #jboss-as7[19:03:51] <bstansberry> should be there. but seeing they weren't is what made me realize my commit never got applied[19:03:56] <hbraun> yes[19:05:52] <bstansberry> hbraun: if I push darranl's security thing, the console will not work. just want to verify that's ok[19:06:00] *** emmanuel has quit IRC[19:06:08] <hbraun> why will it not work?[19:06:16] <bstansberry> don't know. but it doesn't[19:06:31] <hbraun> is authentication required by default?[19:06:33] <bstansberry> so, if you had to ask, that means it's not ok :-)[19:06:55] <hbraun> it depends how much time it takes me to figure out what's wrong[19:06:55] <hbraun> #[19:07:07] *** ALR1 has joined #jboss-as7[19:07:12] <darranl> sorry may have misunderstood your question before bstansberry - did you mean it didn't work when you enabled some security?[19:07:32] <hbraun> darranl: that's what I can live with[19:07:50] <darranl> hbraun, that is what I thought bstansberry meant before but I may be wrong[19:07:55] <hbraun> if it doesn't work w/o authentication enforced it would be bad[19:07:56] <bstansberry> I also don't think it worked without, but I should double-check[19:08:08] <bstansberry> k; testing[19:08:11] <hbraun> tnx[19:09:02] *** pmuir has quit IRC[19:09:22] *** frainone has joined #jboss-as7[19:09:22] *** ChanServ sets mode: +v frainone[19:10:05] <dmlloyd> I don't think we should disable JDBC deployment for web profile[19:10:09] <dmlloyd> jpederse: ^[19:10:24] <dmlloyd> there's no reason to do that, the spec says we can distribute drivers[19:10:54] <bstansberry> hbraun darranl: looks ok. I must not have had a clean config (i.e. no security config) when I tried before[19:10:57] <dmlloyd> it also nicely solves this JSTL issue if we skip this whole DriverManager thing and go to a parallel registry[19:11:03] <dmlloyd> Nihility: wdyt[19:11:10] <hbraun> bstansberry: then we a re good to go[19:11:51] <darranl> bstansberry, that is now what we expect - hbraun next needs to get the username and password for the call[19:12:07] <jpederse> dmlloyd: only the EE 6 feature is supported - web.xml / @DSD[19:12:10] <hbraun> that's fine[19:12:15] <hbraun> i can do that tomorrow[19:12:31] <hbraun> shouldn't we provide a default configuration?[19:12:35] <hbraun> i.e. admin/admin[19:13:04] <darranl> If we want a default I think we should do that as a second step once we have the console supporting it[19:13:05] <hbraun> bstansberry, darranl: why ship it w/o a security config for the http endpoint[19:13:16] <dmlloyd> jpederse: yes but EE says we need a mechanism to distribute drivers, and there's no reason to tie JDBC deployment into JCA[19:13:17] <bstansberry> I pinged Nihility about that ^^^[19:13:25] <dmlloyd> JCA uses JDBC drivers, JDBC drivers do NOT need JCA[19:13:27] <darranl> we will have similar issues with CLI when I see Kabir this week[19:13:34] <dmlloyd> I believe WP allows this[19:13:38] <bstansberry> but yeah, for sure no reason to add it to the config before the console can handle it[19:13:55] <hbraun> bstansberry: it's the other way around[19:14:01] <hbraun> i need to add it myself[19:14:13] <hbraun> how can I verify it works otherwise?[19:14:32] <jpederse> dmlloyd: we need to discuss this with Nihility[19:14:59] <jpederse> dmlloyd: there is no JCA environment in WP[19:15:30] <dmlloyd> I'm well aware of that[19:15:34] <dmlloyd> like I said[19:15:35] <bstansberry> hbraun: well, it's pretty trivial. I can put it in if you want, or give you a branch with it in to test against[19:15:36] <dmlloyd> JCA uses JDBC drivers, JDBC drivers do NOT need JCA[19:15:46] <dmlloyd> we can deploy JDBC drivers without enabling JCA[19:15:50] <hbraun> bstansberry: no, just arguing[19:15:53] <dmlloyd> in fact it would be crazy if we couldn't[19:15:55] <hbraun> bstansberry: I can do it[19:16:02] <hbraun> bstansberry: thanks[19:16:12] <darranl> the default will only support one auth method when the interface will actually have 3 or even 4 auth methods available[19:17:34] <smcgowan> dmlloyd, rmaucher: some behavior has changed (this morning) and even w/o any dep added to the Manifest, the .war is failing to deploy due to: http://pastebin.com/XGw1GMgx[19:17:43] <smcgowan> i'll need a bit more time, just wanted to let you know[19:17:55] <dmlloyd> hmmm[19:18:06] <jpederse> dmlloyd: maybe the solution is to move the pure JDBC services to another extension[19:18:22] <smcgowan> dmlloyd: my thought exactly[19:18:50] <smcgowan> i'm going to kick off run in Hudson which will test with latest AS 7 build and clean TCK env[19:19:04] <dmlloyd> I wonder if Jaikiran's change triggered that[19:19:13] <smcgowan> i think that may be it[19:19:21] <rmaucher> hum, weird error[19:19:45] <darranl> another nice issue on the host controller bstansberry - no paths defined !![19:20:22] <bstansberry> darranl: yes, when i made that suggestion on the thread, i knew there were worms in that can[19:20:33] <darranl> yes I am finding them ;-)[19:20:55] <darranl> I have a nice relative-to working on standalone - and failing on domain mode[19:21:19] <darranl> I will create a Jira - I assume no objections bstansberry to adding the common ones to the HC process?[19:22:00] <bstansberry> no objection tothe ones that are relevant to an HC[19:22:21] <jpederse> ~pgier++[19:22:55] *** sannegrinovero has quit IRC[19:23:10] <hbraun> bstansberry: on host level it's called system-properties, on domain level sytem-property[19:23:30] <bstansberry> hbraun: that's all cleaned up in this patch i mentioned[19:23:35] <hbraun> cool[19:23:39] <bstansberry> but, last chance to make them resources!![19:24:08] <hbraun> no, thanks[19:24:17] <hbraun> it#s an over estimate feature anyway[19:24:26] <hbraun> estimated[19:27:51] *** magesh has quit IRC[19:29:56] *** hbraun has quit IRC[19:32:19] *** smarlow has quit IRC[19:33:49] *** epbernard has joined #jboss-as7[19:33:49] *** epbernard is now known as emmanuel[19:33:49] *** ChanServ sets mode: +v emmanuel[19:33:54] *** emmanuel has quit IRC[19:34:33] *** smarlow has joined #jboss-as7[19:34:33] *** ChanServ sets mode: +v smarlow[19:36:10] <jbossbot> git [jboss-as] push master ff4c6fc.. Darran Lofthouse [AS7-399] Adding authenticators to the domain management HTTP server and introducing an out of the box configuration embedded in the domain model....[19:36:11] <jbossbot> jira [AS7-399] Initial Basic and Digest Authentication for HTTP API [Coding In Progress (Unresolved) Task, Major, Darran Lofthouse] https://issues.jboss.org/browse/AS7-399[19:36:12] <jbossbot> git [jboss-as] push master URL: http://github.com/jbossas/jboss-as/compare/23a4e52...ff4c6fc[19:36:57] *** tcrawley has quit IRC[19:38:34] *** tcrawley has joined #jboss-as7[19:38:34] *** ChanServ sets mode: +v tcrawley[19:44:39] *** tcrawley has quit IRC[19:46:53] *** mmoyses_ is now known as mmoyses[19:48:46] *** darranl has quit IRC[19:49:24] <mmoyses> asaldhan: it makes sense to guard against null key. i will add that check[19:58:26] <smcgowan> dmlloyd,rmaucher: the issue above was an issue with my local env, so pls disregard, with org.apache.derby dependency added to the war MANIFEST, i no longer see the issue noted in JBCTS-1101 but there are a few test failures[19:58:27] <jbossbot> jira [JBCTS-1101] Redirected to: https://issues.jboss.org/login.jsp?permissionViolation=true&os_destination=%2Fsi%2Fjira.issueviews%3Aissue-xml%2FJBCTS-1101%2FJBCTS-1101.xml[19:59:34] <dmlloyd> okay, whew[19:59:39] <dmlloyd> I couldn't figure out what had changed[19:59:54] <dmlloyd> smcgowan: okay, well that's good news[20:00:08] <smcgowan> http://pastebin.com/sB9Cq3wE[20:00:16] <dmlloyd> we're close then... just need to work out this JDBC deployment business, or work out some other way to get that dep in there[20:01:22] <dmlloyd> jamezp: okay, so for the logging stuff[20:01:46] <jamezp> Yes?[20:01:47] <dmlloyd> the project code should be "JBAS" for everything[20:02:02] <jamezp> No problem.[20:02:10] <dmlloyd> let's start allocating codes in chunks of 100 to subsystems[20:02:15] <dmlloyd> can you create a wiki page to track the allocations?[20:02:27] <jamezp> Yes indeed I will.[20:02:43] <dmlloyd> only log messages INFO and higher need codes[20:02:56] <jamezp> I'll add message id and the message associated with it.[20:03:07] <dmlloyd> you don't need to go that far[20:03:13] <jamezp> Okay, no problem.[20:03:17] <dmlloyd> just track blocks of 100 and what subsystem they belong to[20:03:27] <jamezp> Got it, easy enough.[20:03:28] <dmlloyd> then if someone needs a block they can grab it there[20:03:41] <jamezp> Ah, yes. That makes perfect sense.[20:03:44] <dmlloyd> let's start subsystems at 10000 though. Below that we'll reserve for core stuff[20:04:04] <dmlloyd> (also in 100s)[20:04:36] <dmlloyd> finally, I think it may be best (for subsystems at least) to only have one or two interfaces for all message stuff.[20:04:50] <dmlloyd> two if it makes sense to separate logging from messages, which it may.[20:06:14] <jamezp> I agree. I don't think we would need more than one really.[20:06:28] <jamezp> Then again, the one I converted was pretty small.[20:13:34] <jamezp> dmlloyd: We're just looking for something like this https://gist.github.com/977023 correct?[20:14:15] <dmlloyd> yeah[20:14:45] <dmlloyd> we do not currently need to assign codes to exceptions though[20:15:01] <dmlloyd> I've been doing that for my projects but it's not a requirement[20:15:13] <jamezp> Ah, got it. That makes sense actually as they won't be used anyway.[20:15:20] <dmlloyd> oh, really?[20:15:30] <dmlloyd> we should change that[20:15:37] <jamezp> Yes they will. They'll be appended to the message.[20:15:49] <dmlloyd> oh okay.[20:16:00] <jamezp> Sorry, thinking while typing is not always a good thing :-)[20:16:13] <dmlloyd> beats not thinking while typing![20:16:18] <jamezp> Hahaha[20:16:31] *** AndyTaylor has quit IRC[20:16:54] <jamezp> Any particular order I should start converting the subsystems in?[20:17:44] *** lazarotti has joined #jboss-as7[20:17:51] <dmlloyd> nah it shouldn't matter, though you probably want to save ee/ejb/naming for last :)[20:18:26] <jamezp> Will do :-)[20:26:31] *** pil-dinner-bbl is now known as pilhuhn[20:31:06] *** tcrawley has joined #jboss-as7[20:31:06] *** ChanServ sets mode: +v tcrawley[20:37:29] <rmaucher> smcgowan: the JSTL impl that we use can't do anything useful in AS 7[20:43:42] *** lazarotti has quit IRC[20:47:07] <dmlloyd> can't, or doesn't?[20:48:35] <bobmcw> seems to be some issue trying to bind hornetq queues under deep names[20:48:39] <bobmcw> "queue/foo/bar/baz"[20:48:49] <bobmcw> something around how it finds parents, or how MockContext composes names[20:48:59] <bobmcw> sometimes end up with queue/foo/bar/bar/baz[20:49:02] <bobmcw> especially after a redeploy[20:49:17] <bobmcw> I can't see the bug, but I feel it[20:51:18] <Nihility> yuck[20:51:49] <bobmcw> HQ tries to create parent contexts if you hand it a new deep path that doesn't exist[20:51:59] <bobmcw> dunno if the binding traps account for that[20:52:09] <bobmcw> seems like something isn't getting popped off the stack far enough[20:52:22] <bobmcw> so the stack starts "too deep" with a lingering item, for the next push/pop cycle[20:52:44] *** jamezp is now known as jamezp_afk[20:52:48] <Nihility> bobmcw: the good news is that hq is going to switch to using service driven bindings[20:52:49] *** smarlow has quit IRC[20:52:57] <Nihility> bobmcw: although that requires a release[20:53:03] <Nihility> (of hq)[20:54:07] <jbossbot> git [jboss-as] push master 4d9705d.. Jason T. Greene Fix AS-796, don't deploy jdbc drivers as bundles (for now)[20:54:07] <jbossbot> git [jboss-as] push master URL: http://github.com/jbossas/jboss-as/compare/ff4c6fc...4d9705d[20:54:13] <Nihility> jpederse: ^[20:54:31] <jpederse> Nihility: thx[20:55:06] *** smarlow has joined #jboss-as7[20:55:06] *** ChanServ sets mode: +v smarlow[20:55:27] <Nihility> jpederse: btw the conclusion to the call was that we are going to align our DUPs to better fit the OSGi lifecycle[20:55:43] <jpederse> Nihility: sounds good :)[20:55:53] <jpederse> Nihility: damn network crapped out[20:58:58] <dmlloyd> with Nihility's workaround in the interim[21:00:26] <Nihility> bbrowning: hey did you want to do AS7-810, or did you want us to make that change for you[21:00:27] <jbossbot> jira [AS7-810] Change DeploymentRootMountProcessor To Use Marker to Determine Whether to Mount Exploded [Open (Unresolved) Feature Request, Major, Jason Greene] https://issues.jboss.org/browse/AS7-810[21:01:45] *** mlinhard has joined #jboss-as7[21:01:45] *** ChanServ sets mode: +v mlinhard[21:02:11] <bbrowning> Nihility: I have a hacky workaround for TorqueBox at the moment, but I can submit a pull request for the upstream change as well if that's what you're asking :)[21:02:26] <bbrowning> hacky workaround meaning to unmount and remount our .knob file expanded[21:02:42] <Nihility> bbrowning: yeah i wasnt sure if you meant to assign it to me[21:02:46] <dmlloyd> I think it's a good change[21:02:57] <Nihility> bbrowning: i agree with the proposal[21:03:01] <dmlloyd> we should let processors intercede before we mount everything up[21:03:09] <bbrowning> Nihility: I didn't assign it to anyone, just left it blank I thought[21:03:12] <Nihility> bbrowning: i just dont know if you already implemented it[21:03:20] <bbrowning> Nihility: I have not[21:04:10] <Nihility> bbrowning: ok ill stick it in my list of things to do, if however you get impatient ill accept a patch :)[21:04:29] <bbrowning> Nihility: sounds good - thanks![21:04:35] <bobmcw> did that big merge occur?[21:04:40] <bobmcw> should we try it instead of Beta3?[21:04:47] <Nihility> the ejb3 one did[21:04:53] <Nihility> there is a management one coming soon[21:04:56] <bobmcw> if I don't care about ejb3?[21:04:57] <bobmcw> 'k[21:05:02] *** sannegrinovero has joined #jboss-as7[21:05:02] *** sannegrinovero has quit IRC[21:05:02] *** sannegrinovero has joined #jboss-as7[21:05:02] *** ChanServ sets mode: +v sannegrinovero[21:05:16] <Nihility> you would be affected by the management one[21:05:26] <Nihility> as it requires changing operation handler contracts[21:05:28] <bobmcw> is there a URL I can read about that?[21:05:42] <bobmcw> gonna break our FooSubsystemAdds ?[21:05:56] <Nihility> well dmlloyd and bstansberry are trying to port everyones for them[21:06:13] <Nihility> but if yours are not in the tree it wont get ported[21:06:17] <dmlloyd> I meant to be done by now[21:06:19] <dmlloyd> :([21:06:34] <bstansberry> you were dreaming[21:06:54] <dmlloyd> nah just distracted[21:06:58] <Nihility> bobmcw: but the change itself isnt that hard to adapt to, we can give you a hand[21:06:58] *** lazarotti has joined #jboss-as7[21:07:08] <dmlloyd> I still think my total labor estimate is accurate :)[21:07:12] <bstansberry> you were dreaming you wouldn't be distractied[21:07:17] <Nihility> bobmcw: the best URL is probably david's branch[21:07:46] <Nihility> bobmcw: overall its a simplification[21:08:15] <bobmcw> master branch?[21:08:37] <Nihility> https://github.com/dmlloyd/jboss-as/tree/new-controllers-maybe[21:08:40] <dmlloyd> jamezp_afk: when you get a minute, would you mind doing another deprecated method sweep, this time for MSC?[21:08:44] <bobmcw> https://github.com/dmlloyd/jboss-as/blob/new-controllers-maybe/web/src/main/java/org/jboss/as/web/WebSubsystemAdd.java[21:08:49] <dmlloyd> Beta9 is gonna delete a bunch of em[21:09:05] <dmlloyd> bobmcw: I only have one handler ported so far :)[21:09:10] <bobmcw> URL?[21:09:13] <dmlloyd> bobmcw: about 290 to go![21:09:18] <bobmcw> ffs![21:09:24] <Nihility> hahahahahaha[21:09:33] <dmlloyd> to be fair I haven't started yet, I'm still getting the API ready[21:09:41] <Nihility> dmlloyd: feel free to asign a block to me[21:10:02] <dmlloyd> I just about have the service add/remove overlap problem licked here[21:10:19] <Nihility> dmlloyd: can you link that op[21:10:26] <Nihility> i dont see it in your branch[21:10:26] <dmlloyd> remoting/src/main/java/org/jboss/as/remoting/RemotingSubsystemAdd.java[21:11:10] <dmlloyd> sorry org.jboss.as.remoting.NewRemotingSubsystemAdd[21:11:23] <Nihility> oh there it is[21:11:36] <Nihility> https://github.com/dmlloyd/jboss-as/blob/new-controllers-maybe/remoting/src/main/java/org/jboss/as/remoting/NewRemotingSubsystemAdd.java[21:12:18] <bobmcw> NewFoo as in, "create a new foo?" or just it's the new version of Foo?[21:12:47] <Nihility> bobmcw: the latter, the key bits that are different are on line 76 and below[21:13:28] <bobmcw> yeah, where's all the addDeploymentProcessor stuff?[21:13:40] <Nihility> that should all be the same[21:13:47] <Nihility> this one just doesnt happen to do that[21:13:50] <bobmcw> ah, 'k[21:14:06] <dmlloyd> yeah I don't have the hooks on the op context for that yet[21:14:11] <dmlloyd> getting there though...[21:15:36] <Nihility> bobmcw: are you wanting torquebox in CR1?[21:15:53] <bobmcw> Nihility: *in* CR1? No, doubtful[21:16:10] <Nihility> 7.1?[21:16:14] <bobmcw> absolutely[21:16:25] <bobmcw> lemme see how this week goes, maybe 7.0[21:16:32] <Nihility> cool[21:16:33] <bobmcw> how responsive y'all are to our JIRAs[21:16:35] <bobmcw> :)[21:16:49] <Nihility> if you find severe defects[21:16:52] <Nihility> we are pretty fast[21:16:58] <Nihility> mainly due to embarassment[21:16:59] <Nihility> :)[21:18:02] <bobmcw> it's a saved filter, shared with everyone, add it to your dashboard![21:18:08] <Nihility> cool[21:18:11] <Nihility> i was just going to ask for that[21:18:45] <bobmcw> https://issues.jboss.org/secure/IssueNavigator.jspa?mode=hide&requestId=12314573[21:19:00] <bobmcw> maybe easier? named 'TorqueBox AS7 Bugs' if you browse the portlet[21:22:36] *** mgoldmann has quit IRC[21:24:00] <Nihility> how do i add it[21:25:48] <Nihility> never mind[21:40:16] <clebert> Nihility: dmlloyd Hey guys.. quick question... are you guys using @override annotations for interfaces?[21:40:32] <dmlloyd> I'm not, others are[21:40:35] <dmlloyd> we don't have a policy[21:40:55] <clebert> dmlloyd we are having a small discussion within HornetQ if we should do it or not at HornetQ[21:41:05] <clebert> dmlloyd: for me override means overide.. not implements[21:41:07] <dmlloyd> I find @Override more annoying than useful[21:41:47] <Nihility> i use it usually unless its an inner class[21:41:52] <Nihility> then its really damn annoying[21:42:09] <clebert> dmlloyd: Nihility: if you guys can raise the voice here :-) http://community.jboss.org/thread/166830?tstart=0[21:42:16] <clebert> dmlloyd: Nihility if you have 2 minutes only[21:42:34] <Nihility> the main reason i use it though is that my IDE generates it[21:42:38] <Nihility> and im too lazy to turn it off[21:42:43] <clebert> Nihility: :-)[21:42:46] <Nihility> so i actually delete the tags[21:42:49] <Nihility> on inner classes[21:42:53] <Nihility> ok ill say something[21:43:01] <clebert> Nihility: ok.. thanks a lot :-)[21:45:52] <dmlloyd> http://github.com/dmlloyd/jboss-as/tree/new-controllers-maybe[21:45:57] <dmlloyd> updated, rebased, etc.[21:46:04] <dmlloyd> ready for operations to be ported, hopefully[21:46:11] <dmlloyd> at least a trial set of operations[21:46:18] <dmlloyd> we still need some impl business[21:47:02] <dmlloyd> hosts and domains, remoting integration for management protocol, and some other doodads need finishing, including attachments[21:47:09] *** jamezp_afk is now known as jamezp[21:47:24] <dmlloyd> you'll need the latest MSC snapshot though, sorry about that[21:47:24] <jamezp> dmlloyd: Will do on the MSC thing.[21:47:28] <dmlloyd> thanks jamezp[21:49:06] *** smcgowan has quit IRC[21:56:53] *** tdiesler has joined #jboss-as7[21:56:53] *** ChanServ sets mode: +v tdiesler[22:09:00] *** rsvoboda has joined #jboss-as7[22:14:00] *** asoldano_away has quit IRC[22:21:48] *** pferraro has quit IRC[22:31:32] *** rmaucher has quit IRC[22:35:53] *** jfclere has quit IRC[22:43:52] <jamezp> dmlloyd: Maybe I'm crazy, but I'm not finding any deprecated methods or classes in MSC.[22:44:45] <dmlloyd> the stuff in Beta8 that was deprecated has been removed for Beta9[22:44:53] <dmlloyd> hence the need to make sure there's no deprecated usage[22:45:03] <dmlloyd> might just do a build and scan the output for deprecation[22:45:23] <jamezp> Ah, I see :-). No problem.[22:45:53] *** miclorb has quit IRC[22:51:16] *** pilhuhn has quit IRC[22:51:19] <rsvoboda> Hi guys, I'd like to clarify something about as7 management: there are/were planned 3 equivalent ways how to manage AS:[22:51:19] <rsvoboda> 1) CLI 2) Java API 3) Rest API (or in general request/response over http) + Plus GWT based Web UI for more complicated tasks.[22:51:19] <rsvoboda> Is it right? Or what's current plan to support? I tried to search http://community.jboss.org/en/jbossas/dev/jboss_as7_development but without luck.[22:55:01] *** kkhan has joined #jboss-as7[22:55:01] *** ChanServ sets mode: +v kkhan[22:55:30] <rsvoboda> bstansberry , asaldhan, dmlloyd: could you please answer --^ or point me to right person[22:56:04] <bstansberry> rsvoboda: that's the current plan[22:56:09] *** mmoyses has quit IRC[22:57:01] <rsvoboda> bstansberry: thanks, so all I can do in CLI I'll be able to do with Rest API, right?[22:57:52] <bstansberry> rsvoboda: yes and no[22:58:22] <bstansberry> the CLI exposes "low level operations" and then some "high level operations" that make common tasks more convenient[22:59:01] <bstansberry> e.g. "deploy /home/bstansberry/foo.war" is a high level operation; the CLI translates it into low level commands for you[22:59:28] <bstansberry> the HTTP API doesn't have that kind of convenience operation, but you can do all the low level commands[22:59:29] *** mlinhard has quit IRC[22:59:53] <bstansberry> so everything you can do via CLI you can do via HTTP[23:00:05] *** jpederse has quit IRC[23:00:14] <rsvoboda> bstansberry: so I won't be able to deploy war file with rest in one command[23:00:25] <rsvoboda> but with sequence of commands yes ...[23:01:16] <bstansberry> rsvoboda: right[23:02:09] <rsvoboda> bstansberry: you are talking about HTTP API so it won't be based on RESTEasy ?[23:02:45] <rsvoboda> bstansberry: or u just generalize the name of API?[23:02:49] *** misty has joined #jboss-as7[23:03:05] <bstansberry> rsvoboda: right; it's JSON over HTTP but a REST purist would not call it REST, so we stopped calling it REST[23:04:34] <rsvoboda> bstansberry: thank you for answers and your time, I think I'll have more questions in future :)[23:05:05] <bstansberry> great; happy to help[23:05:55] <Nihility> dmlloyd: can you make jbossbot answer that question? :)[23:07:21] *** wolfc has quit IRC[23:07:24] <tdiesler> Nihility, fixed a minor glitch that effectively disabled OSGi all together ;-) https://github.com/tdiesler/jboss-as/commit/19f47dcd1e805577580a7cd03b37e0a04fee53d7[23:07:26] <jbossbot> git [jboss-as] 19f47dc.. Thomas Diesler [AS-796] Fix jdbc driver detection[23:07:46] <Nihility> oops[23:08:16] <Nihility> i wonder how the osgi tests passed[23:09:23] *** rsvoboda has left #jboss-as7[23:09:51] *** kcbabo has quit IRC[23:10:05] <tdiesler> Nihillity, by virtue of @Ignore :-)[23:10:06] <tdiesler> ALR1, aslak, I resolved AS7-773 We can now continue to upgrade that on-demand deployment to the new ARQ version.[23:10:06] <tdiesler> I suggest we do this work in https://github.com/tdiesler/jboss-as/tree/as773 so it woulnt be too disruptive for other folks[23:10:07] <tdiesler> We could merge to master once we have smoke+itegration tests running on the new infrastructure.[23:10:07] <tdiesler> Note, I also migrated all smoke tests (i.e. the once I could) to the managed container. I suggest we first focus on that one[23:10:09] <jbossbot> jira [AS7-773] Replace ARQ subsystem by on-demand deployment [Resolved (Done) Task, Major, Thomas Diesler] https://issues.jboss.org/browse/AS7-773[23:10:57] <jbossbot> git [jboss-as] push master 8bd6472.. Thomas Diesler [AS-796] Fix jdbc driver detection[23:10:57] <jbossbot> git [jboss-as] push master URL: http://github.com/jbossas/jboss-as/compare/4d9705d...8bd6472[23:11:06] <Nihility> tdiesler: thanks, stupid mistake[23:11:58] <Nihility> ah yes[23:12:16] <Nihility> nothing like disabling all our tests[23:12:23] <Nihility> to let regressions through[23:13:53] <ALR1> Woohoo[23:14:01] <tdiesler> I'm tracking this now AS7-814[23:14:03] <jbossbot> jira [AS7-814] Fix or remove ignored smoke tests [Open (Unresolved) Task, Major, Jason Greene] https://issues.jboss.org/browse/AS7-814[23:14:24] <ALR1> Nihility: My SW/ARQ tasks are presently done[23:14:27] <ALR1> With these releases[23:14:33] <ALR1> Today I switched back to AS7 only.[23:14:44] <ALR1> Looking at the testsuites, I see only some are failing now[23:14:53] <ALR1> Was expecting the full integration suites to be failing.[23:14:58] <Nihility> ALR1: cool what happened to the original ALR[23:15:16] <ALR1> The VPN killed him[23:15:17] <ALR1> I think[23:15:39] <ALR1> So where are the failing integration tests at the moment?[23:15:51] <ALR1> (Not the OpenJDK thing)[23:15:52] <ALR1> https://hudson.qa.jboss.com/hudson/view/JBoss%20AS/job/JBoss-AS-7.0.x-testSuite-openjdk/[23:16:06] <ALR1> I think I'll have a window of less than 2 weeks to resolve all these things.[23:16:10] <ALR1> ARQ Beta due out tomorrow.[23:16:21] <ALR1> But the prereq work looks finally cleared now[23:16:28] <Nihility> this is is all in tdiesler 's branch[23:16:39] <Nihility> https://github.com/tdiesler/jboss-as/tree/as773[23:17:17] <ALR1> Nihility: I mean the existing integration test failures?[23:17:34] <ALR1> That were present before JUDCon? Due to POM/dependency problems, if I remember correctly.[23:18:10] <Nihility> AFAIK they are all passing[23:18:33] <Nihility> all 457 of them![23:19:25] <ALR1> OK. As tdiesler noted in the last conf minutes, the 2nd stage after this ARQ release is to make 1 ARQ subsystem in AS.[23:19:34] <ALR1> And I port all over to match, including the new Embedded container.[23:19:46] <ALR1> New ARQ is slated for tomorrow.[23:19:53] <Nihility> ALR1: tdiesler what is this on-demand deployment?[23:20:18] <Nihility> ALR1: tdiesler, do you mean arq itself is a deployment[23:20:23] <ALR1> The AS Arquillian connector service.[23:20:30] <ALR1> For communication w/ ARQ[23:20:37] <Nihility> oh ok[23:20:43] <ALR1> https://github.com/tdiesler/jboss-as/commit/7a3bdff3ff32fa541d446838d1539f4853949bf1[23:20:44] <jbossbot> git [jboss-as] 7a3bdff.. Thomas Diesler [AS7-773] Deploy arquillian service on demand[23:20:45] <jbossbot> jira [AS7-773] Replace ARQ subsystem by on-demand deployment [Resolved (Done) Task, Major, Thomas Diesler] https://issues.jboss.org/browse/AS7-773[23:23:03] <tdiesler> Nihility,ALR1, yes the ARQ subsystem is gone. All containers (i.e. embedded, managed, remote) deploy the ARQ server side connector on server startup[23:23:07] *** spagop has joined #jboss-as7[23:23:25] <jamezp> dmlloyd: Looks like nothing in the current branch of AS7 is using any deprecated stuff in MSC. The build worked with with Beta9.[23:23:43] <ALR1> tdiesler: Supernice stuff.[23:24:19] <tdiesler> In this way we have more flexibility what connector is actually being deployed[23:24:21] <dmlloyd> jamezp: great.[23:24:40] <jamezp> Those are the easy fixes for me :-)[23:25:08] <aslak> tdiesler, we could move that out as a extension, and not hardcode it in the AS containers.[23:25:27] <tdiesler> aslak, yes[23:26:32] <tdiesler> aslak, for now its more or less a deployable replacement of what was there. Next would be the actual ARQ update[23:27:03] *** spagop has quit IRC[23:27:20] <tdiesler> I suggest we get everything going on the new ARQ version with the managed container. Embedded, Remote can be done later[23:29:41] <Nihility> long term thought[23:30:13] <Nihility> if embedded was working, and we wanted to explore parallelizing the testsuite, we could reuse the JVM[23:30:52] <Nihility> you guys should see how ridiculously fast you can cycle an AS instance by calling stop/start on the root service[23:32:00] <Nihility> 16:33:50,706 INFO [org.jboss.as] (MSC service thread 1-4) JBoss AS 7.0.0.Beta4-SNAPSHOT "(TBD)" started in 298ms - Started 106 of 160 services (54 services are passive or on-demand)[23:32:17] <Nihility> thats a stop/start cycle[23:32:27] <Nihility> the difference is purely classloading/initialization[23:33:35] <Nihility> although maybe its possible to do that with managed too[23:33:51] <Nihility> one JVM per concurrent test thread[23:35:02] <tdiesler> I opted for managed because it gives isolation out of the box. When we migrated successfully, we should revisit embedded[23:35:20] <Nihility> yeah im just talking long term[23:35:31] *** mbg has quit IRC[23:37:10] <ALR1> Wow.[23:37:37] <Nihility> if you want to see it in action[23:37:41] <Nihility> ./jboss-admin.sh[23:37:42] <Nihility> connect[23:37:45] <Nihility> :reload[23:38:33] <jbossbot> git [jboss-as] push master c058e9a.. Jason T. Greene Revert "Make Drivers work from DriverManager, for those who want it"...[23:38:33] <jbossbot> git [jboss-as] push master URL: http://github.com/jbossas/jboss-as/compare/8bd6472...c058e9a[23:39:16] <Nihility> dmlloyd: i reverted that idiots patch![23:39:16] <Nihility> hahaha[23:39:16] <dmlloyd> good, we shouldn't accept any more patches from that twit[23:39:49] <Nihility> well at least your patch wasnt as stupid as mine[23:39:55] <Nihility> apparently i cant use VFS[23:41:04] *** kkhan has quit IRC[23:41:23] <dmlloyd> speaking of VFS, it could use a bit of updating[23:41:31] <dmlloyd> get rid of that dumb warning for one thing[23:42:10] <Nihility> yeah i meant to clean that up[23:42:22] <Nihility> im starting a huge list of shit we need to fix thats annoying[23:42:25] <dmlloyd> maybe do a GA release...[23:43:04] *** tdiesler has quit IRC[23:45:24] <Nihility> oh yeah[23:45:27] <Nihility> lets do one of those[23:45:46] <Nihility> were passing TCK now right? :)[23:48:44] <dmlloyd> of VFS I mean![23:49:00] <dmlloyd> it's been in CR since like 2009[23:58:37] *** miclorb has joined #jboss-as7