May 12, 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 12, 2011 NOTICE: This channel is no longer actively logged.
[00:00:16] *** stansilvert has quit IRC[00:00:21] *** misty has quit IRC[00:00:33] <dmlloyd> jamezp: yeah just check for SYSTEM, then IDENTIFIER, then delegate and that's it[00:01:26] <jamezp> dmlloyd: Okay. I'll give it a go.[00:01:31] <jamezp> Thanks.[00:02:55] <dmlloyd> okay ping me when you get something[00:03:04] <dmlloyd> hopefully we can get some functional TCK runs once this is ready[00:03:15] <jamezp> Will do.[00:06:50] *** misty has joined #jboss-as7[00:14:53] <jamezp> dmlloyd: Does this look a little better? https://gist.github.com/967513[00:15:31] <jamezp> I couldn't use the System.getenv().get("CLASSPATH"). I kept getting errors because it had paths like ../lib/[00:16:39] <dmlloyd> hmm interesting[00:17:36] <jamezp> It could be the way I'm creating the java.io.File.[00:17:51] <dmlloyd> well since we're launching jboss-modules.jar via -jar, java.class.path is generally going to be empty when we start[00:19:16] <dmlloyd> where are you getting that[00:19:20] <dmlloyd> I get null for CLASSPATH[00:19:27] <dmlloyd> is it set in your env for other reasons?[00:19:34] <dmlloyd> e.g. echo $CLASSPATH[00:19:49] *** jdcasey has quit IRC[00:19:52] <dmlloyd> you had me thinking that maybe the JVM was setting the env var fo some reason[00:20:00] <dmlloyd> but I think it must be something, well, environmental[00:20:04] <jamezp> I wonder if it's IDEA. Let me try a build outside the IDE.[00:21:28] <jamezp> Yeah, works find if I build from the command line.[00:21:48] <jamezp> IDEA must be setting up something in the class path.[00:22:02] <dmlloyd> how are you launching it in IDEA? just running the main() method?[00:22:19] <dmlloyd> it should still work really[00:22:34] <dmlloyd> unless they put rt.jar into the classpath again or something like that[00:22:36] <jamezp> I created a maven run configuration.[00:22:59] <jamezp> I get "File ../lib/bootstrap.jar in class path not valid."[00:23:18] <dmlloyd> okay so you need to resolve those against user.dir[00:23:20] * jamezp is new to IDEA so probably doesn't get it :-)[00:24:14] <dmlloyd> well the classpath entry is valid[00:24:18] <dmlloyd> it's just a relative path[00:24:26] <dmlloyd> so you need to resolve it[00:24:28] <dmlloyd> I mean it should work[00:26:15] <jamezp> Okay, so if I have a relative path I should try to resolve it.[00:27:54] *** irooskov has quit IRC[00:28:00] *** smcgowan is now known as smcgowan_dinner[00:28:48] <dmlloyd> yeah against the cwd at boot[00:29:10] <dmlloyd> it could change at runtime - if it does then we should continue using the old paths[00:29:44] <jamezp> Sounds good. Writing it up now.[00:31:46] *** irooskov has joined #jboss-as7[00:31:47] *** jamezp has quit IRC[00:32:51] *** asoldano_away is now known as asoldano[00:33:55] <asoldano> dmlloyd, hi! time for a question on possible classloading bug in AS7 ? (or at least different behaviour compared to AS6)[00:34:03] *** jamezp has joined #jboss-as7[00:34:20] <dmlloyd> sure go ahead[00:34:35] <asoldano> dmlloyd, it's basically https://issues.jboss.org/browse/JBWS-3227[00:34:37] <jbossbot> jira [JBWS-3227] handlers configuration file not found on classpath [Open (Unresolved) Sub-task, Major, Unassigned] https://issues.jboss.org/browse/JBWS-3227[00:34:47] <asoldano> dmlloyd, I have an endpoint that is @HandlerChain annotated[00:35:22] <asoldano> dmlloyd, the file attribute in the annotation points to a file that lives in WEB-INF/classes[00:36:02] <asoldano> dmlloyd, the lookup is done by CXF using Class::getResource(String s)[00:36:17] <asoldano> dmlloyd, lookup fails on AS7, works on AS6[00:36:21] *** clebert has joined #jboss-as7[00:36:21] *** ChanServ sets mode: +v clebert[00:39:28] *** frainone is now known as frainone_away[00:39:37] <dmlloyd> ok first what is @HandlerChain :)[00:40:12] <asoldano> dmlloyd, it's annotation to be used to tell the stack that a given handler descriptor needs to be parsed for knowing what jaxws handlers to add to the endpoint[00:40:57] <dmlloyd> okay I don't see any reason why it would fail[00:41:01] <dmlloyd> is it using TCCL?[00:41:21] <asoldano> let me do a quick check with debugger[00:42:10] <asoldano> dmlloyd, oh, well, it uses the classloader that Class impl gets in getResource[00:42:29] <dmlloyd> so it's a deployment Class though?[00:42:47] <asoldano> I think so[00:43:22] <dmlloyd> I'm pretty sure that WEB-INF/classes is a resource root[00:43:27] <dmlloyd> what path is it trying to load?[00:43:36] <asoldano> dmlloyd, yes, confirmed, it's a Modular Classloader, the one for the deployment[00:43:56] <asoldano> dmlloyd, it is trying to load: org/jboss/test/ws/jaxws/jbws3034/../../../../../../handlers.xml[00:44:02] *** Nihility has quit IRC[00:44:10] <asoldano> dmlloyd, which is actually "handlers.xml" in WEB-INF/classes[00:44:58] <dmlloyd> okay is it loading /handlers.xml or is it loading handlers.xml?[00:45:21] <dmlloyd> if the latter, and you're using Class.getResource*(), then it's going to look for it in the package dir and not find it[00:46:17] <asoldano> dmlloyd, well, the annotation says " at HandlerChain(file=" dot ./../../../../../handlers.xml")"[00:46:40] <asoldano> dmlloyd, then CXF does "clz.getResource(filename)"[00:46:53] <asoldano> dmlloyd, with filename = "../../../../../../handlers.xml"[00:47:09] <asoldano> dmlloyd, and the rest is up to Class::getResource impl[00:47:12] <dmlloyd> crikey[00:47:26] <dmlloyd> well maybe we're not handling ..[00:47:33] *** jwulf has quit IRC[00:47:47] <dmlloyd> if you have the debugger up, can you see what the path is when it hits ModularClassLoader?[00:48:15] *** kcbabo has joined #jboss-as7[00:48:15] *** ChanServ sets mode: +v kcbabo[00:48:30] <dmlloyd> or is that the path you saw[00:48:40] <asoldano> dmlloyd, when hitting ConcurrentClassLoader:getResource(final String name), name= org/jboss/test/ws/jaxws/jbws3034/../../../../../../handlers.xml[00:48:48] <dmlloyd> hmm, okay[00:48:52] <dmlloyd> yeah I think that's a bug[00:49:02] <dmlloyd> I'm 90% sure we don't handle ..[00:49:12] <asoldano> ok..[00:49:52] <asoldano> I think we do have other @HandlerChain tests, but they all use things like file="handlers.xml" except for this complex one[00:50:03] <asoldano> and this is the only one failing[00:51:26] <asoldano> dmlloyd, do you want me to fill in another jira for this?[00:52:00] *** miclorb has joined #jboss-as7[00:52:05] <dmlloyd> yeah file a JIRA in MODULES for it[00:52:12] <asoldano> dmlloyd, ok, thanks[00:52:59] *** clebert has quit IRC[00:57:45] <asoldano> dmlloyd, MODULES-86 created, thanks[00:57:46] <jbossbot> jira [MODULES-86] ModuleClassLoader::getResource(String s) does not handle path containing ".." [Open (Unresolved) Bug, Major, David Lloyd] https://issues.jboss.org/browse/MODULES-86[00:58:19] *** jpearlin has joined #jboss-as7[00:58:55] *** asoldano has quit IRC[01:01:46] *** bstansberry has quit IRC[01:17:17] <jamezp> dmlloyd: Not having a lot of luck getting the proper path. At least in IDEA.[01:17:51] <jamezp> Or proper path to the file from the IDEA class path.[01:18:01] *** clebert has joined #jboss-as7[01:18:01] *** ChanServ sets mode: +v clebert[01:20:28] *** clebert has quit IRC[01:27:28] *** jwulf has joined #jboss-as7[01:41:43] <dmlloyd> jamezp, the directory should be resolved against ${user.dir}[01:41:51] <dmlloyd> System.getProperty("user.dir")[01:42:27] <jamezp> Yeah, tried that. Didn't work for IDEA, but worked when just testing with random strings.[01:42:39] <dmlloyd> didn't work how?[01:42:48] <dmlloyd> I mean it's not an error if they list a dir/jar that doesn't exist[01:43:16] <jamezp> Ah, got it! Sorry now I understand.[01:44:30] <jamezp> Basically I shouldn't run the Module.relink() in my test then, correct?[01:44:45] <dmlloyd> there's really no need to[01:44:51] <dmlloyd> relink is only used in special cases[01:45:12] <jamezp> I only did because I tried giving it an invalid dependency and my test still passed.[01:45:37] <jamezp> And that was the only way it failed.[01:45:55] <dmlloyd> probably because whatever your test was doing, it didn't actually load the module and try to use it[01:46:13] *** maeste has quit IRC[01:46:19] <jamezp> Probably because I'm just winging it :-)[01:47:22] <jamezp> Anyway, bottom line is at load time we don't care if the file/directory exists at the moment or not.[01:54:17] *** emuckenhuber has quit IRC[01:59:16] *** alexsmirnov has quit IRC[02:06:40] *** bstansberry has joined #jboss-as7[02:06:40] *** ChanServ sets mode: +v bstansberry[02:07:24] *** bstansberry is now known as bstans_afk[02:09:12] <jamezp> dmlloyd: The classpath is winning. It's working until I hit /usr/java/default/lib/tools.jar then I get a java.util.zip.ZipException: error in opening zip file.[02:09:53] <dmlloyd> what's the error?[02:10:19] <dmlloyd> can you do e.g. "jar tf /usr/java/default/lib/tools.jar" from the commandline?[02:10:35] <jamezp> Yes indeed.[02:11:03] <dmlloyd> there must be something funny about the File[02:11:31] <dmlloyd> in any case, bad JARs should also be skipped afaict[02:11:37] <jamezp> Should I maybe not be using ResourceLoaders.createJarResourceLoader(root.getCanonicalFile().getParent(), new JarFile(root.getName())) for files?[02:12:02] <dmlloyd> hmm yeah don't use getCanonicalFile()[02:12:08] <jamezp> Okay, my issue is in the new JarFile() constructor.[02:12:25] <jamezp> Oh, yeah I was just testing different options. Forgot to remove it.[02:12:52] <dmlloyd> well root.getName() won't work[02:13:00] <dmlloyd> it's just the last segment[02:13:15] <dmlloyd> you'd want getPath(), maybe[02:13:26] <jamezp> Oh, good point. I was getting an NPE without it.[02:13:45] <jamezp> So my problem is probably me not the tools.jar :-)[02:14:13] *** bstans_afk has quit IRC[02:14:54] <jamezp> Somewhere I have a null ModuleIdentifier coming in.[02:27:27] *** sgilda has quit IRC[02:30:27] *** frainone_away is now known as frainone[02:30:29] *** bbrowning has quit IRC[02:33:53] *** miclorb has quit IRC[02:40:33] <jamezp> dmlloyd: I've got to run for now, but it kind of works :-) https://github.com/jamezp/jboss-modules/commit/ea3a8b1ac17bc93c86eef89a333e1c6eb68d0aed[02:40:35] <jbossbot> git [jboss-modules] ea3a8b1.. James Perkins MODULES-85 Add classpath module loader and CLI processing.[02:40:36] <jbossbot> jira [MODULES-85] Add support for class path bootstrap [Open (Unresolved) Enhancement, Major, James Perkins] https://issues.jboss.org/browse/MODULES-85[02:41:13] <jamezp> It's broken when testing from IDEA, but works fine from the command line and there is no classpath.[02:41:57] <jamezp> I know what the line the NPE is happening on, but I'm not sure why yet.[02:42:00] <dmlloyd> okay[02:43:25] *** jamezp is now known as jamezp_afk[02:47:17] *** maxandersen has quit IRC[02:52:10] *** bgeorges has quit IRC[02:53:30] *** jwulf has quit IRC[03:09:08] *** bstansberry has joined #jboss-as7[03:09:08] *** ChanServ sets mode: +v bstansberry[03:09:42] *** miclorb has joined #jboss-as7[03:11:37] *** lgao has joined #jboss-as7[03:13:28] *** pgier has quit IRC[03:34:49] *** stliu has joined #jboss-as7[03:34:49] *** ChanServ sets mode: +v stliu[03:36:29] *** Nihility has joined #jboss-as7[03:36:29] *** Nihility has joined #jboss-as7[03:36:29] *** ChanServ sets mode: +v Nihility[03:37:01] <dmlloyd> jamezp_afk: okay I did a couple fixes and merged it.[03:37:14] <dmlloyd> jamezp_afk: http://github.com/jbossas/jboss-modules/compare/c71b5d1...b2e7860 for the final breakdown.[03:37:30] <dmlloyd> that NPE you found was a bug in the file resource loader[03:39:57] <dmlloyd> not all the tests pass in IDEA but it's because of the system module[03:40:11] <dmlloyd> it's a bit fuzzy about deciding whether to use its parent CL or the system CL for various things[03:40:18] <dmlloyd> most of the time it's the same but sometimes not[03:45:46] <jamezp_afk> dmlloyd: Sweet! Thanks. I'll pull down the changes and do some testing.[03:53:32] *** asaldhan has left #jboss-as7[03:53:58] *** stliu has quit IRC[03:56:19] *** jpearlin has left #jboss-as7[04:02:26] *** torben_ has joined #jboss-as7[04:05:50] *** torben has quit IRC[04:09:03] *** JimMa has joined #jboss-as7[04:17:04] *** torben_ has quit IRC[04:17:29] *** jwulf has joined #jboss-as7[04:21:45] <smarlow> Nihility: the downside of implementing the default Hibernate 2LC based on an BCHM, is that we will support two cache implementations. The upside is that BCHM is probably leaner and possibly faster. I put my "make Infinispan default 2lc" patch on a branch for now. I'm thinking we should bring Infinispan back into standalone (to resolve the failing integration tests), switch to Hibernate 4.0.alpha1 and integrate a BCHM backed 2lc service that[04:21:45] <smarlow> will live in AS7.[04:22:03] *** miclorb has quit IRC[04:22:18] *** stliu has joined #jboss-as7[04:22:18] *** ChanServ sets mode: +v stliu[04:24:08] *** bgeorges has joined #jboss-as7[04:24:08] *** ChanServ sets mode: +v bgeorges[04:33:19] <stuartdouglas> dmlloyd: I added binding description merging on a module level https://github.com/stuartwdouglas/jboss-as/commit/ee68d206c9581aa339dbeccaf775b1afd18e03e0[04:33:21] <jbossbot> git [jboss-as] ee68d20.. Stuart Douglas Add merging of JNDI bindings at a module level[04:33:43] <stuartdouglas> java:app and java:global duplicates will have to be handled by your NewNamingStore thing[04:33:57] <stuartdouglas> but this is enough to get the tests to pass[04:34:17] <dmlloyd> good 'nuff[04:34:39] <dmlloyd> I'll push it up and maybe shelly is still awake and can fire off a new run...[04:34:46] <stuartdouglas> three failures to go[04:35:15] <stuartdouglas> looking at that run it looks like it was actually against master[04:35:31] <stuartdouglas> as there are classes in that server log stack trace that don't exist in the ejb branch[04:37:16] <dmlloyd> can you rebase?[04:37:23] <stuartdouglas> against master?[04:37:29] <dmlloyd> against my ejb[04:37:36] <dmlloyd> we seem to have diverged...[04:37:40] <dmlloyd> I can't merge your branch[04:37:48] <stuartdouglas> ok, rebasing now[04:39:45] *** smarlow has quit IRC[04:42:35] <stuartdouglas> It's rebased, I am just doing a build now to make sure nothing broke[04:44:13] <dmlloyd> okay cool[04:44:35] <stuartdouglas> the weld TCK is also down to 18 failures[04:44:52] <stuartdouglas> most of which are because we have not implemented ejb removal yet[04:45:46] <stuartdouglas> that rebase does not appear to have broken anything[04:45:53] <dmlloyd> great[04:46:02] *** smcgowan_dinner is now known as smcgowan[05:07:44] <jbossbot> git [jboss-invocation] push master eecb524.. David M. Lloyd Prep 1.0.0.Beta5[05:07:45] <jbossbot> git [jboss-invocation] push master c42af60.. David M. Lloyd Next is Beta6[05:07:45] <jbossbot> git [jboss-invocation] push master URL: http://github.com/jbossas/jboss-invocation/compare/f903792...c42af60[05:07:45] <jbossbot> git [jboss-invocation] push 1.0.0.Beta5 URL: http://github.com/jbossas/jboss-invocation/compare/0000000...ca83978[05:09:06] <jbossbot> git [jboss-as] push master 0aa6264.. Carlo de Wolf Ignore hprof dumps[05:09:06] <jbossbot> git [jboss-as] push master c0f9b25.. Carlo de Wolf JBAS-9020: rename DeploymentRepository to ContentRepository and create a 1:N relationship between deployment and content[05:09:07] <jbossbot> jira [JBAS-9020] Redirected to: https://issues.jboss.org/si/jira.issueviews:issue-xml/AS7-431/AS7-431.xml[05:09:08] <jbossbot> jira [AS7-431] Deployment content management enhancement [Open (Unresolved) Task, Major, Carlo de Wolf] https://issues.jboss.org/browse/AS7-431[05:09:08] <jbossbot> git [jboss-as] push master a335f5f.. Carlo de Wolf jboss.server.deployment.scanner.default was silently failing in embedded tests[05:09:08] <jbossbot> git [jboss-as] push master feecd19.. Carlo de Wolf JBAS-9020: remove usage of dmr from runtime operations and make the content come in via a service[05:09:09] <jbossbot> git [jboss-as] push master 9a2ba17.. Carlo de Wolf JBAS-9020: fix writing back standalone.xml[05:09:09] <jbossbot> git [jboss-as] push master d441ece.. Carlo de Wolf JBAS-9020: allow unmanaged content items[05:09:09] <jbossbot> git [jboss-as] push master ae2bd25.. Carlo de Wolf AS7-522: initial welcome app[05:09:09] <jbossbot> jira [AS7-522] Create a "jboss-welcome.war" greeting page application [Open (Unresolved) Task, Major, Unassigned] https://issues.jboss.org/browse/AS7-522[05:09:10] <jbossbot> git [jboss-as] push master bd0afbc.. Carlo de Wolf AS7-431: update welcome content to Beta4[05:09:10] <jbossbot> git [jboss-as] push master 534c956.. Carlo de Wolf AS7-431: unwarred welcome.war and handling exploded unmanaged content[05:09:11] <jbossbot> git [jboss-as] push master ebd3403.. Carlo de Wolf AS7-431: content add operation handling for the domain[05:09:11] <jbossbot> git [jboss-as] push master fd2d9bc.. Carlo de Wolf AS7-431: fixed full-replace and make sure the PathContentServitor service is removed[05:09:12] <jbossbot> git [jboss-as] push master 13948fb.. Carlo de Wolf AS7-431: fixed full-replace for domain[05:09:12] <jbossbot> git [jboss-as] push master 81a24f0.. bstansberry at jboss dot com [AS7-431] Create the correct operation structure for a domain mode deployment add...[05:09:13] <jbossbot> git [jboss-as] push master ed39c25.. bstansberry at jboss dot com [AS7-522] Use system-content instead of content...[05:09:13] <jbossbot> git [jboss-as] push master 442f752.. bstansberry at jboss dot com [AS7-431] Fix deployment content paths when FS scanner has a relative-to...[05:09:14] <jbossbot> git [jboss-as] push master URL: http://github.com/jbossas/jboss-as/compare/e58e55d...442f752[05:24:12] <dmlloyd> stuartdouglas: you around?[05:24:19] <stuartdouglas> yes[05:24:24] <dmlloyd> cool[05:24:32] <dmlloyd> what do you make of this: https://hudson.qa.jboss.com/hudson/view/All/job/JBoss-AS-7.0.x-EJB3/lastCompletedBuild/testReport/org.jboss.as.arquillian.container.managed/IntegrationTestCase/org_jboss_as_arquillian_container_managed_IntegrationTestCase/[05:24:40] <dmlloyd> have you seen this?[05:24:43] <dmlloyd> it hasn't happened to me[05:25:06] <stuartdouglas> thats a new one[05:25:35] <dmlloyd> looks like the url handler wasn't installed or something[05:25:51] <stuartdouglas> yea, which is odd[05:26:11] <dmlloyd> I tagged moduels and invocation, we have no other snapshots...[05:26:31] <dmlloyd> was just one test too[05:27:10] *** irooskov has quit IRC[05:27:51] * dmlloyd reruns[05:28:05] <dmlloyd> maybe I messed up URL handlers just enough...[05:30:00] *** liweinan has joined #jboss-as7[05:30:35] <smcgowan> i'm going to rerun it that job too,unless you guys are looking at it. the test looked familiar as it fails on the Solaris buids; http://hudson.jboss.org/hudson/job/JBoss-AS-7.0.x-Solaris/70/testReport/[05:32:36] <dmlloyd> ah, I was briefly hopeful that it was the same error, but unfortunately not[05:32:52] <dmlloyd> anyway that solaris error is interesting[05:34:05] <smcgowan> yes it's different root cause[05:34:34] *** frainone has quit IRC[05:34:37] <dmlloyd> go ahead and rerun it, I'll try to figure out what happened in the meantime[05:39:28] <smcgowan> still failing[05:39:36] <dmlloyd> shit[05:39:47] <dmlloyd> just the one test?[05:39:52] <stuartdouglas> thats really weird[05:40:09] <dmlloyd> exact same error[05:40:28] <dmlloyd> oh[05:40:29] <dmlloyd> ooooooooh[05:40:32] <dmlloyd> I know the problem![05:40:56] <stuartdouglas> what is it?[05:43:01] <dmlloyd> MODULES-44[05:43:03] <jbossbot> jira [MODULES-44] Provide an API that allows adding of URLStreamHandlerFactories [Resolved (Done) Feature Request, Major, David Lloyd] https://issues.jboss.org/browse/MODULES-44[05:43:12] <dmlloyd> changed the behavior of interpreting the sys prop[05:43:19] <dmlloyd> it only reads it once at boot[05:43:53] <dmlloyd> fix is use the programmatic API to register the module[05:44:10] <dmlloyd> (as soon as IDEA finishes importing the major branch change I just did...)[05:44:47] <dmlloyd> as to why only the one test is affected... I can only speculate[05:45:13] <smcgowan> dmlloyd: i think there may be another problem, i downloaded the zip to test a quick EJbEmbed test, once I configured the TCK, staring the server fails: expected attribute 'cache-type' encountered[05:45:22] <stuartdouglas> maybe because that is the first test? and that fails the build?[05:45:44] <smcgowan> that is used in the new security domain: <security-domain name="other" cache-type="default">[05:47:09] <smcgowan> stuartdouglas, dmlloyd: each test suite is yielding different results depending on the platform they are run on. windows build i think is most accurate: http://hudson.jboss.org/hudson/job/JBoss-AS-7.0.x-testSuite-windows/6/testReport/[05:47:38] <dmlloyd> just testing a fix now[05:47:46] <smcgowan> that should be: Message: Unexpected attribute 'cache-type' encountered[05:48:19] <stuartdouglas> smcgowan: When they are re-run do they all get the same fails each time?[05:48:55] <stuartdouglas> because there is a bug that can cause the server to think the deployment is complete and start the test prematurly[05:49:15] <stuartdouglas> but it happens randomly, and some computers seem more likely to trigger it than others[05:49:29] <dmlloyd> I know that the security guys were making changes to the security cache config[05:49:49] <smcgowan> stuartdouglas: that bug is what we see when run on Linux: http://hudson.jboss.org/hudson/job/JBoss-AS-7.0.x-testSuite-sun16/lastCompletedBuild/testReport/[05:50:04] <smcgowan> but on windows the errors are not initialization errors[05:51:19] <smcgowan> dmlloyd: regarding security, your branch is missing the security updates, is that correct?[05:51:51] <dmlloyd> yeah it hasn't been rebased in a few days...[05:52:05] <dmlloyd> and now *all* the tests fail... ah, because I forgot about embedded bootstrap...[05:52:51] <dmlloyd> these tests take so long to run[05:52:54] <dmlloyd> I need an SSD or something[05:56:44] *** bgeorges has quit IRC[05:57:23] *** magesh has joined #jboss-as7[06:00:07] <dmlloyd> looking better...[06:00:44] <dmlloyd> ok embedded works[06:00:48] <dmlloyd> surefire hangs :|[06:01:10] *** bstansberry has quit IRC[06:01:24] *** nickarls has quit IRC[06:02:25] <dmlloyd> what API do those tests use to create the server?[06:02:39] <dmlloyd> I added the VFS module registration to the embedded API and the main classes[06:03:48] *** nickarls has joined #jboss-as7[06:04:53] <stuartdouglas> I think they are run be the arq embedded container[06:05:03] <stuartdouglas> which starts the server with EmbeddedServerFactory.create([06:05:53] <dmlloyd> hmm[06:06:25] <dmlloyd> I updated that class and that fixed the embedded cases[06:06:39] <dmlloyd> but not the org.jboss.as.test.surefire ones[06:07:56] <stuartdouglas> that may require changes to the surefire plugin[06:08:28] <dmlloyd> I'll set it up in the server BootstrapImpl and see if that helps[06:09:53] <dmlloyd> okay that fixed it, whew[06:10:11] *** magesh has quit IRC[06:10:16] <dmlloyd> smcgowan: pushed![06:10:28] <dmlloyd> all tests are passing for me now[06:10:31] <smcgowan> ok[06:10:54] *** magesh has joined #jboss-as7[06:10:56] <smcgowan> we'll try again[06:11:57] *** miclorb_ has joined #jboss-as7[06:13:10] <smcgowan> of course the hudson queue just filled up, all jobs must be set to daily which kicks off at midnight[06:14:18] <dmlloyd> blah, figures[06:14:33] <dmlloyd> wish I would have realized that this problem would occur[06:16:11] <smcgowan> and i can workaround the security 'cache-type' for this test run[06:16:26] <dmlloyd> okay great[06:16:35] *** liweinan has quit IRC[06:16:37] <dmlloyd> next step would be rebasing against upstream, I guess[06:17:11] *** liweinan has joined #jboss-as7[06:17:16] *** Nihility has quit IRC[06:17:31] <stuartdouglas> the ejb branch? I can do that if you want[06:23:21] *** bgeorges has joined #jboss-as7[06:23:21] *** ChanServ sets mode: +v bgeorges[06:26:28] <smcgowan> dmlloyd: the job completed, all tests passed, i'm just going to test the zip quickly before i do the full test run[06:26:38] <dmlloyd> cool![06:27:56] <dmlloyd> stuartdouglas: oh I already am underway doing it[06:28:02] <dmlloyd> I'll let you do the next one though :)[06:28:20] <stuartdouglas> fun :-)[06:28:40] <dmlloyd> this branch has a friggin lot of commits[06:29:01] <stuartdouglas> yea, the last rebase was not fun[06:29:03] <dmlloyd> the merge will be accompanied by a mushroom cloud[06:31:10] <dmlloyd> okay rebase was a success, but I think smcgowan should continue with the current stuff since it's probably a bit safer :)[06:31:18] <dmlloyd> better than trusting my rebase skills![06:32:06] <smcgowan> dmlloyd: you want me to stay with what I have or do another build with the rebase?[06:32:20] <dmlloyd> stay with what you have[06:32:27] <dmlloyd> next build can be with the rebase[06:34:07] <smcgowan> sounds good[06:37:48] <bgeorges> dmlloyd: smcgowan I got an error while runing the test on latest upstream[06:38:05] <bgeorges> The tets throw an exception but does not fail[06:38:20] <smcgowan> bgeorges: can you paste the error[06:38:41] <bgeorges> T E S T S[06:38:42] <bgeorges> Running org.jboss.as.ejb3.component.EJBComponentDescriptionTestCase[06:38:44] <bgeorges> Tests run: 1, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 0.559 sec[06:38:47] <bgeorges> Running org.jboss.as.ejb3.component.stateful.StatefulSessionSynchronizationInterceptorTestCase[06:38:48] <bgeorges> javax.ejb.EJBException: EJB 3.1 FR 4.6 Stateful instance Mock for StatefulSessionComponentInstance, hashCode: 1113757163 is already associated with tx TX1 (current tx TX2)[06:38:50] <bgeorges> at org.jboss.as.ejb3.component.stateful.StatefulSessionSynchronizationInterceptor.processInvocation(StatefulSessionSynchronizationInterceptor.java:60)[06:38:52] <bgeorges> at org.jboss.as.ejb3.component.stateful.StatefulSessionSynchronizationInterceptorTestCase.testConcurrentTx(StatefulSessionSynchronizationInterceptorTestCase.java:118)[06:38:54] <bgeorges> at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)[06:38:56] <bgeorges> at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:39)[06:38:58] <bgeorges> at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:25)[06:39:00] <bgeorges> at java.lang.reflect.Method.invoke(Method.java:597)[06:39:02] <bgeorges> at org.junit.runners.model.FrameworkMethod$1.runReflectiveCall(FrameworkMethod.java:44)[06:39:04] <bgeorges> at org.junit.internal.runners.model.ReflectiveCallable.run(ReflectiveCallable.java:15)[06:39:06] <bgeorges> at org.junit.runners.model.FrameworkMethod.invokeExplosively(FrameworkMethod.java:41)[06:39:08] <bgeorges> at org.junit.internal.runners.statements.InvokeMethod.evaluate(InvokeMethod.java:20)[06:39:10] <bgeorges> at org.junit.runners.BlockJUnit4ClassRunner.runChild(BlockJUnit4ClassRunner.java:76)[06:39:12] <bgeorges> at org.junit.runners.BlockJUnit4ClassRunner.runChild(BlockJUnit4ClassRunner.java:50)[06:39:12] *** dmison has joined #jboss-as7[06:39:14] <bgeorges> at org.junit.runners.ParentRunner$3.run(ParentRunner.java:193)[06:39:16] <bgeorges> at org.junit.runners.ParentRunner$1.schedule(ParentRunner.java:52)[06:39:18] <bgeorges> at org.junit.runners.ParentRunner.runChildren(ParentRunner.java:191)[06:39:20] <bgeorges> at org.junit.runners.ParentRunner.access$000(ParentRunner.java:42)[06:39:22] <bgeorges> at org.junit.runners.ParentRunner$2.evaluate(ParentRunner.java:184)[06:39:24] <bgeorges> at org.junit.runners.ParentRunner.run(ParentRunner.java:236)[06:39:26] <bgeorges> at org.apache.maven.surefire.junit4.JUnit4TestSet.execute(JUnit4TestSet.java:59)[06:39:28] <bgeorges> at org.apache.maven.surefire.suite.AbstractDirectoryTestSuite.executeTestSet(AbstractDirectoryTestSuite.java:115)[06:39:30] <bgeorges> at org.apache.maven.surefire.suite.AbstractDirectoryTestSuite.execute(AbstractDirectoryTestSuite.java:102)[06:39:32] <bgeorges> at org.apache.maven.surefire.Surefire.run(Surefire.java:180)[06:39:34] <bgeorges> at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)[06:39:36] <bgeorges> at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:39)[06:39:38] <bgeorges> at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:25)[06:39:41] <bgeorges> at java.lang.reflect.Method.invoke(Method.java:597)[06:39:42] <bgeorges> at org.apache.maven.surefire.booter.SurefireBooter.runSuitesInProcess(SurefireBooter.java:350)[06:39:44] <bgeorges> at org.apache.maven.surefire.booter.SurefireBooter.main(SurefireBooter.java:1021)[06:39:46] <bgeorges> Tests run: 2, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 0.325 sec[06:39:48] <bgeorges> Running org.jboss.as.ejb3.deployment.processors.TransactionAttributeAnnotationProcessorTestCase[06:39:50] <bgeorges> Tests run: 2, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 0.066 sec[06:39:52] <bgeorges> Running org.jboss.as.ejb3.deployment.processors.TransactionManagementAnnotationProcessorTestCase[06:39:54] <bgeorges> Tests run: 3, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 0.012 sec[06:39:56] <bgeorges> Running org.jboss.as.ejb3.deployment.processors.BusinessViewAnnotationProcessorTestCase[06:39:58] <bgeorges> Tests run: 3, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 0.022 sec[06:40:00] <bgeorges> Running org.jboss.as.ejb3.component.pool.PooledInstanceInterceptorTestCase[06:40:02] <bgeorges> Tests run: 1, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 0.045 sec[06:40:04] <bgeorges> Results :[06:40:06] <bgeorges> Tests run: 12, Failures: 0, Errors: 0, Skipped: 0[06:40:10] <bgeorges> insight-apj:jboss-as bgeorges$ uname -a[06:40:12] <bgeorges> Darwin insight-apj.sin.redhat.com 10.7.0 Darwin Kernel Version 10.7.0: Sat Jan 29 15:17:16 PST 2011; root:xnu-1504.9.37~1/RELEASE_I386 i386[06:40:14] <bgeorges> insight-apj:jboss-as bgeorges$ java -version[06:40:16] <bgeorges> java version "1.6.0_24"[06:40:18] <bgeorges> Java(TM) SE Runtime Environment (build 1.6.0_24-b07-334-10M3326)[06:40:20] <bgeorges> Java HotSpot(TM) 64-Bit Server VM (build 19.1-b02-334, mixed mode)[06:41:54] <stuartdouglas> It looks like the exception it mean't to occurr, there is just an errant e.printStackTrace() in the test[06:42:22] <bgeorges> ok, thanks , I haven't looked at the test yet. just took rebase and run with it[06:43:56] <dmison> stupid question time: JARs in /modules , are they automatically available to the running server or do they need to be explicitly enabled in configuration as well ?[06:44:33] <stuartdouglas> you need to set up dependencies between the modules[06:44:56] <stuartdouglas> so if your server subsystem needs to use one of the jars, it needs a dependency entry in it's module.xml[06:45:57] <dmison> I mean JARs required by a deployed app, specifically the org.jboss.resteasy ones[06:46:27] <stuartdouglas> they should be available if any resteasy annotations are used in the deployment[06:47:01] <dmison> cool[06:53:41] *** magesh1 has joined #jboss-as7[06:53:50] <smcgowan> dmlloyd: in progress, we'll have results in the morning -[06:53:56] <dmlloyd> great.[06:53:58] <dmlloyd> thanks smcgowan[06:55:08] <smcgowan> dmlloyd: you should get some sleep too . . .[06:55:24] <dmlloyd> I'd like to, but got a few more things to do...[06:55:34] *** magesh has quit IRC[07:00:23] <smcgowan> dmlloyd: the EjbEmbed paths are not set up properly: http://pastebin.com/UyBTWuKA i'll have to work on that a bit more[07:01:11] <dmlloyd> hm okay[07:01:18] <dmlloyd> it was just a guess I admit[07:01:37] <dmlloyd> ah missing /[07:01:49] <smcgowan> better view : http://pastebin.com/226qn8Wz[07:02:49] <dmlloyd> ah since you're using the CLASSPATH env var, you'll need to add the -class switch to the end there[07:02:54] <dmlloyd> just before the $testExecuteClass[07:05:26] *** miclorb_ has quit IRC[07:06:15] *** kcbabo has quit IRC[07:07:40] <misty> I've got some questions to follow up on dmison's questions, if anyone is still here[07:08:03] <dmlloyd> I completely missed them[07:08:20] <misty> we are trying to figure out resteasy and AS7[07:08:39] <misty> stuartdouglas seemed to be saying that we don't need to add the resteasy modules to the standalone.xml[07:08:57] <stuartdouglas> no, it should just work[07:09:06] <stuartdouglas> are you having trouble with it?[07:09:10] <misty> but I'm trying to add resteasy capabilities to a deployment in JBDS, and it does not detect that my server has resteasy[07:09:23] <misty> I tried adding the JARs right into the project but it doesn't work[07:09:35] <misty> I am an extreme newbie so I am sure I am doing something stupid[07:09:40] <dmlloyd> see, I *knew* that would be the first thing people would try :)[07:10:50] <stuartdouglas> so is it just the JBDS part that is causing the problem? Does it actually work when deployed to the server?[07:11:19] <misty> well, I have no idea[07:11:24] <misty> because I can't get that far[07:11:35] <misty> I'm not smart enough to make this app outside of JBDS[07:11:37] <dmlloyd> it should work if you don't include the resteasy JARs in the deployment[07:11:53] <misty> let me see[07:12:00] <dmlloyd> if JBDS makes you have them then max owes you an apology! :)[07:12:37] <stuartdouglas> #jbosstools might be a better place to ask, I have never used JBDS so I am not really sure how it's resteasy support works[07:13:07] <dmlloyd> (we do have jax-rs on the classpath by default, right?)[07:13:23] <stuartdouglas> only if there are jax-rs annotations discovered[07:14:20] <misty> but they won't be discovered until you deploy something[07:14:21] <misty> right?[07:14:30] <stuartdouglas> yes[07:14:55] <misty> and JBDS won't let you deploy something with JAX-RS unless it detects that your server has the capabilities[07:15:01] <misty> if I am understand it properly[07:15:39] <misty> RESTEasy jars not found in the runtime associated with the selected project. Make sure RESTEasy is installed and try again.[07:15:45] <stuartdouglas> It's probably because the AS7 support is not complete in JBDS[07:16:05] <smcgowan> misty: i was just going to say the same thing - stuartdouglas beat me to it[07:16:10] <stuartdouglas> JBDS should realise that AS7 has resteasy support[07:16:19] <smcgowan> i don't think the tools are ready for AS 7[07:16:42] <dmlloyd> max would say AS 7 isn't ready for tools![07:17:10] <dmlloyd> that sounded funnier in my head[07:17:52] <misty> :( but I want to play[07:18:10] <misty> dmlloyd: did you mean tools as in people? ;)[07:18:55] <dmlloyd> I don't know[07:19:06] <dmlloyd> let's say.... no[07:23:15] *** ALR has quit IRC[07:23:22] <misty> lol[07:23:45] <misty> so I have to figure out what JBDS would have put in, put it in manually, and then try to deploy[07:23:58] *** miclorb has joined #jboss-as7[07:24:56] <smcgowan> misty: you can use the tools with AS 6 and then try to deploy to AS 7[07:25:01] <smcgowan> their both EE 6 impls[07:25:57] <misty> haha[07:25:59] <misty> good idea[07:26:11] * misty cracks her knuckles and tries it[07:26:40] <dmlloyd> you still may have to rip out resteasy from the resultant archive though.[07:26:59] <stuartdouglas> didn't AS6 bundle resteasy?[07:27:08] <dmlloyd> I don't know[07:27:49] <stuartdouglas> It did, so JBDS should be smart enough to realise it does not need to bundle[07:31:01] <smcgowan> AS 6 includes RestEasy[07:31:33] <smcgowan> going to catch a nap before my 9 a.m. meeting . . .[07:31:37] *** smcgowan has quit IRC[07:31:49] <misty> ok well AS6 is downloading now[07:36:33] *** liweinan has quit IRC[07:44:08] <misty> yeah I get the same error in AS7[07:44:12] <misty> err AS7[07:44:20] <misty> err AS6, it has been a long day[07:44:42] <stuartdouglas> I would try asking in #jbosstools[07:45:42] <misty> ok[07:47:05] <misty> sholuldn't resteasy be in servers/all/deploy though?[07:47:34] *** maeste has joined #jboss-as7[07:47:40] *** maeste has quit IRC[07:47:40] *** maeste has joined #jboss-as7[07:47:40] *** ChanServ sets mode: +v maeste[07:48:19] <stuartdouglas> ./server/default/deployers/resteasy.deployer[07:48:37] <misty> yes, I have it[08:09:40] *** rawbdor has quit IRC[08:16:26] *** rawbdor has joined #jboss-as7[08:17:51] *** jfclere has joined #jboss-as7[08:17:52] *** ChanServ sets mode: +v jfclere[08:18:30] *** mbg has joined #jboss-as7[08:18:30] *** ChanServ sets mode: +v mbg[08:23:40] *** miclorb has quit IRC[08:24:52] *** tdiesler has joined #jboss-as7[08:24:52] *** ChanServ sets mode: +v tdiesler[08:52:00] *** pferraro has quit IRC[08:52:44] *** miclorb_ has joined #jboss-as7[08:55:55] *** laubai has joined #jboss-as7[08:58:58] *** magesh1 has quit IRC[09:07:00] *** opalka has joined #jboss-as7[09:07:01] *** ChanServ sets mode: +v opalka[09:07:14] <opalka> morning[09:07:23] <stuartdouglas> morning[09:07:58] *** Jaikiran has joined #jboss-as7[09:07:59] *** ChanServ sets mode: +v Jaikiran[09:08:13] *** mbg is now known as mbg|away[09:11:05] *** pil-dinner has joined #jboss-as7[09:11:05] *** ChanServ sets mode: +v pil-dinner[09:12:50] *** mbg|away is now known as mbg[09:13:16] *** mbg has quit IRC[09:15:13] *** Jaikiran has quit IRC[09:16:35] *** Jaikiran has joined #jboss-as7[09:16:35] *** ChanServ sets mode: +v Jaikiran[09:24:53] *** rmaucher has joined #jboss-as7[09:26:35] *** slaboure has joined #jboss-as7[09:30:09] *** Jaikiran has quit IRC[09:37:36] *** asoldano has joined #jboss-as7[09:37:39] *** ChanServ sets mode: +v asoldano[09:43:00] *** maeste has quit IRC[09:51:43] *** dmison has quit IRC[09:51:48] *** laubai has quit IRC[09:51:53] *** hbraun has joined #jboss-as7[09:52:35] <hbraun> good morning[09:54:13] <tdiesler> hbraun, moin[09:54:25] <hbraun> tdiesler: moin ;)[09:55:41] *** pmuir has joined #jboss-as7[09:55:41] *** ChanServ sets mode: +v pmuir[10:01:07] *** emuckenhuber has joined #jboss-as7[10:01:07] *** emuckenhuber has joined #jboss-as7[10:01:07] *** ChanServ sets mode: +v emuckenhuber[10:02:35] *** pil-dinner is now known as pilhuhn[10:02:41] *** pilhuhn has joined #jboss-as7[10:02:41] *** leguin.freenode.net sets mode: +v pilhuhn[10:11:48] *** miclorb_ has quit IRC[10:24:14] <hbraun> tdiesler: i've split the OSGI management task into two separate ones: https://issues.jboss.org/browse/AS7-752 and https://issues.jboss.org/browse/AS7-744[10:24:15] <jbossbot> jira [AS7-752] Expose OSGi management through domain API [Open (Unresolved) Task, Major, Thomas Diesler] https://issues.jboss.org/browse/AS7-752[10:24:17] <jbossbot> jira [AS7-744] OSGI subsystem management through the console [Reopened (Unresolved) Feature Request, Major, Heiko Braun] https://issues.jboss.org/browse/AS7-744[10:24:46] <hbraun> the first one delas with the management API the second is the UI work[10:25:06] <hbraun> your team would need to expose the proper management operations[10:25:09] <tdiesler> hbraun, ok[10:25:22] <hbraun> we would then build the UI for that[10:25:53] <tdiesler> hbraun, my team ... LOL[10:26:26] <hbraun> isn't two people already a team?[10:26:37] <hbraun> ;)[10:40:24] *** miclorb has joined #jboss-as7[10:50:03] *** alesj has joined #jboss-as7[11:08:14] *** maxandersen has joined #jboss-as7[11:08:14] *** ChanServ sets mode: +v maxandersen[11:09:30] *** mlinhard has joined #jboss-as7[11:09:30] *** ChanServ sets mode: +v mlinhard[11:10:27] <mlinhard> hello guys, does anyone know how to run a single testcase in jboss-msc testsuite ?[11:10:40] <mlinhard> I'm trying to run CompleteServiceTestCase and it has some logging problems[11:11:30] <mlinhard> doing mvn test -Dtest=org.jboss.msc.service.CompleteServiceTestCase[11:17:19] *** kkhan has joined #jboss-as7[11:17:19] *** ChanServ sets mode: +v kkhan[11:22:02] *** miclorb has quit IRC[11:22:41] *** bgeorges has quit IRC[11:32:58] *** galderz has joined #jboss-as7[11:32:58] *** ChanServ sets mode: +v galderz[11:34:47] <mlinhard> got it (there are no problems just forgot to compile target/generated-sources/annotations/org/jboss/msc/service/ServiceLogger_$logger.java)[11:41:58] *** darranl has joined #jboss-as7[11:41:58] *** darranl has joined #jboss-as7[11:41:58] *** ChanServ sets mode: +v darranl[11:47:59] *** galderz has quit IRC[11:51:27] *** stliu has quit IRC[11:51:43] *** jcosta has joined #jboss-as7[11:51:43] *** ChanServ sets mode: +v jcosta[11:52:45] *** magesh has joined #jboss-as7[11:52:57] *** Binbin is now known as Binbinaway[12:22:48] *** slaboure has quit IRC[12:36:05] *** jwulf has quit IRC[12:37:22] *** fbricon has joined #jboss-as7[12:37:43] <fbricon> Hello world[12:39:08] <fbricon> I tried to run hudson-2.0.0.war on AS7. App is started, but when accessing a page, I get a java.lang.LinkageError: Failed to link hudson/model/Queue$Task (Module "deployment.hudson-war-2.0.0.war:main" from Service Module Loader)[12:40:51] *** JimMa has quit IRC[12:41:15] <fbricon> if anyone wants to try : http://search.maven.org/remotecontent?filepath=org/jvnet/hudson/main/hudson-war/2.0.0/hudson-war-2.0.0.war (37.2 MB)[12:46:32] <alesj> fbricon: i would guess this is due to new Module CL system[12:46:32] *** adietisheim has joined #jboss-as7[12:46:45] <alesj> it doesn't know how to find some class[12:47:10] *** maxandersen has quit IRC[12:47:55] *** wolfc has joined #jboss-as7[12:47:55] *** ChanServ sets mode: +v wolfc[12:48:44] <fbricon> alesj: is this a known bug or should I open a ticket?[12:49:34] <alesj> it's not a bug[12:49:52] <alesj> it's how AS7 works — no global libs anymore[12:50:32] <fbricon> global lib? Hudson stuff is in WEB-INF/lib[12:50:34] <alesj> perhaps dig into the problem a bit more, before posting any bug/issue[12:50:56] <alesj> sure, but it probably tries to load something it shouldn't, per the spec[12:51:04] <fbricon> thats' why I ask here first :-)[12:51:06] <alesj> otoh, i might be wrong, and it is a bug[12:52:42] *** sgilda has joined #jboss-as7[12:53:07] *** adietisheim has quit IRC[12:55:06] <stuartdouglas> I'm looking into the hudson thing now[12:57:00] <stuartdouglas> at first glance it looks like an AS7 bug[12:57:27] <emuckenhuber> stuartdouglas: yeah i think it's the vfs stuff alesj fixed - at least the last time i looked[12:57:49] <alesj> JBVFS-170?[12:57:53] <jbossbot> jira [JBVFS-170] Support for -Djboss.vfs.forceCaseSensitive=true [Resolved (Done) Feature Request, Minor, Ales Justin] https://issues.jboss.org/browse/JBVFS-170[12:57:55] <emuckenhuber> yes[12:58:08] <alesj> fbricon: try latest VFS3 from git master[12:58:10] <emuckenhuber> because hudson has something like hudson/queue and hudson/Queue[12:58:24] <alesj> https://github.com/jbossas/jboss-vfs[12:58:32] <alesj> fbricon: ^[12:58:48] *** mgoldmann has joined #jboss-as7[12:58:48] *** ChanServ sets mode: +v mgoldmann[13:05:21] <wolfc> stuartdouglas: are you still around?[13:05:32] <stuartdouglas> wolfc: yes[13:05:44] <stuartdouglas> I did a real impl of SessionObjectReference today[13:05:53] <wolfc> bstansberry and I noticed http://pastebin.com/rKfbghHc during AS7-431[13:05:54] <jbossbot> jira [AS7-431] Deployment content management enhancement [Open (Unresolved) Task, Major, Carlo de Wolf] https://issues.jboss.org/browse/AS7-431[13:06:51] <stuartdouglas> where it seems to finish the deployment twice?[13:07:31] <wolfc> No, that's intentional, because it's a full-replace, but notice how MSC starts reporting in the mean time.[13:08:31] <wolfc> Maybe that's tied to that issue of the MSC not yet stabilized.[13:08:54] <wolfc> (https://github.com/stuartwdouglas/jboss-as/commit/26484e2f5eac388f03d6129864107cdced432dda#L9R76 don't like the method name createInstance)[13:08:55] <jbossbot> git [jboss-as] 26484e2.. Stuart Douglas Add correct SessionObjectReference implementation[13:09:22] <wolfc> https://github.com/stuartwdouglas/jboss-as/commit/26484e2f5eac388f03d6129864107cdced432dda#L9R86 is dangerous[13:09:23] <jbossbot> git [jboss-as] 26484e2.. Stuart Douglas Add correct SessionObjectReference implementation[13:09:25] <stuartdouglas> it creates an instance of the bview[13:09:42] <wolfc> ah yes, I was looking cross-eyed[13:09:54] <stuartdouglas> how should I implement removal?[13:10:22] <wolfc> the tricky bit is that the instance might be in use on a transaction[13:10:30] <wolfc> the remove should then happen on commit[13:11:22] <wolfc> I think you should just call component.remove(id); and let ejb3 code handle the details[13:11:54] *** pmuir has quit IRC[13:12:02] *** galderz has joined #jboss-as7[13:12:02] *** ChanServ sets mode: +v galderz[13:12:08] <stuartdouglas> where is that method?[13:12:38] <wolfc> AFAIK it doesn't exist :-)[13:13:08] <stuartdouglas> so I should just add a TODO?[13:14:00] *** torben has joined #jboss-as7[13:14:00] *** ChanServ sets mode: +v torben[13:14:48] <wolfc> https://github.com/stuartwdouglas/jboss-as/blob/master/ejb3/src/main/java/org/jboss/as/ejb3/component/stateful/StatefulSessionSynchronizationInterceptor.java is currently handling the release and should also do the remove at some point[13:14:59] <wolfc> when is SessionObjectRef.remove being called?[13:17:14] <stuartdouglas> when CDI destroys a bean[13:17:24] <stuartdouglas> usually after the request / session ends[13:18:03] <stuartdouglas> all my CDI changes today get me down to 10 failures in the CDI TCK[13:18:22] *** galderz has quit IRC[13:20:20] <wolfc> does/can that happen in a tx?[13:20:46] <stuartdouglas> I don't think so[13:20:49] <stuartdouglas> not 100% sure[13:21:09] <stuartdouglas> having a proper destroy method would be better[13:21:12] *** magesh is now known as magesh_brb[13:21:49] <stuartdouglas> cache.remove calls componentInstance.destroy, will that be where the TX logic goes?[13:24:18] *** rawbdor has quit IRC[13:24:25] *** galderz has joined #jboss-as7[13:24:25] *** ChanServ sets mode: +v galderz[13:24:27] *** mgoldmann_ has joined #jboss-as7[13:24:29] *** mgoldmann has quit IRC[13:26:29] *** bbrowning has joined #jboss-as7[13:26:29] *** ChanServ sets mode: +v bbrowning[13:28:47] *** mgoldmann_ has quit IRC[13:33:12] <wolfc> stuartdouglas: no componentInstance.destroy is just the pre-destroy callback[13:34:19] <stuartdouglas> so where will the destroy logic end up?[13:35:30] *** Jaikiran has joined #jboss-as7[13:35:30] *** ChanServ sets mode: +v Jaikiran[13:35:48] <Jaikiran> stuartdouglas: you around?[13:35:57] <stuartdouglas> yes[13:37:14] <stuartdouglas> david rebased ejb against upstream today[13:37:47] <wolfc> who has the most ahead branch now?[13:37:51] <stuartdouglas> me[13:38:01] <wolfc> github network view sucks with this many branches :-)[13:38:13] <Jaikiran> stuartdouglas: i rebased against yours this morning[13:38:13] <stuartdouglas> not by a huge number of commits though[13:38:19] <Jaikiran> https://github.com/jaikiran/jboss-as/commits/new-ee-framework[13:38:34] <Jaikiran> i was working on the tck stuff locally and have some new fixes which i have pushed[13:38:40] <Jaikiran> looks like you commited some too[13:38:46] <Jaikiran> let me rebase against yours again[13:38:50] <stuartdouglas> I have weld down to 10 failures[13:38:57] <Jaikiran> tck?[13:39:00] <stuartdouglas> yes[13:39:03] <Jaikiran> cool![13:39:13] <Jaikiran> EJB3 tck is getting back to what it was before this refactoring[13:39:22] <Jaikiran> it's not completely there but a big progress[13:39:39] <stuartdouglas> Weld is much better, I think upstream has around 25 failures[13:40:16] <Jaikiran> i think we should be able to merge this upstream by tomorrow[13:40:25] <Jaikiran> dmlloyd was wanting to get this done this week too[13:40:27] <stuartdouglas> yay[13:41:11] *** asoldano is now known as asoldano_lunch[13:41:23] <stuartdouglas> rebasing adds some more test failures, as there are some failing tests in upstream[13:42:43] <fbricon> alesj: stuartdouglas: emuckenhuber: after compiling jboss-vfs, I no longer have the linkage error. However hudson crashes for a strange reason[13:43:16] <stuartdouglas> but out of the ones that are regressions one is due to no concurrency interceptor[13:43:36] <emuckenhuber> fbricon: you mean something like: http://pastebin.com/uTW0CfTy[13:43:37] <fbricon> I have an error page stating : "We detected that your JVM is not supported by Hudson. This is due to the limitation is one of the libraries that Hudson uses, namely XStream. See this FAQ for more details." Detected JVMVendor: Sun Microsystems Inc.Version: 19.1-b02VM Name: Java HotSpot(TM) 64-Bit Server VMOS Name: Linux[13:43:37] <emuckenhuber> ?[13:43:48] <stuartdouglas> another is a JPA one that I am not sure about, I think it may be to do with jpa interceptor ordering[13:44:03] <fbricon> emuckenhuber: yes that's what I got in the console[13:44:13] <pilhuhn> rmaucher :read-resource on connector=http delivers an empty name attribute - even if the description says name is mandatory[13:44:21] <pilhuhn> (and some other fields too)[13:44:29] <wolfc> stuartdouglas: you mean on singleton?[13:44:43] <stuartdouglas> on SFSB[13:44:45] <fbricon> emuckenhuber: detected JVM version : 19.1-b02 !!![13:45:01] <wolfc> there is locking in the interceptor I showed[13:45:05] <stuartdouglas> it is allowing concurrent calls[13:45:33] <stuartdouglas> the interceptor is there, it is just not added to the chain[13:45:41] <wolfc> sigh[13:46:07] <pilhuhn> emuckenhuber what is the general policy on the name attribute on resources in the tree?[13:46:11] *** galderz has quit IRC[13:46:17] <pilhuhn> One needs it for :add of course.[13:47:27] <emuckenhuber> pilhuhn: we usually just don't use the "name" attribute, in case it's part of the address[13:47:43] <stuartdouglas> wolfc: Jaikiran What is the plan for the ejb timer service? Are we using quartz ?[13:47:52] <Jaikiran> stuartdouglas: no we aren't[13:47:53] <pilhuhn> shouldn't that then be left out of the attributes of a resoruce?[13:47:56] <Jaikiran> atleast not in AS6[13:48:08] <Jaikiran> we had a JPA backed timerservice impl in AS6[13:48:18] <kkhan> emuckenhuber / pilhuhn inconsistencies might exist, I tried to get rid of those name attributes a few weeks ago[13:48:20] <wolfc> stuartdouglas, Jaikiran: it's out of scope for AS 7.0[13:48:36] <Jaikiran> timerservice?[13:48:42] <pilhuhn> kkhan you are a hero![13:48:58] <emuckenhuber> fbricon: yeah i doubt it's related to the JVM version - rather to the way hudson is trying to get a hold of resources, which we most likely are not going to support in AS7[13:49:16] <kkhan> pilhuhn: If you see any places where both are being used let me know[13:49:18] <Jaikiran> i thought there was a jira which targetted it for CR1[13:49:21] <Jaikiran> not sure which one[13:49:31] <wolfc> Jaikiran: unless we can do a lite-profile and full-profile rolled into one[13:49:44] <stuartdouglas> there is one[13:49:46] *** magesh_brb is now known as magesh[13:49:58] <wolfc> Basically AS 7.0 'web' must not expose an EJB timer service.[13:50:01] <Jaikiran> oh ok, i get the context of moving it out now[13:50:05] <pilhuhn> kkhan [domain@localhost:9999 connector=http] :read-resource-description[13:51:31] <pilhuhn> kkhan same for virtual-server[13:51:49] <kkhan> pilhuhn: Which subsystem?[13:52:07] <wolfc> crock, all those interceptor additions in the configurations should never have been removed.[13:52:08] <pilhuhn> ah sorry - thought the command line would show it. Web-subsystem[13:52:39] <kkhan> pilhuhn: Right, that is one subsystem I did not touch. It is so big I got scared :-) Maybe jfclere can help[13:52:51] <wolfc> Jaikiran, do you have TX interceptors back in?[13:53:12] <rmaucher> nope, no clue about these descriptions stuff ;)[13:53:20] <Jaikiran> wolfc: in the new EE framework?[13:53:21] <Jaikiran> yes[13:53:26] <wolfc> where is that?[13:53:31] <Jaikiran> but it's very brittle in it's current state[13:53:42] <Jaikiran> and something that stuartdouglas and I have been discussing about changing[13:53:48] <Jaikiran> one sec[13:53:57] *** kcbabo has joined #jboss-as7[13:53:57] *** ChanServ sets mode: +v kcbabo[13:54:42] *** galderz has joined #jboss-as7[13:54:42] *** ChanServ sets mode: +v galderz[13:55:18] *** adietisheim has joined #jboss-as7[13:55:43] <Jaikiran> wolfc: here's one example https://github.com/jaikiran/jboss-as/blob/new-ee-framework/ejb3/src/main/java/org/jboss/as/ejb3/component/session/SessionBeanComponentDescription.java#L461[13:55:43] <kkhan> pilhuhn: Open a jira and assign to me or emuckenhuber[13:55:55] <Jaikiran> is something failing?[13:56:00] <kkhan> I can't look right away, so I don't want to forget[13:56:11] <pilhuhn> okay - thanks kkhan[13:56:22] <emuckenhuber> and thanks for spotting all those nasty details! :)[13:57:03] *** vtunka has quit IRC[13:58:12] <stuartdouglas> wolfc: we have been talking about changing the way that interceptors and configurators are ordered, so the order is by numeric priority (think phase.java)[13:59:12] <stuartdouglas> at the moment it all depends on when the DUP executes, wether the configurator is added to the front or the back of configurators list, and wether the interceptor is added to the front or the back of the chain[13:59:48] <jfclere> [localhost:9999 web] ./:read-resource-description[13:59:49] <stuartdouglas> which makes it really hard to change the order of a given interceptor chain without changing the order of all the DUP's[13:59:55] <wolfc> Jaikiran: StatefulSessionSyncInt is not in place[14:00:03] <jfclere> her you have the descriptions[14:00:10] <Jaikiran> wolfc: yeah, i had to comment it out to make sure i add it at the right place[14:00:15] <stuartdouglas> and even harder to figure out the existing order[14:00:35] <Jaikiran> it's stashed in one of my local changes, waiting to be tested properly[14:00:39] <Jaikiran> before being pushed[14:01:01] <wolfc> stuartdouglas: that's the nature of the beast[14:01:18] <wolfc> basically we need to create proper DUPs that put the interceptor into the description[14:01:55] <stuartdouglas> but that will result in a heap of DUP's that do nothing but add an interceptor in the right place[14:02:01] <wolfc> For example, the tx interceptors should actually be part of the ejb-tx integration submodule[14:02:14] <wolfc> so that if both are enabled, the tx interceptors take hold[14:03:00] <wolfc> that's the theoretical idea[14:03:12] *** bstansberry has joined #jboss-as7[14:03:12] *** ChanServ sets mode: +v bstansberry[14:03:57] *** hardy has joined #jboss-as7[14:11:11] *** vtunka has joined #jboss-as7[14:11:11] *** ChanServ sets mode: +v vtunka[14:13:32] *** jpederse has joined #jboss-as7[14:13:37] *** jpederse has joined #jboss-as7[14:17:23] *** adietisheim has quit IRC[14:26:57] *** bstansberry has quit IRC[14:34:02] <wolfc> Jaikiran: I'm missing per component instance interceptors[14:34:18] <Jaikiran> wolfc: per componentinstance or per component?[14:34:35] <wolfc> Really truly per component instance. :-)[14:34:54] <wolfc> The StatefulSessionSyncInt keeps a transactionKey (per instance).[14:34:54] *** jamezp_afk has quit IRC[14:35:17] * Jaikiran opens code[14:35:28] *** asoldano_lunch is now known as asoldano[14:35:46] <wolfc> It may have to be refactored anyway to allow outside remove requests to be processed correctly.[14:42:50] * Jaikiran is reading spec to understand what the rules are for tsyncregistry[14:50:04] *** opalka has quit IRC[14:50:40] *** Binbinaway has quit IRC[14:52:11] <Jaikiran> wolfc: is this what you meant was missing http://pastebin.com/xwdL05iV ?[14:52:30] *** alesj has quit IRC[14:52:41] <wolfc> Jaikiran: no, that doesn't work[14:52:53] *** maeste has joined #jboss-as7[14:52:53] *** ChanServ sets mode: +v maeste[14:52:58] <wolfc> There can be only one StatefulSessionSyncInt per component instance, else concurrency won't work[14:53:12] *** mmoyses has joined #jboss-as7[14:53:12] *** ChanServ sets mode: +v mmoyses[14:54:39] *** Binbinaway has joined #jboss-as7[14:55:18] <Jaikiran> i'm missing something then[14:55:51] *** alesj has joined #jboss-as7[14:55:58] <Jaikiran> why does it have to be at that level instead of per componentinstance (a.k.a bean instance) level?[14:56:22] *** pmuir has joined #jboss-as7[14:56:22] *** pmuir has joined #jboss-as7[14:56:22] *** ChanServ sets mode: +v pmuir[14:57:13] *** alesj has quit IRC[14:57:16] *** bstansberry has joined #jboss-as7[14:57:16] *** ChanServ sets mode: +v bstansberry[14:57:31] *** alesj has joined #jboss-as7[14:57:36] *** bstansberry has quit IRC[14:57:47] *** smarlow has joined #jboss-as7[14:57:48] *** ChanServ sets mode: +v smarlow[14:59:28] <Jaikiran> smarlow: quick question - are those jpa jndi binding issues in tck sorted out?[14:59:32] *** pilhuhn is now known as pil-meeting[15:00:14] <wolfc> Jaikiran: you now create a new interceptor per view method[15:00:36] <wolfc> or component method.[15:01:54] <smarlow> Jaikiran: I believe so, I think that Shelly started making standalone.xml (for tck test run) changes that were based on the jpa jndi binding issues being sorted out. She mentioned that the tests appeared to be passing but I'm still waiting for the final word that everything is still green.[15:02:42] *** bstansberry has joined #jboss-as7[15:02:42] *** ChanServ sets mode: +v bstansberry[15:03:58] <Jaikiran> wolfc: ah right, i see what you mean[15:04:02] *** galderz has quit IRC[15:04:35] *** mgoldmann has joined #jboss-as7[15:04:35] *** ChanServ sets mode: +v mgoldmann[15:04:48] <Jaikiran> smarlow: ok, thanks. i was looking at the tck run locally and saw some failures (possibly due to other reasons or my local changes) so just wanted to check with you first[15:06:13] <smarlow> I know that we have testsuite/integration failures currently but they are only with the Hibernate second level cache[15:08:27] *** galderz has joined #jboss-as7[15:08:27] *** ChanServ sets mode: +v galderz[15:10:16] *** mbg has joined #jboss-as7[15:10:16] *** ChanServ sets mode: +v mbg[15:12:21] *** tdiesler has quit IRC[15:12:51] *** galderz has quit IRC[15:13:09] *** adietisheim has joined #jboss-as7[15:16:19] *** adietisheim has quit IRC[15:17:48] *** adietisheim has joined #jboss-as7[15:22:35] *** tdiesler has joined #jboss-as7[15:22:40] *** tdiesler has joined #jboss-as7[15:22:40] *** ChanServ sets mode: +v tdiesler[15:34:34] <Jaikiran> wolfc: btw, i see that even for the view interceptors, we have been creating the interceptor one per method[15:34:49] <Jaikiran> so that needs to be fixed[15:44:19] <hbraun> bstansberry: fyi https://issues.jboss.org/browse/AS7-757[15:44:21] <jbossbot> jira [AS7-757] Provide an operation to distinguish domain & standalone [Open (Unresolved) Enhancement, Major, Brian Stansberry] https://issues.jboss.org/browse/AS7-757[15:44:55] <hbraun> bstansberry: a generic meta-data operation would be very helpful[15:45:17] *** balunasj has joined #jboss-as7[15:45:17] *** balunasj has joined #jboss-as7[15:45:17] *** ChanServ sets mode: +v balunasj[15:46:01] <bstansberry> hbraun: yeah, that's an obvious need[15:46:11] <hbraun> bstansberry: fine[15:46:17] <hbraun> bstansberry: even better[15:47:04] <bstansberry> thinking through the various aspects...[15:47:06] <hbraun> bstansberry: then I'll put aside the bucket full of arguments I collected the whole day[15:47:16] <bstansberry> lol. now i feel bad[15:47:18] <hbraun> bstansberry: in order to convince you ;)[15:48:00] <hbraun> bstansberry: yeah, I am fully armed. Couldn't wait for you to show up ...[15:48:17] <hbraun> bstansberry: actually I just bumped into it[15:48:32] <hbraun> could be that the error handling improved[15:48:41] <bstansberry> i was going to say i'm surprised it never came up in all these months; i.e. i never filed the jira, you never filed, heiko, emanuel never filed.[15:49:16] <hbraun> now you see who does all the work[15:49:25] <bstansberry> yep![15:49:30] <hbraun> lol[15:49:38] <bstansberry> so, i'm thinking a bit about variants[15:49:56] <bstansberry> you have the DC, and slave HC, then a server[15:50:26] <bstansberry> server can be running in different modes, standalone, domain, embedded[15:50:51] <bstansberry> the embedded bit is a mixing of concerns[15:52:23] <wolfc> wouldn't the embedded be either domain or standalone?[15:52:58] <dmlloyd> greetings[15:54:29] <hbraun> good morning mr. lloyd[15:58:58] *** ALR has joined #jboss-as7[15:58:58] *** ChanServ sets mode: +v ALR[15:59:05] *** pil-meeting is now known as pilhuhn[15:59:23] *** pgier has joined #jboss-as7[15:59:23] *** ChanServ sets mode: +v pgier[15:59:24] *** maxandersen has joined #jboss-as7[15:59:24] *** ChanServ sets mode: +v maxandersen[16:00:38] *** clebert has joined #jboss-as7[16:00:38] *** ChanServ sets mode: +v clebert[16:01:44] <bstansberry> wolfc: a domain mode server would never be truly embedded -- it's a separate process spawned by the ProcessController[16:02:00] *** tdiesler has quit IRC[16:02:27] <bstansberry> it's possible the ProcessController could be "embedded", which would be relevant to the HostController/DomainController[16:02:57] <bstansberry> so, yeah, that tells me the embedded thing is a separate concern.[16:03:40] <wolfc> I actually hadn't thought of that. I would only expect the DC to be embedded. Why can't we stuff it all in one process?[16:04:25] *** galderz has joined #jboss-as7[16:04:25] *** ChanServ sets mode: +v galderz[16:05:08] <hbraun> then I could run it my android phone[16:05:10] <bstansberry> so the HC/DC modules can be patched[16:06:02] <bstansberry> the PC module is basic, no complex logic, should be bug free and not needing patching[16:06:11] *** Nihility has joined #jboss-as7[16:06:11] *** Nihility has joined #jboss-as7[16:06:11] *** ChanServ sets mode: +v Nihility[16:06:24] *** vtunka has quit IRC[16:08:31] * bstansberry heads to Nihility's[16:08:34] <bstansberry> bbiab[16:08:36] *** bstansberry has quit IRC[16:08:51] <bobmcw> Nihility: hey... in jboss-as-build's pom, could we mark the dependencies as scope=compile or something[16:09:04] <bobmcw> so that my dependency on jboss-as-build.zip doesn't transitively download everything that went into it also?[16:09:21] <bobmcw> I only care about the end-product, not it's full BOM[16:10:33] *** frainone has joined #jboss-as7[16:10:33] *** ChanServ sets mode: +v frainone[16:11:44] <Nihility> bobmcw: sure[16:11:54] <bobmcw> I'm not sure which scope accomplishes what I need[16:12:04] <bobmcw> but just breaking the transitives somehow[16:12:23] *** tdiesler has joined #jboss-as7[16:12:23] *** ChanServ sets mode: +v tdiesler[16:12:37] <Nihility> bobmcw: maybe pgier can offer advice[16:13:00] <dmlloyd> stuartdouglas: still up?[16:13:01] <pgier> for AS6 we had a separate POM for the dist zip[16:13:07] *** lgao has left #jboss-as7[16:13:11] <Nihility> dmlloyd: we are at my place this morning[16:13:19] <dmlloyd> aye aye[16:13:24] <dmlloyd> I'll be there in a while[16:13:27] <Nihility> ok[16:14:00] *** mbg has quit IRC[16:15:13] <dmlloyd> I have to try to time my sleep schedule so I wake up when stuart is still up :)[16:15:30] <pgier> bobmcw: do you already have a jira? I could probably create a new pom just for the zip download[16:15:45] <bobmcw> pgier: no, I have only whinged on IRC thus far[16:16:09] <bobmcw> I'll create one[16:16:37] <tdiesler> ALR, call time?[16:17:42] <tdiesler> ALR, unless there is no update - aslak seems busy[16:17:42] <ALR> tdiesler: Aslak is a GeeCon[16:17:55] <ALR> Yeah, things are centering on his updates ATM[16:18:35] <tdiesler> ALR, shall we re-schedule or do you rather talk now?[16:18:36] <bobmcw> pgier: AS7-758[16:18:37] <jbossbot> jira [AS7-758] jboss-as-build.zip's maven artifact transitively imposes its input dependencies to any consumers of it [Open (Unresolved) Enhancement, Major, Jason Greene] https://issues.jboss.org/browse/AS7-758[16:19:08] <pgier> bobmcw: thanks, I'll take a look[16:19:11] <ALR> tdiesler: Tomorrow OK for you?[16:19:21] <ALR> (It's what Aslak proposed on the email thread)[16:19:36] <ALR> Also I invited AndyM, if he wanted to sit in.[16:20:02] *** slaboure has joined #jboss-as7[16:20:03] <hbraun> emuckenhuber: ping[16:20:09] <emuckenhuber> hbraun: pong[16:20:20] <hbraun> emuckenhuber: got a minuute?[16:20:25] <emuckenhuber> hbraun: yes[16:20:30] <tdiesler> ALR, me still on UMTS so only checking email rarely - hang on ...[16:20:37] <hbraun> can you take a look at the messaing security settings? /profile=default/subsystem=messaging:read-resource[16:20:45] *** vtunka has joined #jboss-as7[16:20:45] *** ChanServ sets mode: +v vtunka[16:20:58] <hbraun> i am wondering how to add security-setting's[16:21:05] <ALR> tdiesler: :)[16:21:06] <hbraun> shouldn't this be a subresource?[16:21:57] <emuckenhuber> hbraun: hmm, yes it actually should[16:22:06] <emuckenhuber> but not sure if we can do that without the hornetQ team[16:22:08] <hbraun> emuckenhuber: that#s the general idea, right?[16:22:16] <tdiesler> ALR, ok. lets talk tomorrow[16:22:17] <hbraun> add() only works on subresources[16:22:31] <ALR> tdiesler: Super. :) Seeya then, and I'll likely have SW Beta out[16:22:40] <emuckenhuber> yes, afaik we did not do it in the first place because we have to create the complete configuration when starting AS[16:22:44] <hbraun> emuckenhuber: same problem with addressing-setting[16:23:04] <hbraun> emuckenhuber: means it#s r/o?[16:25:03] <emuckenhuber> hmm, well it shouldn't be - however not sure how much work it would be to make that a subModel[16:25:37] <hbraun> emuckenhuber: ok, I create an issue then[16:26:22] <emuckenhuber> ok[16:26:56] <hbraun> emuckenhuber: : should I assign somebody right away?[16:27:24] <bbrowning> A big thank you for making the CLI retain command history even after exiting and restarting :)[16:27:49] <emuckenhuber> hbraun: not at the moment.. i think[16:27:57] <hbraun> ok[16:28:24] *** jamezp has joined #jboss-as7[16:29:04] *** balunasj has quit IRC[16:32:25] *** pmuir has quit IRC[16:34:48] *** jdcasey has joined #jboss-as7[16:35:27] *** bstansberry has joined #jboss-as7[16:35:27] *** ChanServ sets mode: +v bstansberry[16:37:09] <bstansberry> back[16:37:23] <bstansberry> emuckenhuber: let me know when ice cream is done :)[16:37:31] *** dimitris_ has joined #jboss-as7[16:37:41] *** dimitris_ has quit IRC[16:37:41] *** dimitris_ has joined #jboss-as7[16:37:41] *** ChanServ sets mode: +v dimitris_[16:40:47] *** mgoldmann has quit IRC[16:40:52] <wolfc> Jaikiran, I think I know how to re-introduce per ComponentInstance interceptors[16:41:52] <Jaikiran> listening[16:41:57] <wolfc> stuartdouglas: you wouldn't happen to be still awake?[16:43:45] <emuckenhuber> bstansberry: ok, so do you think it would make sense to separate the navigation through the mgmt tree from the actual model?[16:44:57] <bstansberry> emuckenhuber: this is basically an issue for read-resource(recursive=true), right?[16:45:07] <emuckenhuber> yes and read-children-names[16:45:12] <wolfc> Jaikiran: it would appear that componentInterceptors are now per ComponentInstance[16:45:27] <emuckenhuber> basically could be useful for "deployments", "host", "server"[16:45:27] <asoldano> mmoyses, Hi! On AS6 we used to have a JBossWS security domain coming by default on the AS; can we evaluate doing the same for AS7?[16:45:40] *** pmuir has joined #jboss-as7[16:45:40] *** ChanServ sets mode: +v pmuir[16:45:50] <Jaikiran> wolfc: yes, they are triggered for the ComponentInstance[16:45:52] <bstansberry> emuckenhuber: yes, that makes sense. the MNR is authoritative for the model structure[16:46:23] <bstansberry> i.e. if someone sticks some stuff in the model that looks like a child resource, it's not a child resource unless there is an entry in the MNR[16:46:27] <wolfc> Jaikiran: ah yes, not entirely. At least the InterceptorFactoryContext is tied to the ComponentInstance and then per method[16:46:41] <mmoyses> asoldano: look for the security subsystem in the standalone.xml. you can put a <security-domain> element there for JBossWS[16:46:45] <wolfc> I'll fiddle a bit[16:46:49] <bstansberry> i.e. you can't invoke an operation against that address[16:46:50] <Jaikiran> ok[16:46:56] <mmoyses> asoldano: let me know if you need help setting the options[16:47:07] <asoldano> mmoyses, is there some doc anywhere?[16:47:21] <mmoyses> not yet :([16:47:39] <bstansberry> emuckenhuber: the split between model and registry is a bit odd.[16:47:41] <asoldano> mmoyses, in any case, my question was mainly if I can add that together with the "other" one by default[16:48:01] <emuckenhuber> bstansberry: yeah, maybe navigation should be separate from the model and registration...[16:48:03] <asoldano> mmoyses, jbossws testsuite have many tests/demos that used that in the past[16:48:06] <emuckenhuber> well maybe that's a bit odd as well[16:48:53] <mmoyses> asoldano: sure. i expect more security domains will be added there. go ahead and add the one you need[16:49:02] <asoldano> mmoyses, ok, thanks[16:49:08] <mmoyses> asoldano: it should be very similar to AS6[16:50:30] <bstansberry> emuckenhuber: the thing that's not nice about combining is forcing op handlers to see child resources via something more complex than parentNode.get(type, name)[16:58:08] <bstansberry> darranl: baileyje is working on audit logging; we're looking for info on how we're going to get user identity info to log[16:58:26] *** magesh has left #jboss-as7[16:58:46] *** slaboure has quit IRC[16:59:55] <bstansberry> darranl: iow, what's up?[17:00:02] *** kcbabo has quit IRC[17:00:03] <emuckenhuber> bstansberry: yeah, i don't really think they can be combined - especially since we have wildcard registries it does not really represent the state[17:00:15] <bstansberry> good point[17:01:06] *** frainone is now known as frainone_lunch[17:03:50] <darranl> bstansberry, just adding a second user store now so will have two types of verification and then will start to have something to pass in - has the location the audit log will happen been identified?[17:04:04] <asoldano> mmoyses, ok, I had a look at the login module declaration in the domain; now, I add my JBossWS domain, using a UsersRoles login module ... I assume I can provide a text file or something with the allowed user/roles... ?[17:04:17] <bstansberry> darranl: baileyje has a branch[17:05:16] <bstansberry> darranl: https://github.com/baileyje/jboss-as/commits/AS7-444[17:05:18] <jbossbot> jira [AS7-444] Audit logging [Open (Unresolved) Task, Critical, John Bailey] https://issues.jboss.org/browse/AS7-444[17:06:12] <darranl> bstansberry, ok will take a look[17:08:12] <hbraun> bstansberry: quick question, if you don't mind?[17:08:19] <bstansberry> sure[17:08:36] <hbraun> bstansberry: subsystem=web connector socket-binding[17:08:53] <hbraun> it refers to a socket-binding name[17:08:57] <hbraun> i.e. "http"[17:09:09] <bstansberry> yep[17:09:09] <hbraun> how is the socket-binding-group identified?[17:09:28] <bstansberry> on a server there is only one[17:09:34] <hbraun> or is it not necessary?[17:09:42] <hbraun> ah, makes sense[17:09:48] <bstansberry> on a domain it would need to be resolved on a per-server-group basic[17:09:48] <hbraun> so, you assume it exists[17:09:50] <bstansberry> basis[17:10:26] *** mbg has joined #jboss-as7[17:10:26] *** ChanServ sets mode: +v mbg[17:10:29] <hbraun> tnx[17:10:44] *** mbg has quit IRC[17:10:44] *** mbg has joined #jboss-as7[17:10:44] *** leguin.freenode.net sets mode: +v mbg[17:11:02] <tdiesler> dmlloyd, ping[17:11:02] *** hardy has quit IRC[17:11:13] <dmlloyd> what's up tdiesler[17:11:36] *** hardy has joined #jboss-as7[17:11:42] <tdiesler> dmlloyd, are you planing a modules release any time soon?[17:12:11] <dmlloyd> just did one last night[17:12:12] <dmlloyd> CR1[17:12:23] <pilhuhn> jfclere for subsystem=web,virtual-server=x there is element rewrite of type LIST and element access-log of type OBJECT - it that right?[17:12:37] <tdiesler> dmlloyd, I have a long trailback of commits that are waiting https://github.com/tdiesler/jboss-as/commits/as722/[17:12:58] <hbraun> one access log, multiple reqrite rules[17:13:53] <tdiesler> dmlloyd, ok checking ...[17:14:37] <pilhuhn> hbraun ok, that makes sense - thanks[17:14:52] <hbraun> pilhuhn: i am working on these as we speak[17:15:42] <mmoyses> asoldano: yes. if don't add any module-options to this particular login module it will look for a users.properties and roles.properties in the classpath[17:15:45] *** hardy has quit IRC[17:16:12] <wolfc> Jaikiran, https://github.com/wolfc/jboss-as/commit/d6fbab9ebfdb709ea21049f61e526dea4a2dc2a8[17:16:13] <jbossbot> git [jboss-as] d6fbab9.. Carlo de Wolf Re-enabled StatefulSessionSynchronizationInterceptor[17:17:08] <asoldano> mmoyses, ok, what's the format for specifying the module-options?[17:17:26] <pilhuhn> tdiesler why umts? Have you moved to you new house?[17:17:39] <asoldano> mmoyses, and by classpath you mean in the classpath related to the module classloader of the deployment?[17:17:48] <tdiesler> pilhuhn, yes[17:17:57] <pilhuhn> cool, congrats tdiesler[17:18:14] <mmoyses> asoldano: yes. so normally WEB-INF/classes[17:18:22] <dmlloyd> tdiesler: you may want to wait a day or two. the integration of the new jboss-modules is in the ejb branch which is a bit dynamic at the moment[17:18:32] <dmlloyd> it includes the protocol handler fixes for example[17:18:38] <dmlloyd> anyway once that's in you'll want to rebase[17:18:41] <mmoyses> asoldano: <module-option name="usersProperties" value="something"/>[17:18:51] *** bstansberry has quit IRC[17:19:06] <mmoyses> asoldano: and <module-option name="rolesProperties" value="somethingelse"/>[17:19:25] <tdiesler> dmlloyd, you did not seem to have included my preDefine hook[17:19:47] *** pferraro has joined #jboss-as7[17:19:47] *** ChanServ sets mode: +v pferraro[17:19:48] <tdiesler> dmlloyd, this dooms OSGi 4.2 TCK compliance[17:19:50] <asoldano> mmoyses, ok, and those "something" should be the path to a prop file in the AS ?[17:20:04] <dmlloyd> tdiesler: hmmm I thought that was in there...[17:20:05] <mmoyses> asoldano: for those options you can use a file path as the value yes[17:20:13] *** alesj has quit IRC[17:20:22] <tdiesler> dmlloyd, https://github.com/tdiesler/jboss-modules/commit/a4cb6f0598d532[17:20:23] <jbossbot> git [jboss-modules] a4cb6f0.. Thomas Diesler [MODULES-65] Add a pre-define hook[17:20:24] <jbossbot> jira [MODULES-65] Deadlock on circular class load [Resolved (Done) Bug, Critical, Thomas Diesler] https://issues.jboss.org/browse/MODULES-65[17:20:32] <mmoyses> asoldano: like ../standalone/configuration/jbossws-users.properties or something like that[17:20:40] <dmlloyd> tdiesler: looks like there'll be a CR2... :)[17:20:56] <tdiesler> dmlloyd, yes seems like it ;-)[17:21:37] <asoldano> mmoyses, ok, assuming I use a relative path as above, the where is the root?[17:21:38] <jfclere> pilhuhn: a list of Aliases (string), access-log and rewrite rules[17:22:00] <pilhuhn> jfclere got it - hbraun explained it already[17:22:15] *** bstansberry has joined #jboss-as7[17:22:16] *** ChanServ sets mode: +v bstansberry[17:22:20] <mmoyses> asoldano: root is where the jvm is started. if you are using standalone.sh to start it the root is JBOSS_HOME/bin[17:22:49] <asoldano> mmoyses, ok, thanks![17:23:10] <mmoyses> asoldano: i will try and get system properties parsed there so you add a value like ${jboss.home}/path[17:23:41] <Nihility> guys just so everyone knows[17:23:49] <Nihility> the CR1 date is going to be pushed[17:24:04] <hbraun> all ears[17:24:18] <Nihility> tentative new date is 15th of june[17:24:32] <hbraun> Final still 2 weeks after that?[17:24:35] <Nihility> so not by alot but enough that you dont have to panic[17:25:03] <asoldano> Nihility, panic delayed by 2 weeks ;-) (joking)[17:25:13] <Nihility> hahahahaha[17:25:30] <Nihility> yeah likely 2 weeks[17:25:39] <hbraun> great[17:25:43] <Nihility> basically we need to be done by CR1[17:26:17] <Nihility> the only stuff after that shuold be bugs comunity finds[17:26:18] <hbraun> are we doing onother interim?[17:26:22] <hbraun> liek beta4?[17:26:33] <Nihility> no plans for that[17:26:46] <dmlloyd> tdiesler: Ah I see you moved the hooks to a different method. Was that in order to get outside of the lock?[17:28:05] <jamezp> Nihility: Anything you'd like me to work on? Do we still need to create the logging interfaces?[17:29:43] <tdiesler> dmlloyd, hmm avoiding the lock was not on purpose. Maybe you could pull as is and we move it back iside the lock in CR3. I'd need to verify this against the TCK[17:30:23] <dmlloyd> yeah I pulled it.[17:30:34] <dmlloyd> I don't think the lock is super important[17:30:40] <dmlloyd> in fact it may be better this way[17:31:08] <tdiesler> dmlloyd, it might even be ok outside the lock. dependes on the contract. exception handling I'm also not too sure about[17:33:57] <tdiesler> dmllock, a failure in either hook will result in CNFE. Perhaps the hooks should throw ClassNotFoundException.[17:35:12] <tdiesler> dmlloyd, never mind we can revisit if there is an issue[17:36:57] <dmlloyd> tdiesler: okay yeah let's revisit... you may be right that CNFE is better...[17:37:03] <dmlloyd> jamezp: okay so about logging[17:37:21] <jamezp> dmlloyd: Yes?[17:37:43] <dmlloyd> jamezp: we're chatting about it right now and I think we've agreed that we probably need 6-digit codes rather than 5[17:37:55] <dmlloyd> just looking forward, since we really can never reuse a code once it has been allocated[17:38:08] <jamezp> Right, that makes sense.[17:38:59] *** pilhuhn has quit IRC[17:40:57] *** alexsmirnov has joined #jboss-as7[17:43:32] *** galderz has quit IRC[17:44:11] *** stliu has joined #jboss-as7[17:44:11] *** ChanServ sets mode: +v stliu[17:46:44] <asoldano> mmoyses, ok, it seems we'll need that system property support, as for instance the path above won't work if I start the AS with ./bin/standalone.sh instead of ./standalone.sh[17:46:51] <asoldano> mmoyses, should I create a jira for that?[17:47:52] <dmlloyd> jamezp: and then what we'll probably do is start i18n of a subsystem, something small[17:48:08] <asoldano> mmoyses, alternatively the path should be allowed to be relative to where the domain conf file is ...?[17:48:11] <dmlloyd> then when it's done we'll look at it and see if we can derive some best practices and that sort of thing[17:48:20] <asoldano> mmoyses, as if was on AS6 with login-config.xml, I think[17:48:25] <asoldano> s/if/it/[17:48:38] <jamezp> dmlloyd: Sounds good to me.[17:48:48] <dmlloyd> jamezp: transactions is probably a good one to start at[17:48:53] <mmoyses> asoldano: in AS6 conf/ was in the classpath. we don't have any directories in the classpath now[17:49:27] <mmoyses> asoldano: i'm about to ask for a pull request. i'll make sure the system property parsing is there[17:49:32] <hbraun> bstansberry: In case you run out of work. Here's another interesting one: https://issues.jboss.org/browse/AS7-762[17:49:34] <jbossbot> jira [AS7-762] Provide an operation to request service status report after server has been started [Open (Unresolved) Feature Request, Major, Brian Stansberry] https://issues.jboss.org/browse/AS7-762[17:49:36] *** jcosta has quit IRC[17:49:37] <asoldano> mmoyses, cool, thanks[17:50:47] <bstansberry> hbraun: k[17:51:32] <bstansberry> makes sense[17:51:52] <jamezp> dmlloyd: Sounds good. That looks small enough. I'm currently working on logging tool now for overloaded method support.[17:51:54] <hbraun> yeah, it#s a weak spot atm[17:52:54] *** pferraro has quit IRC[17:53:14] <dmlloyd> jamezp: great.[17:53:25] <dmlloyd> that's going to be important I think.[17:53:59] <jamezp> I think so too. I got a good start on my flight to Raleigh.[17:59:48] *** smcgowan has joined #jboss-as7[17:59:48] *** ChanServ sets mode: +v smcgowan[17:59:50] *** mlinhard has quit IRC[18:03:49] <hbraun> bstansberry: you know what's funny?[18:04:08] <hbraun> bstansberry: now would be the first time I could make use of the rollback operations[18:04:28] *** kcbabo has joined #jboss-as7[18:04:28] *** ChanServ sets mode: +v kcbabo[18:04:33] <bstansberry> well, they aren't dead yet[18:04:51] <bstansberry> but they are on failing life support[18:05:04] <bstansberry> what's the usage?[18:06:14] <hbraun> {"New missing/unsatisfied dependencies" => ["service jboss.binding.http2 (missing)"][18:06:30] <hbraun> creating a connector w/o the socket binidng in place[18:06:56] *** fbricon is now known as fbricon_out[18:09:12] <hbraun> bstansberry: actually this brings up another question[18:09:16] *** stliu has quit IRC[18:09:22] <hbraun> in case of a failing operation[18:09:35] <hbraun> should the resource be written or not?[18:09:50] <hbraun> who's responsibility to clean up?[18:10:12] <hbraun> in my example the operation fails, but the resource is still persistted[18:10:16] <bstansberry> the default is to rollback, and it's the server's responsibility to do it[18:11:00] *** slaboure has joined #jboss-as7[18:11:06] <bstansberry> so you get "outcome" => "failed" but it persists?[18:11:44] <hbraun> well it exists in memory at least. let me check the xml[18:12:13] <hbraun> yeah, it#s written to disk[18:12:23] *** vtunka has quit IRC[18:13:13] <hbraun> general problem or error in the operation handler?[18:16:53] *** slaboure has quit IRC[18:19:31] *** kcbabo is now known as babo_afk[18:23:39] *** hbraun has quit IRC[18:26:17] *** frainone_lunch is now known as frainone[18:27:22] <asoldano> mmoyses, sorry, I have another question: is there currently a way of getting the AuthenticationManager on AS7? I used to do a jndi lookup on AS6 that's failing on AS7...[18:29:11] <mmoyses> asoldano: jndi lookup still works with java:jboss/jaas/<secDomain> but preferably you should inject the appropriate SecurityDomainService when possible[18:29:37] <mmoyses> asoldano: this service has a SecurityDomainContext as the value where you can get the authentication manager[18:30:19] <asoldano> mmoyses, interesting, thanks. Let me see how I can deal with this...[18:30:35] <mmoyses> asoldano: see WarDeploymentProcessor for an example of the service being injected in the Realm service[18:31:16] <bstansberry> hbraun: sounds like a general problem[18:32:00] <ALR> emuckenhuber: Ping. Got a support issue w/ PS needing some quick resolve.[18:32:58] <bstansberry> there's a known issue we are chasing where service problems are being detected too late[18:33:27] <emuckenhuber> ALR: hehe, ok[18:33:36] <ALR> emuckenhuber: :) #jbossas?[18:33:57] <emuckenhuber> ALR: sure[18:34:50] *** jfclere has quit IRC[18:39:20] *** darranl has quit IRC[18:44:10] *** babo_afk has quit IRC[18:44:21] *** kcbabo has joined #jboss-as7[18:44:21] *** ChanServ sets mode: +v kcbabo[18:45:46] *** galderz has joined #jboss-as7[18:45:46] *** ChanServ sets mode: +v galderz[18:45:53] *** galderz has quit IRC[18:57:05] *** asoldano is now known as asoldano_afk[18:57:08] *** jdcasey has quit IRC[18:59:33] *** emuckenhuber has quit IRC[19:00:41] *** mmoyses is now known as mmoyses_[19:02:25] *** pmuir has quit IRC[19:09:04] <bobmcw> question... would it be silly for TorqueBox components to implement org.jboss.as.ee.component.Component and play in that world?[19:09:58] *** emuckenhuber has joined #jboss-as7[19:09:59] *** emuckenhuber has joined #jboss-as7[19:09:59] *** ChanServ sets mode: +v emuckenhuber[19:13:19] *** tdiesler has quit IRC[19:15:08] *** fbricon_out is now known as fbricon[19:17:59] *** mmoyses_ is now known as mmoyses[19:20:46] *** mgoldmann has joined #jboss-as7[19:20:46] *** ChanServ sets mode: +v mgoldmann[19:27:55] <bstansberry> hbraun: ok, I bet I know the issue. you are seeing the op apply successfully to the domain, but it fails on all servers[19:29:36] *** mgoldmann has quit IRC[19:29:52] *** mgoldmann has joined #jboss-as7[19:29:52] *** ChanServ sets mode: +v mgoldmann[19:31:34] *** jamezp has left #jboss-as7[19:34:37] *** jamezp has joined #jboss-as7[19:37:24] *** sannegrinovero has joined #jboss-as7[19:37:24] *** ChanServ sets mode: +v sannegrinovero[19:39:05] <mmoyses> bstansberry: can you please pull https://github.com/mmoyses/jboss-as/commit/e6d138b86572c510432235f6dbf887f46a80410c ?[19:39:06] <jbossbot> git [jboss-as] e6d138b.. Marcus Moyses AS7-763: adding list-cached-principals and flush-cache operations[19:39:08] <jbossbot> jira [AS7-763] Add operations for security domains [Open (Unresolved) Feature Request, Major, Marcus Moyses] https://issues.jboss.org/browse/AS7-763[19:39:10] <smarlow> Nihility: I'd like to merge my hibernate40_2 branch into my master and then submit a merge request to bring Hibernate 4.0.Alpha3 into as7. I'll then bring Hibernate 4.0.0.Beta1 in on June 1st (APIs will be frozen) unless something comes up to change the plan.[19:39:46] <smarlow> Nihility: does that make sense?[19:40:19] <bstansberry> mmoyses: ok[19:40:42] <mmoyses> bstansberry: thanks[19:41:25] *** Jaikiran has quit IRC[19:48:15] <bstansberry> mmoyses: did you have another commit on that branch?[19:48:48] <mmoyses> bstansberry: i do have but it should not be pulled yet[19:49:03] <bstansberry> don't send me that kind of thing please[19:49:08] <bstansberry> send me a clean branch[19:49:32] <mmoyses> ok. let me create a new branch...[19:49:49] <bstansberry> it's ok, just for the future[19:49:58] <mmoyses> sorry[19:50:43] <mmoyses> out of curiosity, is it because merging is easier than cherry picking?[19:51:39] <dmlloyd> we'll just say "yes"[19:51:40] <dmlloyd> :)[19:52:16] <mmoyses> :)[19:52:37] <bstansberry> mostly it's because I don't want to test stuff that hasn't been tested independent of other changes[19:52:49] <mmoyses> i see[19:57:24] <jbossbot> git [jboss-as] push master bac5c8c.. Marcus Moyses AS7-763: adding list-cached-principals and flush-cache operations[19:57:25] <jbossbot> jira [AS7-763] Add operations for security domains [Open (Unresolved) Feature Request, Major, Marcus Moyses] https://issues.jboss.org/browse/AS7-763[19:57:25] <jbossbot> git [jboss-as] push master URL: http://github.com/jbossas/jboss-as/compare/442f752...bac5c8c[19:57:49] *** bstansberry has quit IRC[19:59:56] <mmoyses> asoldano_afk: system properties now work for UserRoles login module. use ${jboss.home.dir}[20:00:05] <asoldano_afk> mmoyses, thank you![20:00:13] *** asoldano_afk is now known as asoldano_dinner[20:01:33] *** rmaucher has quit IRC[20:02:39] *** jdcasey has joined #jboss-as7[20:04:04] *** sannegrinovero has quit IRC[20:07:53] *** smcgowan has quit IRC[20:09:51] *** tdiesler has joined #jboss-as7[20:09:51] *** ChanServ sets mode: +v tdiesler[20:11:26] *** ALR has quit IRC[20:12:16] *** torben has quit IRC[20:14:33] *** frainone is now known as frainone_away[20:14:33] *** jamezp has quit IRC[20:24:25] *** jamezp has joined #jboss-as7[20:25:03] <bobmcw> dmlloyd: question about ee.component stuff...[20:25:18] <bobmcw> in torquebox, we have components, like scheduled jobs, or services[20:25:32] <bobmcw> each is uniquely named, but backed by a RubyJava.class or a RubyService.class in java[20:25:50] <bobmcw> EEModuleDescription wants every component to have a unique class and a unique name[20:26:15] <bobmcw> I'd really like (I think?) to add 5 components backed by the same java class, but with some uniqueness in their config[20:26:26] <bobmcw> ie, the name of the ruby file that really powers it[20:26:37] <bobmcw> am I doing it wrong, or do I just not fit into this model?[20:27:45] <bobmcw> Nihility: or do you have any wisdom?[20:28:03] <baileyje> bobmcw: I think it should be able to work.[20:28:10] <baileyje> Not sure if it does now though[20:28:39] <baileyje> Since the EE world has views and are generally a specific class, but you should be able to use the name to make it unique.[20:29:01] <bobmcw> unfortunately the componentsByClassName Map freaks out[20:29:47] <baileyje> bobmcw: Yeah. What I meant, was it should be possible, but currently isn't. We may need to look into it.[20:30:04] <baileyje> I am guessing dmlloyd and Nihility are at lunch right now.[20:30:27] *** maxandersen has quit IRC[20:31:50] <bobmcw> baileyje: viewClass is typically a java interface?[20:32:16] <baileyje> yeah typically. But could be a class for managed beans and no-interface ejbs[20:32:30] <bobmcw> 'k[20:32:31] <baileyje> and servlets for that matter[20:33:03] <bobmcw> and if I conform to all of this component interface/impl stuff, what does that buy me, really? Do I get easier remoting? @Inject'ability into other Java bits?[20:33:20] <baileyje> wolfc: There?[20:33:35] <bobmcw> or should I just juggle msc-level Service bits?[20:33:37] <wolfc> baileyje: yes[20:33:52] <baileyje> bobmcw: Correct. You get EE based injection out of the box. As well as easy jndi binding, etc[20:34:05] <baileyje> wolfc: Did you see my change for AS7-254[20:34:07] <jbossbot> jira [AS7-254] Undeclared checked exception is thrown from EJB invocation [Open (Unresolved) Bug, Critical, John Bailey] https://issues.jboss.org/browse/AS7-254[20:34:24] <bobmcw> baileyje: 'k, I like that. I'll make the effort, and nag dmlloyd about this uniqueness stuff[20:34:31] <bobmcw> thanks![20:34:44] <bbrowning> was that a yes to easier remoting as well? that would be awesome![20:34:56] <wolfc> bobmcw, baileyje: ps. there can be different components with the same class in one module[20:35:11] <bobmcw> wolfc: I'm failing to figure out how[20:35:11] <baileyje> wolfc: There seems to be an issue though.[20:35:24] <wolfc> if not => bug[20:35:24] <baileyje> I think it is allowed by spec, but not by our current impl[20:35:33] <bobmcw> https://gist.github.com/d86210d06ad6dd7e7cf7[20:35:33] <wolfc> baileyje: where is that change?[20:35:42] <bobmcw> I'm looking at Beta3, mind you[20:35:46] *** maxandersen has joined #jboss-as7[20:35:46] *** ChanServ sets mode: +v maxandersen[20:35:47] <baileyje> Right now we have a 1:1 mapping of class to component[20:35:53] <bobmcw> that's from EEModuleDescription[20:36:04] *** opalka has joined #jboss-as7[20:36:04] *** opalka has joined #jboss-as7[20:36:04] *** ChanServ sets mode: +v opalka[20:36:09] <baileyje> wolfc: Yeah. Let me look at the ejb branch and see if it is still an issue[20:36:22] *** jdcasey has left #jboss-as7[20:36:29] <baileyje> bobmcw: Checking to see if it is fixed in the latest component work[20:36:35] <wolfc> baileyje: where is that change for AS7-254?[20:36:36] <bobmcw> 'k, thanks[20:36:36] <jbossbot> jira [AS7-254] Undeclared checked exception is thrown from EJB invocation [Open (Unresolved) Bug, Critical, John Bailey] https://issues.jboss.org/browse/AS7-254[20:37:18] <wolfc> I think I saw a signature change somewhere. Or was that in the CDI codebase.[20:37:27] <baileyje> wolfc: There are two. One in jboss-invocation -- https://github.com/baileyje/jboss-invocation/commit/f90379281298227709146dbc07897ea40a5c4c34[20:37:28] <jbossbot> git [jboss-invocation] f903792.. John E. Bailey Add an initial interceptor capable of handling undeclared exceptions[20:38:20] <baileyje> wolfc: and the corresponding change in AS --https://github.com/baileyje/jboss-as/commit/b79f31cb08380c41b37bb9b450341af03b6abc04[20:38:21] <jbossbot> git [jboss-as] b79f31c.. John E. Bailey [AS7-254] - Make sure UndeclaredThrowable is thrown for any exception thrown by an interceptor that is not declared by the intercepted method[20:38:21] <jbossbot> jira [AS7-254] Undeclared checked exception is thrown from EJB invocation [Open (Unresolved) Bug, Critical, John Bailey] https://issues.jboss.org/browse/AS7-254[20:39:08] <baileyje> bobmcw: Looks like it would still be an issue with the latest component changes. Still required unique class names[20:39:23] <bobmcw> and my use-case isn't unreasonable?[20:39:34] *** pferraro has joined #jboss-as7[20:39:34] *** ChanServ sets mode: +v pferraro[20:39:45] <baileyje> bobmcw: no. Your use case makes sense. I think you can do the same thing in EJB[20:40:04] <baileyje> one class multiple beans. Correct wolfc?[20:40:20] <baileyje> they just need separate names and jndi bindings.[20:40:35] <wolfc> yes, one class can be used in multiple components[20:40:47] <baileyje> So it is really a bug at this point.[20:41:04] <baileyje> I think the EEModuleDescription needs to have a list of components for each class name[20:42:09] <wolfc> baileyje: why did you remove testCCE?[20:42:36] <baileyje> Oh. I there was an issue I still need to talk to you about.[20:43:05] <baileyje> The EJBException is not declared on the bean method, correct?[20:43:28] <wolfc> correct[20:43:53] <baileyje> It is a RuntimeException correct?[20:43:56] <wolfc> yes[20:44:00] <baileyje> For somereason that wasn't getting thrown[20:44:34] <baileyje> Is there an EJB interceptor that should be catching and rethrowing as EJBException?[20:45:00] <wolfc> are you using the new ejb branch?[20:45:07] <baileyje> yeah[20:45:14] <baileyje> Let me add it back in a test.[20:45:18] <wolfc> the interceptors are not properly in place there[20:46:11] <baileyje> wolfc: ok. What is the plan for adding them back in properly?[20:46:43] <wolfc> Jaikiran and I are figuring out where to put them.[20:47:31] <wolfc> what I'm unsure of (and I think it's out of scope) is whether the causedBy should actually be a CCE or a UndeclaredT[20:48:02] <baileyje> ahhh. Not sure on that one[20:49:37] <wolfc> interceptor spec doesn't say anything sensible, so I think the causedBy is implementation specific. We need a check to make sure nothing else is going wrong.[21:06:01] *** jamezp is now known as jamezp_afk[21:09:57] *** asoldano_dinner is now known as asoldano[21:15:30] *** fbricon has left #jboss-as7[21:24:22] *** asaldhan has joined #jboss-as7[21:24:22] *** ChanServ sets mode: +v asaldhan[21:27:56] *** dimitris_jboss has joined #jboss-as7[21:27:56] *** dimitris_jboss has joined #jboss-as7[21:27:56] *** ChanServ sets mode: +v dimitris_jboss[21:32:04] *** dimitris_ has quit IRC[21:38:22] <stuartdouglas> morning[21:41:20] *** tdiesler has quit IRC[21:47:25] *** irooskov has joined #jboss-as7[21:52:48] <asaldhan> dmlloyd: http://community.jboss.org/thread/166646[21:53:27] <dmlloyd> hey stuart[21:53:39] <dmlloyd> two words: JSF injection! :)[21:54:14] <stuartdouglas> I was thinking about that yesterday, it's one of the failures I had on the CDI TCK[21:54:16] <asaldhan> stuartdouglas: interesting that you are already our future. 12+ hours ahead. :)[21:54:43] <stuartdouglas> nearly 6:00am here[21:54:58] <dmlloyd> stuartdouglas, it should be pretty straightforward if a bit ugly[21:55:29] <dmlloyd> we'll have to do redundant annotation scanning+dd processing for the JSF crap so we know to track it and create injection interceptor chains for it[21:55:44] <dmlloyd> and there's the issue of locating a deployment from a JSFMB instance[21:55:54] <dmlloyd> once those are solved it should be pretty simple[21:56:23] <stuartdouglas> ok, I will have a look today[21:56:53] <stuartdouglas> if I get that working CDI TCK failures will be down in the single digits[21:58:37] <stuartdouglas> I'm going to head to the gym in a few minutes, then I am heading up to the Sydney office, so I might not be online that much for the next few hours[22:02:15] *** jamezp_afk is now known as jamezp[22:03:58] *** smcgowan has joined #jboss-as7[22:03:58] *** ChanServ sets mode: +v smcgowan[22:04:06] *** jpederse has quit IRC[22:04:37] *** opalka has quit IRC[22:05:32] <dmlloyd> okay thanks stuartdouglas[22:06:13] *** stansilvert has joined #jboss-as7[22:23:52] *** frainone_away is now known as frainone[22:37:23] *** mgoldmann has quit IRC[22:38:14] *** maeste has quit IRC[22:39:00] *** dimitris_jboss has quit IRC[22:39:27] *** asoldano has quit IRC[22:44:38] *** mgoldmann has joined #jboss-as7[22:44:38] *** ChanServ sets mode: +v mgoldmann[22:53:16] *** ALR has joined #jboss-as7[22:53:16] *** ChanServ sets mode: +v ALR[22:55:46] <bobmcw> dmlloyd: hey, didja see history wrt uniqueness of classnames of components in a module?[22:57:57] *** wolfc has quit IRC[22:59:13] <dmlloyd> jsut chatted about that with john[22:59:23] <dmlloyd> curious though, how did you run into that?[22:59:34] *** mgoldmann_ has joined #jboss-as7[22:59:34] <bobmcw> theoretical machinations, so far[22:59:48] <bobmcw> but added two RubyAppDescriptions to a Module[22:59:49] <dmlloyd> we'll definitely make sure to address that[22:59:51] <bobmcw> ModuleDesc[23:00:01] <bobmcw> I'm trying to make my stuff masquarade as bonafide ee componentry[23:00:07] <dmlloyd> ah I see[23:00:15] <dmlloyd> ERB instead of EJB? :)[23:00:17] <bobmcw> I'm told that helps with remoting, jndi, and injectability[23:00:34] <bobmcw> if I conform to the Component((Instance)Stuff)[23:00:53] *** smarlow has quit IRC[23:01:00] <dmlloyd> yeah it's a cool idea[23:01:17] <bobmcw> then be able to @Inject @RubyThing(name='tacos')[23:01:18] <bobmcw> or somesuch[23:01:27] <bobmcw> will be minimal interfaces, but at least a bridge[23:01:39] *** kcbabo has quit IRC[23:01:39] *** maxandersen is now known as max_crashed[23:01:41] *** mgoldmann has quit IRC[23:01:42] <bobmcw> or even just for remoting purely from ruby->ruby[23:01:50] *** kkhan has quit IRC[23:02:10] <bobmcw> dmlloyd: can you point me to some information about this ComponentView stuff? I'm not sure what to do with that[23:02:29] <dmlloyd> well... every EJB has views, like a remote interface or local interface[23:02:45] <bobmcw> I can ignore that, and still benefit from all of this stuff?[23:02:52] <bobmcw> is an MBean considered a view?[23:02:55] <dmlloyd> if you want to actually be able to generate something EJB-ish then it has to act like an EJB[23:03:08] <dmlloyd> you mean ManagedBean? no, those cannot be remotely invoked[23:03:22] * dmlloyd notes that JMX MBean != ManagedBean[23:03:29] <bobmcw> is EJB our only remoting thing at the moment?[23:03:31] <dmlloyd> to the confusion of all[23:03:33] <bobmcw> no remote POJO capabilities?[23:03:39] <bobmcw> or is that EJB3?[23:03:42] <dmlloyd> well we really have no remote anthing yet[23:03:44] * bobmcw doesn't know ejb3[23:03:58] <dmlloyd> EJB will definitely be remotable though[23:04:10] <dmlloyd> anything else has to be thought out a little[23:04:29] <bobmcw> 3-6wks until CR1? For reals?[23:04:33] <dmlloyd> :D[23:05:10] <dmlloyd> we've pulled off crazier feats, just fyi :)[23:07:43] <bobmcw> also, why all the Description/Configuration/Component/ComponentInstance hierarchy? Every example I'm finding all seems to be about the same boiler-plate[23:07:47] * bobmcw is wary of boilerplate[23:08:01] <bobmcw> seems like a desc just creates a config just to create a component[23:08:07] <dmlloyd> yes[23:08:11] <dmlloyd> the desc is pre-classloading[23:08:28] <dmlloyd> the config is post-class loading and also adds some more stuff at that time[23:08:34] <dmlloyd> then the components get created from that[23:08:44] <dmlloyd> it all changes a little after the big merge tomorrow[23:08:52] <bobmcw> so, let's say I have a ScheduledJob in my world, that needs a lifecycle and injection[23:09:03] <bobmcw> I'd been thinking of a Service<ScheduledJob> and injecting my ruby bits and stuff into that[23:09:12] <bobmcw> but now I'm thinking ScheduledJobComponent** stuff[23:09:14] <dmlloyd> that's a reasonable approach[23:09:26] <dmlloyd> what's the motivation for going to a component?[23:09:30] <dmlloyd> it is substantially more complex[23:09:42] <bobmcw> my issue/concern, is I want to be able to inject from JNDI/CDI/other-components into my Service<ScheduledJob>[23:09:44] <dmlloyd> and it's a bit of a slave to the EE specs[23:10:07] <dmlloyd> oaky well the beauty of the way we're doing EE components is that you can inject EE stuff and JNDI stuff into any MSC service[23:10:20] <dmlloyd> I am not sure about CDI though, or what its relationship to services are[23:10:28] <dmlloyd> you'd need to check with stuartdouglas[23:10:30] <bobmcw> and if I have some HASingleton-type ruby service on node #1, I'd love to be able to inject some remote reference to it on node #2[23:10:52] <bobmcw> dmlloyd: how do I inject EE or JNDI into Service<T>?[23:11:06] <bobmcw> I think I found a way to get various naming contexts injected, but then I still poll?[23:11:13] <dmlloyd> the type of any JNDI binding is a ManagedReferenceFactory[23:11:26] <dmlloyd> all you do is inject that, then get a new ManagedReference out of it[23:11:31] <dmlloyd> from that you can get the actual instance[23:11:42] <dmlloyd> just make sure you free it when you're done with it[23:11:49] <bobmcw> how do I get the MRFactory to start?[23:11:56] <dmlloyd> inject it MSC-style[23:12:08] <bobmcw> yes, but I dunno what ServiceName is[23:12:11] <bobmcw> I guess[23:12:24] <dmlloyd> the service name can be calcualted from a JNDI name using the ContextNames class[23:12:46] <dmlloyd> (there's two right now, you have to find the right one - we fixed that in tomorrow's merge though)[23:13:03] <dmlloyd> basically every valid JNDI name has an equivalent service name[23:13:08] <bobmcw> serviceName = contextServiceNameOfApplication( "comp/foo/whatever/in/jndi" )[23:13:12] <bobmcw> or whatnot?[23:13:16] <dmlloyd> yeah more or less[23:13:48] <bobmcw> if I need a jms.Queue, is that a magic injection, or just a jndi lookup also?[23:14:04] <dmlloyd> since your stuff is not properly a component, you can access any of the "java:jboss", "java:global", "java:app", or "java:module" namespaces[23:14:14] <dmlloyd> JNDI lookup is probably the easiest[23:14:30] <bobmcw> yah, for CDI, all I ended up doing was dipping down via java:global and finding their BeanManager[23:14:33] <dmlloyd> though most things that are bound into JNDI *also* are Service<MRF>[23:14:34] *** mmoyses has quit IRC[23:14:37] <bobmcw> but that was also poll-ish, not true injection[23:15:15] <dmlloyd> the thing with JNDI bindings is, if you inject something from JNDI and its corresponding service is shut down, all the dependents (including you) are shut down first[23:15:27] <dmlloyd> that way if someone replaces some resource everything restarts with the replacement[23:16:39] <dmlloyd> I'm not sure CDI works the same way because their lifecycles are different[23:16:44] <dmlloyd> stuart will be able to help you there[23:17:06] <bobmcw> 'k[23:17:26] <bobmcw> if I go Service<T>, while I can inject in, I'm not necessarily injectable via @Inject back into java land[23:17:33] <bobmcw> but that's fine, if I can skip a lot of this boiler-plate[23:18:18] <dmlloyd> you may be, in the end[23:18:39] <bobmcw> allrighty[23:18:42] <bobmcw> screw ee![23:18:43] <dmlloyd> I'm not super-clear on how CDI does its stuff but a while back we talked about a @ServiceNamed annotation so you can have:[23:18:55] <dmlloyd> @Inject @ServiceNamed("org.jboss.foobar") void setFooBar(...[23:19:08] <dmlloyd> so that's an option at least[23:26:24] *** sannegrinovero has joined #jboss-as7[23:26:24] *** sannegrinovero has quit IRC[23:26:24] *** sannegrinovero has joined #jboss-as7[23:26:24] *** ChanServ sets mode: +v sannegrinovero[23:32:31] *** kkhan has joined #jboss-as7[23:32:38] *** ChanServ sets mode: +v kkhan[23:33:45] *** kkhan has quit IRC[23:35:52] *** kkhan has joined #jboss-as7[23:35:52] *** ChanServ sets mode: +v kkhan[23:39:06] *** kkhan has quit IRC[23:51:51] <misty> fyi, I am still working on the RESTeasy + AS6/7 problem, and #jbosstools has now sent me to Bill Burke :/[23:53:44] <smcgowan> misty: stuartdouglas did the resteasy int in AS 7[23:54:26] <misty> oh excellent, and he is in AU so I should be able to catch him[23:54:45] <misty> but I Think he is the one who asked me to check #jbosstools[23:54:50] <misty> maybe I should just file a bug against JBDS[23:55:18] *** mbg has quit IRC[23:56:55] *** adietisheim has quit IRC[23:58:23] <sguilhen> smcgowan: ping, do I need to use an specific maven version for the specs project?[23:58:52] <smcgowan> sguilhen: are you having a problem building?[23:59:36] <sguilhen> smcgowan: yea, I've added the new module to jboss-javaee-6/pom and when I try to run the build from jboss-spec-parent I get an error saying the new module is missing the version declaration[23:59:59] <smcgowan> sguilhen: don't worry about that pom, i'll take care of the parent and the aggregate