[00:03:21] *** rruss has joined #jbosstesting
[00:04:22] *** michaelschuetz has quit IRC
[00:24:50] *** ldimaggi has joined #jbosstesting
[00:36:09] *** aslak has quit IRC
[00:36:18] *** ALR has quit IRC
[01:30:38] *** lightguard_jp has quit IRC
[01:50:42] *** rruss has quit IRC
[02:54:50] *** rruss has joined #jbosstesting
[04:18:40] *** jdlee has quit IRC
[04:37:02] *** jdlee has joined #jbosstesting
[04:40:29] *** bgeorges has joined #jbosstesting
[05:06:49] *** ldimaggi has quit IRC
[05:30:33] *** bleathem_ has joined #jbosstesting
[05:33:45] *** bleathem has quit IRC
[05:42:06] *** bleathem_ is now known as bleathem
[05:43:19] *** rruss has quit IRC
[06:04:50] *** rruss has joined #jbosstesting
[06:05:29] *** rruss has quit IRC
[07:12:03] *** bleathem has quit IRC
[07:59:52] *** mgoldmann has joined #jbosstesting
[08:03:58] *** jharting has joined #jbosstesting
[08:50:08] *** ge0ffrey has joined #jbosstesting
[08:59:08] *** kpiwko has joined #jbosstesting
[08:59:19] *** jeand has joined #jbosstesting
[09:16:30] *** bgeorges has quit IRC
[09:32:59] *** lfryc has joined #jbosstesting
[09:39:43] *** aslak has joined #jbosstesting
[09:39:43] *** aslak has quit IRC
[09:39:43] *** aslak has joined #jbosstesting
[09:41:02] *** alesj has joined #jbosstesting
[09:42:43] *** wolfc has joined #jbosstesting
[09:48:25] <aslak> kpiwko, good morning
[09:48:33] <aslak> kpiwko, seen this ? http://community.jboss.org/message/592616#592616
[09:48:49] <kpiwko> aslak, morning
[09:50:38] <kpiwko> aslak, yes, I have unfortunately I wasn't able to dig out the reason why SeleniumConfiguration is not created by simple repetition...I'll look at it during weekend, I've more priority things in the queue now
[09:51:03] <aslak> kpiwko, aa, good.. .)
[09:51:04] <aslak> :)
[09:51:37] <nickarls> mornings
[09:51:45] <aslak> nickarls, morning
[09:54:18] *** avogt_ has joined #jbosstesting
[10:08:42] *** Jaikiran has joined #jbosstesting
[10:16:29] *** avogt has joined #jbosstesting
[10:32:17] *** michaelschuetz has joined #jbosstesting
[10:46:14] *** avogt has quit IRC
[11:00:21] *** michaelschuetz has quit IRC
[11:02:04] *** michaelschuetz has joined #jbosstesting
[11:49:30] *** Elisha has quit IRC
[11:49:55] *** Elisha has joined #jbosstesting
[11:55:04] *** michaelschuetz has quit IRC
[11:55:44] *** avogt_ has quit IRC
[11:55:50] *** michaelschuetz has joined #jbosstesting
[12:01:50] *** michaelschuetz has left #jbosstesting
[12:01:59] *** michaelschuetz has joined #jbosstesting
[12:36:08] *** stuartdouglas has joined #jbosstesting
[12:53:32] *** bgeorges has joined #jbosstesting
[13:12:26] *** avogt has joined #jbosstesting
[13:20:27] *** ldimaggi has joined #jbosstesting
[13:32:47] *** lfryc has quit IRC
[14:14:15] *** echelog-2` has joined #jbosstesting
[14:20:13] *** echelog-2 has quit IRC
[14:27:36] *** avogt has quit IRC
[14:35:48] *** mgoldmann is now known as mgoldmann|away
[15:13:28] *** rruss has joined #jbosstesting
[15:24:38] *** bgeorges_ has joined #jbosstesting
[15:24:38] *** bgeorges has quit IRC
[15:33:29] *** bgeorges_ has quit IRC
[15:36:58] *** alesj has left #jbosstesting
[15:40:16] *** kpiwko has quit IRC
[15:57:45] <bobmcw> hey guys, I'm getting some RMI-related SocketTimeoutException during my ARQ test runs against a managed AS6
[15:57:53] <bobmcw> my code doesn't touch RMI, as far as I know
[15:57:59] <bobmcw> wasn't sure if this was how ARQ talks to a managed AS?
[15:58:32] <bobmcw> aslak: hi!
[15:58:50] <aslak> bobmcw, heya
[15:59:02] <aslak> bobmcw, pastebin a stacktrace
[15:59:12] <bobmcw> no stack, just console output, inside my windows vm
[15:59:14] <bobmcw> lemme see if I can copy
[16:00:01] <bobmcw> https://gist.github.com/865977
[16:00:16] <bobmcw> the destroying is me hitting ^C
[16:00:22] <bobmcw> it sorta hung at that point
[16:00:38] <bobmcw> that test works if run one-off via -Dtest=*FormHandlingTest
[16:00:56] <aslak> does it actually fail your tests?
[16:01:00] <bobmcw> which is why I'm wondering if it's passing, and then things freak out when reporting back, or attempting to undeploy
[16:01:19] <bobmcw> aslak: I was too impatient to wait, lemme try again, I'll ping in 30min :)
[16:01:44] <aslak> looks similar to some managed container shutdown i've seen before, but it's basically the RMI MBean protocol that 'times' out during shutdown i think
[16:01:47] <bobmcw> aslak: I guess, my question is, does ARQ use RMI to chat with a managed AS6?
[16:01:55] <bobmcw> 'k
[16:01:57] <bobmcw> I'll poke around
[16:02:24] <aslak> it uses MBean Server which under the hood use RMI.
[16:02:49] <bobmcw> 'k, cool, thanks
[16:02:52] <bobmcw> I'll poke around more
[16:02:53] <aslak> i think it comes when the ServerManager kills the process during shutdown, so the MBean connection is retrying to connect
[16:38:14] <bobmcw> aslak: I think we're just facing memory issues after 15 deployments into our AS
[16:38:31] <bobmcw> I thought there was a way to ask ARQ to restart the AS after N deployments, but I can't find that?
[16:39:48] *** mgoldmann|away is now known as mgoldmann
[16:39:57] <aslak> bobmcw, yea, 2 sec
[16:40:11] <aslak> https://github.com/weld/core/blob/master/tests-arquillian/src/test/resources/arquillian.xml
[16:40:14] <aslak> maxDeploymentsBeforeRestart
[16:40:23] <bobmcw> is that inside <engine> in arq.xml?
[16:40:46] <aslak> yea
[16:40:52] <bobmcw> thanks!
[16:42:15] <aslak> reminds me i need to rename that one, maxDeploymentBeforeRestart doesn't make much sense anymore when you can have multiple deployments pr TestClass. shame to restart the server in the middle of the test.. :)
[17:05:56] *** lightguard_jp has joined #jbosstesting
[17:20:56] *** jharting has quit IRC
[17:33:54] *** ge0ffrey_ has joined #jbosstesting
[17:35:15] *** ge0ffrey has quit IRC
[17:35:52] *** ge0ffrey_ has quit IRC
[18:01:49] *** bobmcw has quit IRC
[18:15:41] *** bobmcw has joined #jbosstesting
[18:21:04] *** michaelschuetz has quit IRC
[18:21:23] *** jeand has quit IRC
[18:24:48] *** jeand has joined #jbosstesting
[18:28:03] *** jc3 is now known as jc3-is-away
[18:49:15] *** bobmcw has quit IRC
[18:51:30] *** bobmcw has joined #jbosstesting
[18:58:07] *** michaelschuetz has joined #jbosstesting
[19:51:49] *** jc3-is-away is now known as jc3
[19:57:20] *** kpiwko has joined #jbosstesting
[20:12:47] *** aslak has quit IRC
[20:13:44] *** aslak has joined #jbosstesting
[20:14:35] *** aslak has quit IRC
[20:15:15] *** aslak has joined #jbosstesting
[20:17:29] *** ALR has joined #jbosstesting
[20:17:36] *** michaelschuetz1 has joined #jbosstesting
[20:21:08] *** michaelschuetz has quit IRC
[20:21:08] *** aslak has quit IRC
[20:31:29] *** michaelschuetz has joined #jbosstesting
[20:32:35] *** Tashtego has joined #jbosstesting
[20:33:23] *** mgoldmann has quit IRC
[20:34:53] *** michaelschuetz1 has quit IRC
[20:42:46] *** ccrouch has joined #jbosstesting
[20:44:54] <ccrouch> quick question about Arquillian
[20:44:54] <ccrouch> will it let me use the relevant part of EAP5.1 as an "embedded container"?
[21:02:03] *** michaelschuetz1 has joined #jbosstesting
[21:05:30] *** michaelschuetz has quit IRC
[21:07:11] *** michaelschuetz has joined #jbosstesting
[21:08:57] *** alesj has joined #jbosstesting
[21:10:12] *** alesj has left #jbosstesting
[21:10:17] *** michaelschuetz1 has quit IRC
[21:12:51] *** michaelschuetz1 has joined #jbosstesting
[21:15:46] *** michaelschuetz has quit IRC
[21:20:04] *** michaelschuetz1 has quit IRC
[21:23:10] *** Jaikiran has quit IRC
[22:10:22] *** jdlee has quit IRC
[22:18:33] *** ALR has quit IRC
[22:44:40] *** ALR has joined #jbosstesting
[22:46:46] *** jdlee has joined #jbosstesting
[22:46:46] *** jdlee has joined #jbosstesting
[22:50:55] *** Tashtego has quit IRC
[22:52:53] <ccrouch> ALR: any thoughts? ^
[22:53:03] <ALR> ccrouch: EAP 5.1
[22:53:04] <ALR> ?
[22:53:18] <ALR> The difficult bit there is in constructing your application classpath
[22:53:23] <ALR> EAP artifacts aren't in a Maven repo
[22:54:01] <ALR> So you can't rely on Maven/Surefire to set up the classpath as we do in say, AS, where we have the POM "depchain".
[22:54:24] <ccrouch> ok, so you're talking about one or two deps here
[22:54:36] <ccrouch> sorry
[22:54:36] <ALR> Unfortunately, no.
[22:54:47] <ccrouch> you're NOT talking about one or two...
[22:54:48] <ccrouch> :-)
[22:54:50] <ALR> "depchain" exports out a boatload transitively.
[22:54:51] <ccrouch> thought so
[22:54:51] <ALR> Yeah.
[22:55:12] <ALR> Basically *everything* runtime in AS goes on the appCL to make it work Embedded
[22:55:29] <ALR> In AS7 this isn't the case, as there's module loading and forked Surefire to accommodate that
[22:55:36] <ccrouch> right
[22:55:53] <ALR> ccrouch: I have some idea though
[22:55:57] <ccrouch> if AS7 was an option i would be all of it
[22:56:02] <ALR> ccrouch: Make a Manifest JAR for EAP
[22:56:10] <ccrouch> be all *over* it
[22:56:15] <ALR> ccrouch: And in ClassPath of the manifest...
[22:56:23] <ALR> ...point to the other libs in the distro
[22:56:29] <ALR> Then just put that one JAR on the appCL
[22:56:35] <ALR> I think that should work nicely.
[22:56:50] <ALR> And in Maven it'd be one "system" dependency
[22:56:53] <ALR> ie. On the filesystem
[22:56:58] <ALR> So we get around the repo issue.
[22:57:09] <ccrouch> what would be the next blocker :-)
[22:57:24] <ALR> Technically, I think that's the only one.
[22:57:42] <ALR> EAP 5.1 is based on an AS version proven to work fine Embedded
[22:57:55] <ccrouch> ok that sounds promising
[22:58:07] <ALR> ccrouch: We can prioritize that.
[22:58:24] <ALR> ccrouch: All I need to know is the company/internal stance of adding a new JAR to the distro
[22:58:29] <ALR> Or allowing one to be plugged in.
[22:58:43] <ALR> And how that cooperates w/ support, product engineering, and the like
[22:58:59] <ALR> Let me draft an email and engage some of the decision-makers there
[22:59:20] <ALR> ccrouch: Any recommendations for who should be on that Thread?
[22:59:45] <ALR> Bruno, Sharples, Dimitris, you, me, ... who else?
[22:59:48] <ALR> Anne-Louise?
[23:00:00] <ccrouch> ALR this is still super preliminary right now
[23:00:11] <ALR> ccrouch: I'll want it for JBW.
[23:00:39] <ALR> Customers deserve the same type of usability support we're baking in the community too.
[23:00:45] <ALR> IMO
[23:01:19] <ccrouch> if you're going to do this anyway then i'll watch :-), but its I don't have enough business driver to push this right now
[23:01:22] <ccrouch> thats very true
[23:01:33] <ALR> ccrouch: The driver is somewhere, of that I'm sure.
[23:01:45] <ALR> PM will be able to advise further.
[23:01:59] <ALR> So I just need to know who to include in the mail
[23:02:09] *** aslak has joined #jbosstesting
[23:02:09] <ccrouch> yeah as long as its pluggable, i.e. we dont put EAP jars on the web, it should be fine i imagine
[23:02:09] *** aslak has quit IRC
[23:02:10] *** aslak has joined #jbosstesting
[23:02:23] <ccrouch> you can cc me
[23:02:28] <ALR> Sure.  Who else?
[23:02:35] <ALR> RichS is enough from PM?
[23:02:37] *** aslak has quit IRC
[23:03:04] *** kpiwko has quit IRC
[23:03:11] <ccrouch>  I would include Bruno and Dimitris, at least for the heads up
[23:03:16] *** aslak has joined #jbosstesting
[23:04:16] <ccrouch> ALR: so back to the tech :-)
[23:04:16] <ccrouch> what we discussed above could act as a replacement for all the /jboss-embedded-all-beta3.SP9.jar cruft ?
[23:05:41] <ALR> ccrouch: That thing is ancient.
[23:05:59] <ALR> ccrouch: And I'm hesitant to fully "support" Embedded in EAP.
[23:06:07] <ALR> As the whole stack has not been tested in Embedded mode
[23:06:25] <ALR> My standup integration tests for EJB, JMS, Web work.
[23:06:35] <ALR> But they don't begin to encompass the full AS testsuite.
[23:11:17] <ccrouch> (4:03:38 PM) ALR: ccrouch: That thing is ancient.
[23:11:18] <ccrouch> true, but I know we're still using it for EJB and possibly JMS testing, and I know other projects that are using it too :-/
[23:11:18] <ccrouch> (4:03:56 PM) ALR: ccrouch: And I'm hesitant to fully "support" Embedded in EAP.
[23:11:18] <ccrouch> if you could support atleast as much as the old stuff did that would probably be sufficient for a start, e.g. EJB/JMS
[23:11:55] <ALR> ccrouch: To support "EJB" means to support all code paths also tested by the standalone mode.
[23:12:14] <ALR> And without a suite that does the *exact same thing* as the existing one, only in Embedded mode, I can't guarantee that.
[23:12:22] <ALR> In theory it should all work fine.
[23:12:43] <ALR> Unless we come across a code path in a subproject where they make some ClassLoading assumption.
[23:12:56] <ALR> (Which I'd come across before when making AS Embeddable in the first place)
[23:15:44] <ccrouch> gotya
[23:17:12] <ccrouch> but wouldn't the old stuff have similar problems?
[23:17:12] <ccrouch> i'm trying to determine whether Arq would be a suitable replacement/improvement over beta3.SP9.jar etc
[23:17:13] <ALR> ccrouch: Sent.
[23:17:26] <ALR> ccrouch: That old JAR you have is a *forked* AS.
[23:17:54] <ALR> ccrouch: It's frozen in time.
[23:18:09] <ALR> And doesn't look anything like EAP will
[23:18:09] <ccrouch> right, so I'm thinking Arq has got to be better than that :-)
[23:18:15] <ALR> Yes, much.
[23:18:31] <ALR> But ARQ Embedded just won't be as heavily tested as say, EAP standalone.
[23:18:44] <ALR> So what we'd have would be a marked improvement.
[23:19:10] <ccrouch> right, but its really comparing ARQ to beta3.SP9.jar etc
[23:19:22] <ccrouch> not to standalone EAP
[23:19:34] <ccrouch> in the embedded container test case
[23:20:26] <aslak> just shooting in my 2 cents here.. i would go with Managed over Embedded for EAP 5.1. Managed being controlled by arq but a remote jvm. it's much safer on every level then embedded, and since arq is controlling the start up of the container, and is testing in container. it gives you what embedded would without the extra 'hassle'
[23:21:13] <ALR> aslak: Yes, but he asked about Embedded :)
[23:21:20] <ALR> aslak: At the least we also need the managed case.
[23:21:27] <ALR> Which we've seen work.
[23:21:31] <ccrouch> true ALR
[23:22:12] <aslak> ALR, sure, but is he really asking for Embedded, or is the use case startup control.. which I would say is 99% of the usecase for embedded
[23:22:29] <aslak> ccrouch, didn't mean to talk about you in 3. person.. hehe
[23:22:36] <ALR> Definitely not 99% from my perspectve :)
[23:22:51] <ccrouch> :-)
[23:22:52] <ccrouch> aslak: what are the startup/footprint differences between Managed and Embedded wrt EAP5.1 ?
[23:22:57] <ALR> I love Embedded stuff because I can share a VM and get all sorts of awesome shared memory access :)
[23:23:10] <ALR> ccrouch: A new process is Managed
[23:23:12] <aslak> ALR, you get the same with incontainer testing
[23:23:16] <ALR> Not too different in terms of footprint
[23:23:23] <ALR> You still need the RAM.
[23:23:25] <ALR> It's not faster.
[23:23:30] <ALR> It's just a different process.
[23:23:48] <ccrouch> ALR: i would have thought Managed would have been slower
[23:23:50] <ALR> incontainer has its own cavets.  Modifying the @Deployment.
[23:23:56] <ALR> ccrouch: Why
[23:23:57] <ALR> ?
[23:24:14] <ccrouch> but maybe that just because I thought Embedded would start fewer services
[23:24:23] <ALR> ccrouch: The one you're using now does.
[23:24:30] <ALR> The current Embeddded *is* AS.
[23:24:36] <aslak> ALR, vs modifying the app classpath..  leaking state etc.. :)
[23:24:37] <ALR> Just ... in your existing VM
[23:24:37] <ccrouch> or allow fewer to be started more easily
[23:24:50] <ALR> aslak: Yep, which is why it's nice we have 3 options to offer. :P
[23:25:00] <aslak> heh
[23:25:04] <ALR> ccrouch: Nope.  That's the point I made when making Embedded.
[23:25:08] <ALR> "Embedded" is one concern.
[23:25:17] <ALR> Configuration and startup speed, those are others.
[23:25:34] <ALR> As far as I'm concerned, if it starts in an existing VM, it's "Embedded"
[23:25:41] <ccrouch> thats fair
[23:26:07] <ALR> Not to discount your very valid points.
[23:26:12] <ALR> It should be easy/fast
[23:26:18] <ALR> AS7 has the speed.
[23:26:35] <ccrouch> i wonder what the startup/footprint differences are between  beta3.SP9.jar etc and embedded/managed EAP5.1
[23:26:42] <ALR> Likely a lot.
[23:26:58] <ALR> beta3.* was stripped of services
[23:27:03] <ALR> For just that reason.
[23:27:04] <ccrouch> right, with EAP5.1 being slower ?
[23:27:07] <ccrouch> gotya
[23:27:41] <ccrouch> does ARQ support similar "stripping"?
[23:28:34] <ALR> ccrouch: By the same way EAP does.
[23:28:39] <ALR> You make a server config
[23:28:42] <ALR> Strip it all you want.
[23:28:54] <ALR> Then use Embedded or ARQ to call upon the server config you created.
[23:29:04] <ccrouch> gotya
[23:29:16] <ALR> IMO that's solving problems at the right level.
[23:29:49] <ALR> Embedded is to launch in a VM.  ARQ is to integrate with a test framework.  Server config is defined by the server.
[23:34:13] <ccrouch> makes sense
[23:38:42] <ccrouch> i replied with some clarification :-)
[23:42:45] <ALR> Thanks
[23:42:51] *** jeand has quit IRC
[23:43:26] <ALR> ccrouch: Ah sorry, I got that wrong.  Thanks for correcting.
[23:43:29] <ALR> (In the email)
[23:51:28] <ccrouch> np
[23:52:45] *** jc3 has left #jbosstesting





top