May 5, 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 5, 2011 NOTICE: This channel is no longer actively logged.
[00:01:26] <jamezp> dmlloyd: Do you have a quick moment for a pm? Just two general questions.[00:01:53] <dmlloyd> go for it[00:08:50] <jamezp> Yay! I've got mail, but it was rather painful getting there :-D[00:13:45] <stuartdouglas> has anyone else been getting random test failures on master?[00:15:06] *** magesh has joined #jboss-as7[00:16:07] *** magesh1 has quit IRC[00:22:00] *** kcbabo has joined #jboss-as7[00:22:00] *** ChanServ sets mode: +v kcbabo[00:25:28] *** adietisheim has quit IRC[00:29:33] *** rawbdor has joined #jboss-as7[00:32:49] *** kcbabo has quit IRC[00:37:55] *** jpearlin has joined #jboss-as7[00:43:46] *** miclorb has joined #jboss-as7[00:47:40] <jamezp> I haven't been getting any failures.[00:48:37] <stuartdouglas> it seems to fail for me on different tests fairly consistently[00:48:58] <stuartdouglas> I will investigate some more once I have finished this rebase[01:04:02] *** emuckenhuber has quit IRC[01:47:01] *** alexsmirnov has quit IRC[01:49:29] <dmlloyd> just have to finish dinner cleanup and kids->bed and I'll be back on the job[02:07:31] *** alesj has joined #jboss-as7[02:18:08] *** fnasser has quit IRC[02:26:53] *** jamezp is now known as jamezp_afk[02:27:20] <stuartdouglas> jamezp_afk: I am only seeing the failures on the integration tests, the smoke tests seem to be fine[02:27:34] *** lgao has joined #jboss-as7[02:28:19] <jamezp_afk> Ah, that I have seen. In fact it's been a while since I've even been able to build with the integrated tests.[02:28:37] <stuartdouglas> I have a feeling I know what it is[02:29:22] *** ccrouch has joined #jboss-as7[02:29:22] *** ChanServ sets mode: +v ccrouch[02:43:03] <stuartdouglas> I'm pretty sure it is ServerStateMonitorListener that is to blame[02:43:51] *** alesj has quit IRC[02:45:44] *** sgilda has quit IRC[02:57:55] *** clebert has joined #jboss-as7[02:57:55] *** ChanServ sets mode: +v clebert[03:02:02] *** clebert has quit IRC[03:04:13] <dmlloyd> I think the handler does what it should, but the controller doesn't have a proper relationship with it[03:04:20] <dmlloyd> I'm looking into fixing that right now[03:04:54] <stuartdouglas> Have those msc race condition's that we talked about before I went on holidays been fixed? Otherwise I would say that they were to blame[03:05:21] <dmlloyd> just hat to break for a couple hours to get five kids: fed dinner, in pajamas, brushed of teeth, and in bed, and then do laundry, dishes, and clean the house[03:05:34] <dmlloyd> yeah MSC should be solid now, at least upstream[03:05:49] <dmlloyd> I was holding off on tagging until my branch is done[03:06:53] <stuartdouglas> I have rebased my ejb branch[03:08:04] <dmlloyd> awesome.[03:09:42] <stuartdouglas> If I want to add an interceptor to a component is the correct way to add a configurator that iterates over the components methods and adds the interceptor to all the methods deque's ?[03:12:06] <stuartdouglas> One concern that I have is that it is not going to be easy to figure out interceptor order just by looking at the code, and having interceptor order dependent on processor order does not seem very nice[03:12:39] <stuartdouglas> I was thinking maybe we could do something like Phase.java but for the configurator's, so the configurators run in a well defined order?[03:15:50] <dmlloyd> can we wait and see how it looks after a few are implemented?[03:16:15] <stuartdouglas> sure[03:51:31] *** magesh has left #jboss-as7[04:01:22] *** JimMa has joined #jboss-as7[04:01:53] *** magesh has joined #jboss-as7[04:02:58] *** magesh has quit IRC[04:05:29] *** clebert has joined #jboss-as7[04:05:29] *** ChanServ sets mode: +v clebert[04:14:47] *** misty has quit IRC[04:37:08] *** jpearlin has left #jboss-as7[04:48:28] *** clebert has quit IRC[04:49:23] *** stliu has joined #jboss-as7[04:49:23] *** ChanServ sets mode: +v stliu[04:49:50] <stuartdouglas> dmlloyd: any objection to changing the missing dep report to tell you what is depending on the missing dependencies as well?[04:51:13] <dmlloyd> sure, though we might want to limit the list somehow, and make the full list available via management op or something[04:51:34] <dmlloyd> so you don't get a list of 50,000 services waiting for that one EJB to start or whatever[04:51:43] <stuartdouglas> good point[04:52:06] <stuartdouglas> I was thinking just the immediate deps[05:23:43] <dmlloyd> http://github.com/dmlloyd/jboss-as/compare/new-controllers-maybe[05:24:11] * dmlloyd really wants to show this to brian[05:25:22] *** ccrouch has quit IRC[05:26:56] *** irooskov has quit IRC[05:31:56] *** misty has joined #jboss-as7[05:47:43] *** frainone_afk has quit IRC[06:03:56] <stuartdouglas> weird[06:04:54] <stuartdouglas> I have a case where one test's deployment seems to be using the module definition for a previous test with the same archive name[06:23:12] <stuartdouglas> ah, someone has been messing with the module loader[06:44:38] *** jing has joined #jboss-as7[06:45:58] *** jing is now known as jinli[06:45:59] *** jamezp_afk has quit IRC[06:46:29] *** jamezp has joined #jboss-as7[07:02:47] *** rawbdor has quit IRC[07:32:20] *** jinli is now known as jing[07:32:35] *** jing is now known as jingli[07:37:53] *** pferraro has quit IRC[07:45:55] *** bobmcw has joined #jboss-as7[07:45:55] *** ChanServ sets mode: +v bobmcw[07:46:51] *** bobmcw has quit IRC[08:10:33] *** jfclere has joined #jboss-as7[08:10:33] *** ChanServ sets mode: +v jfclere[08:14:27] *** opalka has joined #jboss-as7[08:14:27] *** ChanServ sets mode: +v opalka[08:15:41] *** magesh has joined #jboss-as7[08:39:29] *** rawbdor has joined #jboss-as7[08:48:20] *** pilhuhn has joined #jboss-as7[08:48:20] *** pilhuhn has joined #jboss-as7[08:48:20] *** ChanServ sets mode: +v pilhuhn[08:48:21] *** misty has quit IRC[09:04:06] *** emuckenhuber has joined #jboss-as7[09:04:06] *** emuckenhuber has joined #jboss-as7[09:04:06] *** ChanServ sets mode: +v emuckenhuber[09:08:38] *** rawblem has joined #jboss-as7[09:17:17] *** opalka has quit IRC[09:17:20] *** rawbdor has quit IRC[09:18:01] *** rawblem has quit IRC[09:19:35] *** opalka has joined #jboss-as7[09:19:35] *** ChanServ sets mode: +v opalka[09:23:22] *** adietisheim has joined #jboss-as7[09:23:23] *** ChanServ sets mode: +v adietisheim[09:33:29] *** miclorb has quit IRC[09:47:52] *** AndyTaylor has joined #jboss-as7[09:47:53] *** AndyTaylor has joined #jboss-as7[09:47:53] *** ChanServ sets mode: +v AndyTaylor[09:49:54] *** AndyTaylor has quit IRC[09:50:00] *** AndyTaylor has joined #jboss-as7[09:50:00] *** ChanServ sets mode: +v AndyTaylor[10:08:22] *** miclorb_ has joined #jboss-as7[10:11:34] *** cvasilak has joined #jboss-as7[10:13:57] *** maeste has joined #jboss-as7[10:13:57] *** ChanServ sets mode: +v maeste[10:15:27] *** jcosta has joined #jboss-as7[10:15:27] *** ChanServ sets mode: +v jcosta[10:30:43] *** emuckenhuber has quit IRC[10:33:50] *** kkhan has joined #jboss-as7[10:33:50] *** ChanServ sets mode: +v kkhan[10:41:31] *** emuckenhuber has joined #jboss-as7[10:41:31] *** ChanServ sets mode: +v emuckenhuber[10:44:57] *** rawbdor has joined #jboss-as7[11:00:24] *** hbraun has joined #jboss-as7[11:01:38] *** hbraun has quit IRC[11:11:27] <pilhuhn> JimMa : about as7-697 and as7-698 : this feels wrong. I think the endpoints should directly show in the tree as children of the WS-subsystem. Like http-connectors show as children of the web subsystem.[11:11:47] <pilhuhn> Requiring an operation to list them is not orthogonal to how other subsystems work[11:12:37] <pilhuhn> Also the metrics should be available on a :read-resource(runtime=true) and not by some special operation for the webservices subsystem[11:22:57] *** miclorb_ has quit IRC[11:27:38] *** zroubali has joined #jboss-as7[11:34:52] <opalka> pilhuhn, could U elaborate more on :read-resource(runtime=true)[11:35:06] <opalka> pilhuhn, is there some sample code how to do it right ?[11:35:16] *** darranl has joined #jboss-as7[11:35:16] *** darranl has joined #jboss-as7[11:35:16] *** ChanServ sets mode: +v darranl[11:35:47] <pilhuhn> opalka just digging - web subsystem[11:36:13] <opalka> pilhuhn, anyway thanks for your feedback - highly appreciated![11:36:49] <pilhuhn> have a look at [domain@localhost:9999 connector=http] :read-resource-description[11:37:14] <opalka> pilhuhn, thx for info[11:37:24] <pilhuhn> here the connectors are individual children of subsystem=web[11:37:40] <pilhuhn> and then for each connector they include the metrics[11:37:57] <pilhuhn> like e.g.[11:37:58] <pilhuhn> },[11:37:58] <pilhuhn> "bytesSent" => {[11:37:58] <pilhuhn> "type" => INT,[11:37:58] <pilhuhn> "access-type" => "metric",[11:37:58] <pilhuhn> "storage" => "runtime"[11:37:58] <pilhuhn> },[11:38:36] <pilhuhn> and then you can do [domain@localhost:9999 connector=http] :read-attribute(name=bytesSent)[11:38:36] <pilhuhn> {[11:38:36] <pilhuhn> "outcome" => "success",[11:38:36] <pilhuhn> "result" => "no metrics available",[11:38:36] <pilhuhn> "compensating-operation" => undefined[11:38:37] <pilhuhn> }[11:39:28] <pilhuhn> read-attribute is an operation available on all subsystems, so users know about it.[11:42:12] <opalka> pilhuhn, ok[11:47:42] <pilhuhn> opalka the other thing (but you probably have not yet finished that ) -- currently there is no way of modifying the configuration[11:48:06] <pilhuhn> :add says (add a new webservices subsystem), which does not sound what I want[11:48:42] <pilhuhn> and the config attributes have (as not otherwise specificed , and if I understood emuckenhuber correctly) read-only access[11:49:28] <opalka> pilhuhn, you're saying we shouldn't provide add operation? Just remove?[11:49:49] <pilhuhn> also I wonder why the attributes are below a special configuration element[11:50:07] <pilhuhn>[11:50:07] <pilhuhn> "attributes" => {"configuration" => {[11:50:08] <pilhuhn> "description" => "Holder attribute for the actual configuration values",[11:50:08] <pilhuhn> "type" => OBJECT,[11:50:08] <pilhuhn> "required" => true,[11:50:08] <pilhuhn> "value-type" => {[11:50:08] <pilhuhn> "modify-soap-address" => {[11:51:35] <opalka> pilhuhn, they shouldn't be wrapped, right?[11:52:06] <pilhuhn> don't know about :add and :remove -- if that would allow to add endpoints, I'd say ok. But for the subsystem itself it sounds odd. But then this may be correct - perhaps ask emuckenhuber or brian[11:52:10] <pilhuhn> yes, no wrapping[11:52:28] <pilhuhn> they should directly go below attributes[11:53:10] <pilhuhn> have a look at [domain@localhost:9999 connector=http] :read-resource-description[11:53:44] <opalka> pilhuhn, ok[11:53:58] <opalka> pilhuhn, currently there's no such functionality available - to add endpoints[11:54:08] *** zroubali has quit IRC[11:54:11] <opalka> pilhuhn, both add/remove endpoints are not supported[11:54:26] <pilhuhn> I don't request that - I do not have enough knowledge here[11:54:28] <pilhuhn> :[11:54:30] <pilhuhn> :)[11:54:44] <opalka> pilhuhn, NP ;) Just informing you[11:57:03] <opalka> pilhuhn, OK, I see now what U mean[11:57:14] <opalka> pilhuhn, I'll fix all these issues today[11:57:23] <pilhuhn> cool, great[11:57:35] <opalka> pilhuhn, once again thanks for your feedback[11:57:47] <pilhuhn> You're very welcome[11:58:03] <opalka> pilhuhn, appreciated![12:01:59] *** asoldano has joined #jboss-as7[12:01:59] *** ChanServ sets mode: +v asoldano[12:05:19] *** jwulf has quit IRC[12:08:37] <jbossbot> git [jboss-as] push master d2a0573.. Alexey Loubyansky AS7-709 updated create/delete JMS resources commands to work in domain[12:08:38] <jbossbot> jira [AS7-709] update commands creating JMS resources to work in domain [Open (Unresolved) Task, Major, Alexey Loubyansky] https://issues.jboss.org/browse/AS7-709[12:08:39] <jbossbot> git [jboss-as] push master 69ce113.. Stefano Maestri AS7-677 Add a <cached-connection-manager debug="false" error="false"/> configuration element to the <connector> subsystem.[12:08:39] <jbossbot> jira [AS7-677] CCM configuration [Open (Unresolved) Task, Major, Stefano Maestri] https://issues.jboss.org/browse/AS7-677[12:08:39] <jbossbot> git [jboss-as] push master b0227f7.. Paul Ferraro AS7-694 Upgrade Infinispan to 5.0.0.CR1[12:08:40] <jbossbot> jira [AS7-694] Upgrade Infinispan to 5.0.0.CR1 [Coding In Progress (Unresolved) Component Upgrade, Major, Paul Ferraro] https://issues.jboss.org/browse/AS7-694[12:08:40] <jbossbot> git [jboss-as] push master bc5a6cd.. kabir Connector subsystem add operation should use 'request-properties', not 'attributes'[12:08:40] <jbossbot> git [jboss-as] push master URL: http://github.com/jbossas/jboss-as/compare/2440284...bc5a6cd[12:10:12] <JimMa> pilhuhn, Regarding the endpoints list operation , the problem is that the ws endpoint is created by deploying an webservice war or ejb jar, there is no approach to create by add a children under webservice subsystem like the http connector does.[12:11:10] <pilhuhn> jimma I understand[12:12:06] <pilhuhn> JimMa the problem with the special operation - epecially wrt metrics is that this needs special code in each and every client[12:12:21] <pilhuhn> that talks to the management api[12:12:46] <jbossbot> git [jboss-as] push master ae063b8.. Emanuel Muckenhuber fix CF describe handler[12:12:46] <jbossbot> git [jboss-as] push master URL: http://github.com/jbossas/jboss-as/compare/bc5a6cd...ae063b8[12:14:10] *** magesh has quit IRC[12:14:50] <JimMa> pilhuhn, OK. I'll look at if we can add some hook during in the deployment process to add some children nodes , or other better solution ..[12:16:31] *** magesh has joined #jboss-as7[12:18:12] <pilhuhn> maeste subsystem=connector is jca ?[12:18:32] <maeste> pilhuhn: general JCA config[12:18:46] <maeste> pilhuhn: we have 3 subsystem[12:18:57] <jbossbot> git [jboss-as] push master 7432351.. kabir Fix up create-jms-cf description[12:18:57] <jbossbot> git [jboss-as] push master URL: http://github.com/jbossas/jboss-as/compare/ae063b8...7432351[12:19:10] <maeste> pilhuhn: connector = general config for all deploymnets and other 2 subsystem[12:19:38] <maeste> pilhuhn: for example thread usage, validation activation, cached connection manager use and logging infos[12:19:49] <maeste> pilhuhn: then a datasources subsystem[12:20:48] <maeste> pilhuhn: and a resource-adapters subsystem for vendor (jboss) specific parameter of ra (not in JCA standard) like security and advanced transaction support or admin objects supports[12:21:06] <pilhuhn> k[12:21:40] <pilhuhn> so a configuration of a RA is partially in resource-adapters and partially in connectors. Equivalent for a datasouce[12:22:19] <maeste> pilhuhn: well what you find in connector are more config of the engine[12:22:30] <maeste> pilhuhn: not of a single ra[12:22:34] <maeste> pilhuhn: or ds[12:22:50] *** jingli has quit IRC[12:22:59] <maeste> pilhuhn: any configs in connector affect all the ra and ds in a domain[12:23:15] <maeste> pilhuhn: because it's on the JCA engine itself[12:23:16] <pilhuhn> ok - let me ask the other way around: will an end user when confiuring a DS or RA need to tweak the connector[12:23:30] <maeste> pilhuhn: not at all[12:23:44] <pilhuhn> ok, great[12:24:09] <maeste> pilhuhn: tweaking the connector could be done for general performance tuning for example[12:24:20] <maeste> pilhuhn: giving JCA a more aggressive thread usage[12:24:31] <maeste> pilhuhn: but it's not matter of single ds[12:24:49] <pilhuhn> k[12:25:29] <pilhuhn> I think we will want to hide that somehow in the default RHQ ui - like the underyling HornetQ subsystem for JMS[12:25:38] <pilhuhn> Thanks maeste[12:25:47] <maeste> pilhuhn: yup maybe a good idea[12:26:36] *** JimMa has quit IRC[12:28:51] *** rawbdor has quit IRC[12:42:11] <pilhuhn> Tests in error:[12:42:11] <pilhuhn> getDefaultCache(org.jboss.as.clustering.infinispan.DefaultEmbeddedCacheManagerTest)[12:42:11] <pilhuhn> getCache(org.jboss.as.clustering.infinispan.DefaultEmbeddedCacheManagerTest)[12:42:14] <pilhuhn> lunch time[12:42:19] *** pilhuhn is now known as pil-lunch-bbl[13:00:09] *** sgilda has joined #jboss-as7[13:10:36] *** magesh has quit IRC[13:19:37] *** asoldano is now known as asoldano_lunch[13:27:23] *** ccrouch has joined #jboss-as7[13:27:23] *** ChanServ sets mode: +v ccrouch[13:35:26] *** Jaikiran has joined #jboss-as7[13:35:26] *** ChanServ sets mode: +v Jaikiran[13:39:43] <kkhan> pil-lunch-bbl: Those tests passed for me, trying to run them again[13:40:45] <kkhan> They pass here[13:41:31] *** JimMa has joined #jboss-as7[13:45:22] *** pil-lunch-bbl is now known as pilhuhn[13:45:52] <Jaikiran> emuckenhuber: you around?[13:48:50] *** liweinan has joined #jboss-as7[13:50:19] <pilhuhn> kkhan Indeed now after lunch they all pass. They were hungry too :)[13:50:33] <kkhan> cool[13:50:49] * pilhuhn is embarrassed about false alarm[13:51:18] *** kkhan is now known as kabir_afk[13:53:41] <stuartdouglas> Jaikiran: got a minute?[13:53:47] <Jaikiran> stuartdouglas: sure[13:54:02] <emuckenhuber> Jaikiran: yes[13:54:19] <stuartdouglas> I have been working on dmlloyd's ejb branch, and have it nearly working[13:54:22] <Jaikiran> emuckenhuber: i was looking at alexey's testcase failures and he mentioned you too are running into the same?[13:54:28] <Jaikiran> stuartdouglas: the web part?[13:54:34] <stuartdouglas> the main thing that is left to do is setting up the ejb interceptor chains[13:54:42] <Jaikiran> i was working on it too, but i needed some help for weld/web to get it working[13:54:54] *** frainone has joined #jboss-as7[13:54:55] *** ChanServ sets mode: +v frainone[13:55:02] <Jaikiran> some of the ejb interceptor chains are done[13:55:06] <stuartdouglas> not so much the web part, just getting stuff to run[13:55:22] <Jaikiran> i have it failing in the smoke tests right now[13:55:40] <Jaikiran> so just trying to get past that issue before taking it forward w.r.t ejbs[13:55:48] <Jaikiran> do you have your changes somewhere?[13:55:50] <emuckenhuber> Jaikiran: yeah, sometimes... mostly when doing a clean build[13:55:58] <Jaikiran> emuckenhuber: you use OpenJDK?[13:56:19] <emuckenhuber> no, the sun jdk[13:56:24] <Jaikiran> ah ok[13:56:44] <Jaikiran> i am using sun jdk 1.6.0 24 update and alexey is using OpenJDK. so just checking if it's an JDk issue[13:56:57] <Jaikiran> i haven't been able to reproduce it yet with sun jdk[13:57:03] <Jaikiran> let me keep trying[13:57:21] <Jaikiran> the next time you run into this issue, please send me the -output.txt and the surefire reports[13:58:53] <stuartdouglas> Jaikiran: https://github.com/stuartwdouglas/jboss-as/tree/ejb most of the test failures are to do with sfsb's[13:59:23] <stuartdouglas> If you get a chance to look at it it would be appreciated[13:59:48] <stuartdouglas> it has also been rebased against master, so is not really compatible with david version any more[14:00:02] <Jaikiran> stuartdouglas: sure. infact, that's on top priority right now. so i'll start looking right away[14:00:16] <Jaikiran> btw, is your branch updated with the latest from dmlloyd's ejb branch?[14:00:33] <Jaikiran> ah, ignore that. i saw your previous message now[14:00:35] <stuartdouglas> yes, and then rebased[14:00:40] <Jaikiran> ok[14:00:50] <pilhuhn> maeste another question - this time on data sources[14:01:16] <stuartdouglas> David was going to reset his branch against mine, but I don't think he did before he went to bed[14:01:23] <maeste> pilhuhn: shot[14:01:45] <pilhuhn> maeste there are config elements like "blocking-timeout-wait-millis" that are r/w , while others are r/o - did you just not get to modify the others to r/w or are there other reasons for them being r/o?[14:02:09] <Jaikiran> right. i was going to discuss with david today about getting the latest from upstream so that we don't lag behind[14:02:23] <Jaikiran> i'll start working off your branch locally[14:02:45] <maeste> pilhuhn: no, it's modifiabl only supported ones, IOW ones that change pool config atm[14:02:47] <stuartdouglas> cool[14:03:14] <pilhuhn> maeste k - will that change in the future?[14:03:29] <maeste> pilhuhn: well depends[14:03:41] <maeste> pilhuhn: we have discusse it a bit[14:03:49] <maeste> pilhuhn: we = me and jesper[14:04:32] <maeste> pilhuhn: we don't full agree since he think is better to support only ones supported by jca subsystem[14:04:42] *** Jaikiran has quit IRC[14:05:00] *** stliu has quit IRC[14:05:21] <maeste> pilhuhn: because all the other require a server restart to take effect[14:05:44] <maeste> pilhuhn: if you have a different use case, please share on ML[14:05:47] <pilhuhn> Well, I see it from the usability side. Forcing the user to remove a DS and then adding it with 30 options is not user firendly[14:05:54] <pilhuhn> I'll dig it out[14:06:28] <maeste> pilhuhn: oki[14:07:34] <maeste> pilhuhn: it would be good to open a discussion in general on this point[14:07:41] <maeste> pilhuhn: not only for ds I mean[14:08:15] <maeste> pilhuhn: since atm AS support only r/w attribute that directly change the runtime status or r/w that require server restart[14:08:24] <maeste> pilhuhn: but for ds for example[14:08:31] <pilhuhn> yes - I get that feeling more and more when looking at the various pieces of configuration[14:08:37] <maeste> pilhuhn: would be possible to restart the ds itself[14:08:42] <pilhuhn> yep[14:08:59] <maeste> pilhuhn: that would be better from a user point of view of course[14:09:05] *** Jaikiran has joined #jboss-as7[14:09:05] *** ChanServ sets mode: +v Jaikiran[14:09:24] <maeste> pilhuhn: something like a deferred remove and add operation with new value[14:09:43] <maeste> pilhuhn: deferred = nice shutdown of the resource[14:10:02] <pilhuhn> I'll write an email - and you can chime in then :)[14:10:08] *** miclorb_ has joined #jboss-as7[14:10:13] <maeste> pilhuhn: yup[14:11:03] *** frainone has quit IRC[14:12:41] *** AndyTaylor has quit IRC[14:17:44] <Jaikiran> stuartdouglas: just to be sure these are the failures you are running into right? http://pastebin.com/GVjh1jk0[14:18:33] *** jamezp has quit IRC[14:18:34] <stuartdouglas> Jaikiran: yes, and some in the integration tests module to[14:18:45] <Jaikiran> ok. i'll start working on it[14:18:51] <stuartdouglas> my internet is so slow tonight[14:19:16] *** jamezp has joined #jboss-as7[14:21:26] *** asoldano_lunch is now known as asoldano[14:26:59] *** kabir_afk is now known as kabir[14:28:14] *** jwulf has joined #jboss-as7[14:29:59] *** smarlow has joined #jboss-as7[14:29:59] *** ChanServ sets mode: +v smarlow[14:31:04] *** ccrouch has quit IRC[14:40:24] *** miclorb_ has quit IRC[14:47:23] *** smarlow has quit IRC[14:52:22] *** mmoyses has joined #jboss-as7[14:52:22] *** ChanServ sets mode: +v mmoyses[14:55:00] *** JimMa has quit IRC[14:55:38] *** JimMa has joined #jboss-as7[15:00:24] *** jdcasey has joined #jboss-as7[15:01:30] *** AndyTaylor has joined #jboss-as7[15:01:30] *** ChanServ sets mode: +v AndyTaylor[15:03:16] *** clebert has joined #jboss-as7[15:03:16] *** ChanServ sets mode: +v clebert[15:13:50] *** dimitris_ has joined #jboss-as7[15:13:50] *** dimitris_ has joined #jboss-as7[15:13:50] *** ChanServ sets mode: +v dimitris_[15:14:51] <JimMa> pilhuhn, ping[15:15:08] <pilhuhn> JimMa pong[15:15:34] <JimMa> pilhuhn, After finished the work at hand , I looked at this again. There is no such api or extension point to add children node to the subsystem and also synchronise with war ejb deployments. Currently this could be the simple way to get the deployed endpoints .But I agree your suggestion , and that will be the ideal way to manage the deployed endpoint[15:16:24] *** bstansberry has joined #jboss-as7[15:16:24] *** ChanServ sets mode: +v bstansberry[15:16:29] <pilhuhn> jimma I can unfortunately not help you with the Jbas internals - perhaps emuckenhuber knows or bstansberry :)[15:16:35] <JimMa> pilhuhn, Regarding the WSEndpointMetrics, the client side does not need some special code(our ws code), the return objects are all simple type. But as you suggested, if it can be an attribute in Endpoint. That would be ideal.[15:16:58] <JimMa> pilhuhn, I will talk with Brain today to see if we can add some api or useful things to handle this kind of deployments to create children elements.[15:16:58] *** bstansberry has quit IRC[15:17:25] <emuckenhuber> maybe that's related to AS7-367 ?[15:17:26] <jbossbot> jira [AS7-367] Expose deployment details via the domain management API [Open (Unresolved) Feature Request, Critical, Unassigned] https://issues.jboss.org/browse/AS7-367[15:18:07] <pilhuhn> JimMa I think it is important (for RHQ and console) for users to have the same semantic all over the place. Users exepct to get at a metric via :read-attribute etc.[15:18:44] <JimMa> emuckenhuber, thanks. That's part of what I need.[15:19:16] <JimMa> pilhuhn, Understand that. Then we can use one plugin to get all the infos.[15:20:14] <jbossbot> git [jboss-as] push master 1a9cc0e.. Emanuel Muckenhuber handle coordinator-env default-timeout[15:20:15] <jbossbot> git [jboss-as] push master URL: http://github.com/jbossas/jboss-as/compare/7432351...1a9cc0e[15:20:30] <pilhuhn> yep. and also admins may use curl or wget to obtain metrics and deployed endpoints etc. in a common way[15:21:41] <mmoyses> kabir: if you have some time can you please pull the request i sent now?[15:23:15] <kabir> mmoyses: I'll try to take a look soon[15:23:30] <mmoyses> thanks[15:26:10] *** rawbdor has joined #jboss-as7[15:26:11] *** jpederse has joined #jboss-as7[15:26:23] *** jdcasey has quit IRC[15:29:06] *** bgeorges has joined #jboss-as7[15:31:56] *** jcosta has left #jboss-as7[15:32:00] *** jcosta has joined #jboss-as7[15:32:00] *** ChanServ sets mode: +v jcosta[15:36:19] *** frainone has joined #jboss-as7[15:36:19] *** ChanServ sets mode: +v frainone[15:52:50] *** mgoldmann has joined #jboss-as7[15:52:57] *** mgoldmann has joined #jboss-as7[15:52:57] *** ChanServ sets mode: +v mgoldmann[15:56:16] *** baileyje has joined #jboss-as7[15:56:16] *** ChanServ sets mode: +v baileyje[15:57:31] *** alesj has joined #jboss-as7[16:03:45] *** baileyje has quit IRC[16:09:35] *** baileyje has joined #jboss-as7[16:09:35] *** ChanServ sets mode: +v baileyje[16:12:16] *** jamezp is now known as jamezp_afk[16:12:58] *** hardy has joined #jboss-as7[16:13:55] *** bobmcw has joined #jboss-as7[16:13:55] *** ChanServ sets mode: +v bobmcw[16:14:39] *** pgier has joined #jboss-as7[16:14:39] *** ChanServ sets mode: +v pgier[16:21:33] *** adietisheim has quit IRC[16:24:26] *** galderz has joined #jboss-as7[16:24:26] *** ChanServ sets mode: +v galderz[16:26:25] *** jamezp_afk is now known as jamezp[16:28:52] *** smarlow has joined #jboss-as7[16:28:52] *** ChanServ sets mode: +v smarlow[16:29:01] *** bobmcw has quit IRC[16:29:45] *** opalka has quit IRC[16:35:23] *** maxandersen has joined #jboss-as7[16:35:23] *** ChanServ sets mode: +v maxandersen[16:39:30] *** cvasilak has quit IRC[16:47:45] *** kcbabo has joined #jboss-as7[16:47:53] *** ChanServ sets mode: +v kcbabo[16:53:16] *** aloubyansky has joined #jboss-as7[16:57:32] *** lgao has quit IRC[16:59:50] *** baileyje has quit IRC[17:00:14] <Jaikiran> dmlloyd: are you around for a quick discussion, later today around 1.5 hours from now?[17:01:47] <dmlloyd> I think so[17:02:10] <Jaikiran> ok[17:02:30] <Jaikiran> basically, i want to find a way to associate some session info with a ComponentViewInstance[17:03:01] <Jaikiran> creating the info and setting it in the interceptor context of ComponentViewInstance's postconstruct interceptor won't work[17:03:23] <Jaikiran> since that interceptor context will not be available to the *invocation* interceptor chain[17:03:28] *** kcbabo has quit IRC[17:03:46] <Jaikiran> so i'm looking at a way of extending the ViewInstance and setting such info within there[17:03:49] <Jaikiran> or something along those lines[17:03:57] *** pferraro has joined #jboss-as7[17:03:57] *** ChanServ sets mode: +v pferraro[17:03:57] <dmlloyd> there's a way to do it[17:04:01] <dmlloyd> I can explain if you have a minute[17:04:04] <Jaikiran> sure[17:04:08] <dmlloyd> first I have to refresh my memory :)[17:04:11] <Jaikiran> :)[17:04:19] <Jaikiran> btw, i am working off stuart's branch currently[17:04:37] *** frainone has quit IRC[17:04:38] <dmlloyd> yeah that's fine, I haven't had a chance to reset yet[17:06:28] <Jaikiran> btw on a slightly different topic, wouldn't the logic of storing j.l.r.Method(s) of a view as entry points and associating interceptors with each of those methods be a memory hog?[17:06:56] <dmlloyd> it shouldn't be if we make sure to cache instances[17:07:42] <Jaikiran> component view instance, you mean?[17:07:51] <dmlloyd> no, interceptor instances[17:08:05] <dmlloyd> so if an interceptor applies to 10 methods there's only one instance[17:08:25] <Jaikiran> yeah, but the memory store will increase with increase in methods exposed by the view[17:08:40] <Jaikiran> i was thinking of a notion of having interceptors associated to "all methods" of a view[17:08:44] <Jaikiran> and not store the methods at all[17:08:46] *** smcgowan has joined #jboss-as7[17:08:46] *** ChanServ sets mode: +v smcgowan[17:08:53] <Jaikiran> not sure how that works out, but just a thought[17:09:13] *** baileyje has joined #jboss-as7[17:09:14] *** ChanServ sets mode: +v baileyje[17:09:19] <dmlloyd> we have to associate the source method with the target at some point[17:09:27] * Jaikiran is rechecking the code to make sure he's read it right[17:09:28] <dmlloyd> but I'm open to alternatives[17:09:57] <dmlloyd> using identity lookups is better than using a plain map becuase Method has a poorly-designed equals/hashCode[17:10:14] *** mmoyses has quit IRC[17:10:39] <Jaikiran> yeah.[17:10:47] *** kcbabo has joined #jboss-as7[17:10:47] *** ChanServ sets mode: +v kcbabo[17:11:01] <Jaikiran> i'll take a look at this later and see if it really is going to impact the memory perf[17:11:13] <dmlloyd> okay so back to your first question, when are you creating the session ID? and can a view ever change session IDs? (I'm sure I've asked this before but I forget)[17:12:00] *** kcbabo has quit IRC[17:12:04] <Jaikiran> the session id is created on a lookup of a stateful component proxy via jndi[17:12:10] <Jaikiran> SFSB[17:12:35] <Jaikiran> so a user does, a lookup of java:global/sfsb and that in turn results in a session being started[17:13:01] <Jaikiran> we do so by creating a id for that session and keep it associated with that proxy through the lifetime of that sfsb instance[17:13:06] <dmlloyd> okay so the construct interceptor should create it, and an invocation interceptor should associate it, is that a correct statement? in broad terms?[17:13:13] <dmlloyd> by "it" I mean the session ID[17:13:19] <Jaikiran> yes[17:13:24] <Jaikiran> to be precise[17:13:29] <dmlloyd> oh, and the destroy interceptor should remove it[17:13:31] <Jaikiran> post construct interceptor the view[17:13:33] <Jaikiran> right[17:13:49] <Jaikiran> creating it in the post construct part is easy[17:13:50] <dmlloyd> okay so you'll want to take a look at how I do injection because this is very similar[17:14:09] <Jaikiran> it's how we pass on that info to the invocation interceptor chain is what i am trying ot find out[17:14:14] <dmlloyd> what I do is create an AtomicReference to hold the target data, within the interceptor factory, which is associated under a unique key[17:14:39] <dmlloyd> then, the construct/invoked/destroy interceptors are constructed with the same AtomicReference (which it gets from the interceptor factory context)[17:15:05] <dmlloyd> so in your case it'd just be an AtomicReference<SessionIDType>[17:15:19] <Jaikiran> i see[17:15:24] <Jaikiran> makes sense[17:15:38] <dmlloyd> it's not obvious I know, I was hoping to document it a little better[17:15:46] <Jaikiran> np, for now :)[17:16:46] * Jaikiran will bbl in a hour and implement this[17:16:51] *** Jaikiran is now known as Jaikiran|AFK[17:18:12] <dmlloyd> cool[17:29:10] *** JimMa has quit IRC[17:33:37] *** alexsmirnov has joined #jboss-as7[17:35:11] *** maxandersen has quit IRC[17:36:38] *** maxandersen has joined #jboss-as7[17:36:38] *** ChanServ sets mode: +v maxandersen[17:43:07] *** maxandersen has quit IRC[17:45:17] *** maeste has quit IRC[17:46:27] *** fnasser has joined #jboss-as7[17:47:59] *** bgeorges has quit IRC[17:48:32] *** jpederse has quit IRC[17:49:15] *** smarlow has quit IRC[17:50:17] *** smcgowan has quit IRC[17:50:25] *** dimitris_ has quit IRC[17:53:18] *** pilhuhn has quit IRC[17:53:29] <baileyje> dmlloyd: Can you take a look at https://github.com/baileyje/jboss-as/commits/JBAS-9221[17:53:30] <jbossbot> jira [JBAS-9221] Redirected to: https://issues.jboss.org/si/jira.issueviews:issue-xml/AS7-254/AS7-254.xml[17:53:31] <jbossbot> jira [AS7-254] Undeclared checked exception is thrown from EJB invocation [Open (Unresolved) Bug, Critical, Unassigned] https://issues.jboss.org/browse/AS7-254[17:55:14] <dmlloyd> hm, okay, the solution (in the ejb branch of course) is to insert a first client interceptor which checks any thrown checked Throwables against the declared checked exception types, and wraps if necessary[17:55:29] <dmlloyd> we might already have a first client interceptor, in which case we can just tack this on there[17:55:31] *** jcosta has quit IRC[17:56:18] <baileyje> dmlloyd: Sure. I can put it in there. Or would it make sense to add another interceptor, or to update that existing first interceptor?[17:56:54] <baileyje> dmlloyd: I gotta run. I will update it after I grab some lunch[17:56:56] <dmlloyd> if there is a first interceptor, I'd just modify that. If not then add one.[17:57:01] <dmlloyd> maybe in jboss-invocation even.[17:57:14] <baileyje> ok[17:57:19] <baileyje> back in a bit[17:57:21] *** baileyje has quit IRC[18:05:09] *** aslak has joined #jboss-as7[18:05:09] *** ChanServ sets mode: +v aslak[18:11:36] *** mgoldmann has quit IRC[18:11:54] *** mgoldmann has joined #jboss-as7[18:11:54] *** ChanServ sets mode: +v mgoldmann[18:15:11] *** alesj has quit IRC[18:18:08] *** mgoldmann has quit IRC[18:25:33] *** sguilhen has quit IRC[18:39:43] *** ALR has joined #jboss-as7[18:39:43] *** ChanServ sets mode: +v ALR[18:51:34] *** asoldano has quit IRC[18:55:12] *** AndyTaylor has left #jboss-as7[18:55:49] *** emuckenhuber has quit IRC[18:59:31] *** jfclere has quit IRC[19:16:21] *** frainone has joined #jboss-as7[19:16:21] *** ChanServ sets mode: +v frainone[19:18:23] *** mmoyses has joined #jboss-as7[19:18:23] *** ChanServ sets mode: +v mmoyses[19:18:39] *** sguilhen has joined #jboss-as7[19:21:36] *** pilhuhn has joined #jboss-as7[19:21:36] *** pilhuhn has joined #jboss-as7[19:21:36] *** ChanServ sets mode: +v pilhuhn[19:22:06] *** darranl has quit IRC[19:26:33] *** adietisheim has joined #jboss-as7[19:27:18] *** Jaikiran|AFK has quit IRC[19:28:23] *** adietisheim1 has joined #jboss-as7[19:30:25] *** kabir has quit IRC[19:31:11] *** adietisheim has quit IRC[19:34:04] *** hardy has quit IRC[19:34:04] *** jamezp has quit IRC[19:34:34] *** jamezp has joined #jboss-as7[19:43:09] *** pferraro has quit IRC[19:46:24] *** pferraro has joined #jboss-as7[19:46:25] *** ChanServ sets mode: +v pferraro[19:55:46] *** mgoldmann has joined #jboss-as7[19:55:46] *** ChanServ sets mode: +v mgoldmann[20:06:05] *** ALR has quit IRC[20:09:06] *** kkhan has joined #jboss-as7[20:09:06] *** ChanServ sets mode: +v kkhan[20:17:24] *** mgoldmann has quit IRC[20:23:50] *** ALR has joined #jboss-as7[20:23:50] *** ChanServ sets mode: +v ALR[20:36:27] *** galderz has quit IRC[20:41:42] *** galderz has joined #jboss-as7[20:41:43] *** ChanServ sets mode: +v galderz[20:48:31] *** galderz has quit IRC[20:57:01] *** maxandersen has joined #jboss-as7[20:57:02] *** ChanServ sets mode: +v maxandersen[21:05:54] *** maxandersen has quit IRC[21:05:58] *** jamezp is now known as jamezp_afk[21:22:47] *** smarlow has joined #jboss-as7[21:22:50] *** ChanServ sets mode: +v smarlow[21:30:17] *** smarlow has quit IRC[21:30:45] *** emmanuel has joined #jboss-as7[21:30:45] *** ChanServ sets mode: +v emmanuel[21:31:05] *** smarlow has joined #jboss-as7[21:31:06] *** ChanServ sets mode: +v smarlow[21:35:57] *** bgeorges has joined #jboss-as7[21:40:26] *** stansilvert has quit IRC[21:41:38] *** bgeorges has quit IRC[21:42:33] *** bgeorges has joined #jboss-as7[21:42:35] *** emmanuel has quit IRC[21:42:43] *** smcgowan has joined #jboss-as7[21:50:40] <jbossbot> git [jboss-as] push master 47b7a51.. Paul Ferraro Cave to the anti-JAXB lobby. Replace externalized infinispan defaults with hard coded defaults.[21:50:40] <jbossbot> git [jboss-as] push master URL: http://github.com/jbossas/jboss-as/compare/1a9cc0e...47b7a51[21:51:40] *** smarlow has quit IRC[21:54:30] *** bgeorges has quit IRC[21:56:03] *** baileyje has joined #jboss-as7[21:56:03] *** ChanServ sets mode: +v baileyje[21:57:58] *** kkhan has quit IRC[21:58:21] *** alesj has joined #jboss-as7[22:02:52] *** jamezp_afk is now known as jamezp[22:03:26] <dmlloyd> pferraro++[22:03:42] <dmlloyd> he can see which way the wind blows![22:14:09] *** irooskov has joined #jboss-as7[22:16:50] *** smcgowan has quit IRC[22:17:48] *** smcgowan has joined #jboss-as7[22:17:52] *** ChanServ sets mode: +v smcgowan[22:18:23] *** baileyje has quit IRC[22:19:16] *** baileyje has joined #jboss-as7[22:19:16] *** ChanServ sets mode: +v baileyje[22:30:02] *** pilhuhn has quit IRC[23:03:51] *** mmoyses has quit IRC[23:06:06] *** misty has joined #jboss-as7[23:20:15] *** jamezp has left #jboss-as7[23:20:50] *** emuckenhuber has joined #jboss-as7[23:20:50] *** ChanServ sets mode: +v emuckenhuber[23:21:30] *** jamezp has joined #jboss-as7[23:21:58] *** jamezp has left #jboss-as7[23:22:12] *** jamezp has joined #jboss-as7[23:22:24] *** jamezp has left #jboss-as7[23:22:32] <stuartdouglas> morning[23:22:53] *** jamezp has joined #jboss-as7[23:23:51] <alesj> stuartdouglas: hey[23:23:55] <alesj> how was vacation?[23:24:09] <stuartdouglas> good[23:24:25] <stuartdouglas> how was judcon?[23:24:31] <alesj> good as well[23:24:36] <alesj> a lot of people[23:24:49] <alesj> good hackfest and very cool demo[23:24:54] <alesj> at keynote[23:32:17] *** emuckenhuber has quit IRC[23:33:09] <baileyje> dmlloyd: https://github.com/baileyje/jboss-invocation/commit/78906a67d886c8943017afe387388e7be2f6e379[23:33:09] <jbossbot> git [jboss-invocation] 78906a6.. John E. Bailey Add an initial interceptor capable of handling undeclared exceptions[23:33:15] <baileyje> finally got some more time[23:33:43] <dmlloyd> e should be Throwable since Throwable is checked[23:34:03] <dmlloyd> otherwise looks good[23:34:20] <dmlloyd> which means you need a catch (Error e) as well[23:35:18] *** miclorb_ has joined #jboss-as7[23:37:05] *** dimitris_ has joined #jboss-as7[23:37:05] *** dimitris_ has joined #jboss-as7[23:37:05] *** ChanServ sets mode: +v dimitris_[23:37:34] *** baileyje has quit IRC[23:38:30] <misty> how's it going at JBW??[23:39:59] <dmlloyd> too late, he's gone :)[23:41:39] *** bgeorges has joined #jboss-as7[23:41:47] <misty> hi bgeorges[23:41:58] <misty> dmlloyd: we only have one attendee in the channel? :([23:42:13] <dmlloyd> well, they keep dropping off as they go to do really fun stuff[23:42:49] *** baileyje has joined #jboss-as7[23:42:50] *** baileyje has quit IRC[23:42:50] *** baileyje has joined #jboss-as7[23:42:50] *** ChanServ sets mode: +v baileyje[23:43:16] <misty> lol[23:44:10] *** jamezp has left #jboss-as7[23:44:52] *** fnasser has quit IRC[23:45:02] *** adietisheim1 has quit IRC[23:46:21] *** bgeorges has quit IRC[23:48:01] <stuartdouglas> dmlloyd: Do you know if jaikiran made any progress on the ejb branch? I can't see anything in his github[23:48:33] <dmlloyd> he said he was going to implement SFSB session association[23:48:41] <dmlloyd> we talked about it about 6 hours ago[23:48:42] *** jamezp has joined #jboss-as7[23:48:52] <dmlloyd> not sure if he got sidetracked[23:49:55] <stuartdouglas> ok, that and duplicate jndi bindings seem to be the only two major outstanding issues at the moment[23:50:40] <dmlloyd> okay, I'll see what I can dredge up on the JNDI thing[23:51:49] <alesj> dmlloyd: what's the reason not to include TxM in JNDI?[23:51:59] <dmlloyd> what's the reason _to_ include it?[23:52:21] <alesj> 1M frameworks expecting it there, including Seam[23:52:32] <dmlloyd> the txn guys think it's not a great idea, since there's really no defined contract[23:52:53] <alesj> what contract would they want/need?[23:53:01] <dmlloyd> not sure![23:53:20] <dmlloyd> if you can get jonathan on board then I'm okay with adding it[23:53:35] *** jamezp has quit IRC[23:53:45] *** jamezp has joined #jboss-as7[23:53:55] <alesj> hehe, jonathan said you said it should not get in[23:54:01] <alesj> and that he has no issue[23:54:01] <alesj> :-0[23:54:07] <dmlloyd> hmmm :)[23:54:17] <alesj> so, resolved … it's in[23:54:18] <alesj> :-)[23:54:44] <misty> hahaha[23:54:51] <dmlloyd> it's better to use it via service injection though[23:55:04] <alesj> definitely[23:55:09] <dmlloyd> anything which is non-spec that we put in JNDI has to be in java:jboss[23:55:14] <stuartdouglas> but then your portable frameworks needs deps on AS7 classes[23:55:15] <misty> hi alesj, btw. First time I have seen you on IRC and not just in email![23:55:32] <alesj> misty: hey, well, we're in diff timezones[23:55:32] <dmlloyd> sure, stuartdouglas, I'm thinking mainly of subsystems.[23:55:39] <alesj> misty: i just happen to be in US atm[23:55:45] <misty> right[23:56:44] <alesj> dmlloyd: and not all sub-systems want to depend on TM[23:57:06] <alesj> e.g. i opened a thread on CDI[23:57:36] <alesj> where it was the same argument; spec doesn't deal with TxM, hence no usage[23:57:53] <alesj> but, otoh, 1M of frameworks expects it there[23:58:07] <alesj> all having same lookup pattern then[23:58:27] <dmlloyd> Jonathan says that things should usually use UserTransaction iirc[23:58:30] <alesj> infinispan (now moving to injection), spring, seam, ...[23:58:33] <alesj> yes[23:58:37] <alesj> for the user code[23:58:38] <stuartdouglas> UserTransaction is limited[23:58:44] <alesj> but frameworks need a bit more[23:58:47] <stuartdouglas> e.g. you can't register synchronizations[23:58:58] <stuartdouglas> can't implement stuff like REQUIRES_NEW[23:58:58] <alesj> XA handling, synchs, suspend/resume, ...[23:59:53] <stuartdouglas> not having it in there is just going to piss people off in a big way