May 6, 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 6, 2011 NOTICE: This channel is no longer actively logged.
[00:00:14] <alesj> yup[00:00:29] <stuartdouglas> well, it will piss of people that write frameworks that deal with transactions anyway :-)[00:02:11] <alesj> any semi-complex framework eventually deals with tx :-)[00:03:13] <dmlloyd> do things usually look for it in a common location?[00:03:36] <alesj> hmm, mostly there is no common location[00:03:48] <alesj> so there is try/catch lookup[00:04:00] <alesj> for diff jndi names[00:04:04] <stuartdouglas> yep[00:04:06] <alesj> trying to find the right one[00:04:44] <stuartdouglas> and then you need special code for websphere as well, because it uses a UOWManager rather than a TransactionManager[00:04:45] <alesj> or, an exact TMLookup impl; e.g. JBossTMLookup[00:05:06] <stuartdouglas> It would be much better if the spec just said 'TM goes here'[00:05:12] <alesj> yup[00:05:45] <baileyje> dmlloyd: There?[00:06:27] <alesj> baileyje: did you manage to finish that gigantic whiskey glass? :-)[00:06:45] <baileyje> alesj: of course... :)[00:06:50] <baileyje> Not a super good idea[00:07:07] <stuartdouglas> I'm sure it seemed like one at the time[00:07:15] <misty> drunk coding is never as good an idea as it seems at the time[00:11:47] *** clebert has quit IRC[00:13:36] <dmlloyd> baileyje: yeah[00:13:53] <dmlloyd> baileyje: I put a line comment on your commit, in case you missed my irc remarks[00:14:21] <baileyje> So I think for this to work, we will need Interceptor to throw Throwable not Exception, since Throwable could be decalred on the target method.[00:14:34] <baileyje> dmlloyd: Yeah, I saw the commit comment, but not the IRC remarks[00:14:39] <dmlloyd> ah good point.[00:14:51] <dmlloyd> might have to always wrap Throwable[00:15:01] <baileyje> To make it an exception?[00:15:07] <dmlloyd> as Undeclared[00:15:17] <dmlloyd> just because we can't throw it doesn't mean that something we call can't[00:15:18] <baileyje> What if they did declare throwable?[00:15:26] <dmlloyd> then tough crap :)[00:15:40] <dmlloyd> unless we think of something better[00:15:47] <dmlloyd> I think that's good enough thouhg[00:15:55] <baileyje> I don't think there is really better.[00:16:02] * dmlloyd is a big fan of "good enough" these days[00:16:17] <baileyje> We just either allow the Interceptor interface to throw Throwable, or we have to wrap[00:16:34] <baileyje> Throwable matches the invocation handler interface as well[00:16:46] <dmlloyd> well there is a way[00:17:13] <dmlloyd> we can do: public class Dummy { public Dummy(Throwable t) throws Throwable { throw t; }}[00:17:19] <dmlloyd> then do Dummy.class.newInstance(t)[00:17:21] <dmlloyd> :-)[00:17:28] * dmlloyd barfs![00:17:39] <dmlloyd> it's the old newInstance hole[00:18:38] <dmlloyd> you can do that if you really really want to[00:18:42] <dmlloyd> but I don't care either way[00:19:08] <dmlloyd> if you do though, only do it for Throwables which aren't Errors or Exceptions[00:19:22] <dmlloyd> minimize the awfulness :)[00:19:33] <baileyje> Will that actually give the caller the Throwable?[00:19:40] *** alesj has quit IRC[00:19:44] <dmlloyd> yeah it'll throw it like normal, bypassing the declared exceptions[00:19:59] <dmlloyd> with the original stack trace and everything[00:21:16] <dmlloyd> that's why they always say use getConstructor().newInstance() instead[00:21:21] <dmlloyd> it wraps all thrown exceptions[00:21:32] <dmlloyd> newInstance() dates back to days before reflection[00:22:12] <dmlloyd> okay I gotta step out to get some dinner, anything else before I go?[00:23:10] <dmlloyd> going once[00:23:20] <dmlloyd> going twice[00:23:29] <dmlloyd> gone![00:23:30] <dmlloyd> bbiab[00:27:08] *** jpearlin has joined #jboss-as7[00:27:55] *** dimitris_ has quit IRC[00:33:06] *** jamezp has quit IRC[00:33:19] *** jamezp has joined #jboss-as7[00:44:05] *** bobmcw has joined #jboss-as7[00:44:05] *** ChanServ sets mode: +v bobmcw[00:45:03] <baileyje> dmlloyd: I pushed a change for you to review when you get back. I just left it throwing an Undeclared for all throwables. The Invoking interceptor won't allow a target method to throw a thowable anyway. It will turn it into an undeclared throwable. If some clown really has throwable as their declared type, we can deal with it then.[00:49:09] *** baileyje has quit IRC[00:52:25] <jbossbot> git [jboss-invocation] push master f903792.. John E. Bailey Add an initial interceptor capable of handling undeclared exceptions[00:52:25] <jbossbot> git [jboss-invocation] push master URL: http://github.com/jbossas/jboss-invocation/compare/4dce7ff...f903792[00:54:32] *** frainone has quit IRC[01:04:27] *** miclorb_ has quit IRC[01:07:39] *** baileyje has joined #jboss-as7[01:07:39] *** ChanServ sets mode: +v baileyje[01:11:49] *** bgeorges has joined #jboss-as7[01:18:12] *** miclorb_ has joined #jboss-as7[01:21:11] *** alexsmirnov has quit IRC[01:22:04] *** baileyje has quit IRC[01:23:21] *** bgeorges has quit IRC[01:25:18] *** mgoldmann has joined #jboss-as7[01:25:18] *** ChanServ sets mode: +v mgoldmann[01:26:01] *** bobmcw has quit IRC[01:51:15] *** jamezp is now known as jamezp_afk[02:16:39] *** pferraro has quit IRC[02:24:44] *** lgao has joined #jboss-as7[02:30:16] *** pferraro has joined #jboss-as7[02:30:16] *** ChanServ sets mode: +v pferraro[02:32:01] *** pferraro has left #jboss-as7[02:37:30] *** liweinan_ has joined #jboss-as7[02:37:35] *** liweinan has quit IRC[02:37:35] *** liweinan_ is now known as liweinan[02:56:19] *** jwulf_ has joined #jboss-as7[02:59:39] *** jwulf has quit IRC[03:13:15] *** jdcasey has joined #jboss-as7[03:14:09] *** smcgowan has quit IRC[03:18:53] *** jamezp_afk has quit IRC[03:18:54] *** pgier has quit IRC[03:19:35] *** pgier has joined #jboss-as7[03:19:35] *** ChanServ sets mode: +v pgier[03:19:38] *** rawbdor has quit IRC[03:21:03] *** jamezp has joined #jboss-as7[03:22:48] *** mgoldmann has quit IRC[03:35:38] *** jpearlin has left #jboss-as7[03:49:44] *** jwulf_ has quit IRC[03:50:41] *** jwulf_ has joined #jboss-as7[03:52:52] *** pgier has quit IRC[04:07:09] *** JimMa has joined #jboss-as7[04:08:16] *** stliu has joined #jboss-as7[04:08:17] *** ChanServ sets mode: +v stliu[04:14:24] *** jwulf_ has quit IRC[04:30:00] *** jdcasey has quit IRC[04:41:48] *** miclorb_ has quit IRC[04:42:08] *** miclorb__ has joined #jboss-as7[04:47:01] <dmlloyd> http://github.com/dmlloyd/jboss-as/commit/132d061[04:47:01] <jbossbot> git [jboss-as] 132d061.. David M. Lloyd Shared bindings part 1: allow comparison of binding injection source[04:59:50] <jamezp> Good stuff.[05:01:18] <stuartdouglas> I just noticed that in the EJB spec under the timers section there is this line: "By default, each Schedule annotation corresponds to a single persistent timer, regardless of the number of JVMs across which the container is distributed."[05:01:41] <stuartdouglas> does that mean that if you have 5 jvm's running the same app, the scheduled event should only be run on one of them>[05:01:43] <stuartdouglas> ?[05:03:54] <dmlloyd> yeah, though the spec is pretty loose about what defines a container[05:08:14] <stuartdouglas> That makes implementing timers considerable harder :-([05:09:45] <dmlloyd> indeed[05:10:01] <dmlloyd> when in doubt, create an SPI with a trivial implementation![05:11:01] <stuartdouglas> first server up runs all the task :-)[05:11:30] <stuartdouglas> I was thinking I may have a look at implementing this over the weekend without using quartz[05:12:30] <dmlloyd> yeah I think the multi-server behavior overlaps with clustering somewhat[05:12:54] <stuartdouglas> yea[05:13:24] <dmlloyd> I think infinispan has some job distribution thing though[05:13:31] <dmlloyd> so when the time comes I'm sure there will be a way[05:14:07] <stuartdouglas> ok, if I get to it I will try and abstract all that sort of stuff out as much as possible[05:15:20] <stuartdouglas> a really simple sort of way would be have each job as an infinispan node, first node to lock the job wins, however that is probably not hugely scalable / performant[05:15:39] *** jwulf has joined #jboss-as7[05:16:46] <dmlloyd> yeah maybe the infinispan guys could throw in some ideas[05:17:21] <dmlloyd> if only there was a venue where a bunch of folks from the clustering & infinispan team could meet with the EJB team and the core AS team...[05:17:23] <dmlloyd> oh wait[05:17:28] <dmlloyd> :)[05:17:46] <stuartdouglas> hehe[05:32:57] *** jamezp is now known as jamezp_afk[05:33:27] *** irooskov has quit IRC[05:38:45] *** bgeorges has joined #jboss-as7[05:46:42] *** magesh has joined #jboss-as7[05:47:01] *** bgeorges has quit IRC[06:06:29] *** jwulf has quit IRC[06:08:35] <dmlloyd> I'm on to the hard part about the naming stuff now. Gonna head to bed and work on it after rest.[06:09:04] <stuartdouglas> ok[06:16:48] *** liweinan has quit IRC[06:22:02] *** frainone has joined #jboss-as7[06:22:02] *** ChanServ sets mode: +v frainone[06:23:57] *** lgao has quit IRC[06:24:57] *** lgao has joined #jboss-as7[06:24:58] *** liweinan has joined #jboss-as7[06:34:09] *** magesh1 has joined #jboss-as7[06:34:34] *** magesh has quit IRC[07:20:39] *** magesh has joined #jboss-as7[07:21:28] *** magesh1 has quit IRC[07:27:41] *** frainone has quit IRC[07:28:21] *** bobmcw has joined #jboss-as7[07:28:21] *** ChanServ sets mode: +v bobmcw[07:58:47] *** bobmcw has quit IRC[08:04:47] *** rawbdor has joined #jboss-as7[08:13:19] *** jfclere has joined #jboss-as7[08:13:19] *** ChanServ sets mode: +v jfclere[08:17:31] *** aslak has quit IRC[08:21:52] *** magesh has left #jboss-as7[08:31:25] *** tdiesler has joined #jboss-as7[08:31:25] *** ChanServ sets mode: +v tdiesler[08:37:26] *** jwulf has joined #jboss-as7[08:47:11] *** opalka has joined #jboss-as7[08:47:11] *** ChanServ sets mode: +v opalka[08:47:27] <opalka> morning[08:47:54] *** misty has left #jboss-as7[08:50:22] *** aloubyansky has quit IRC[08:51:45] *** maeste has joined #jboss-as7[08:51:45] *** ChanServ sets mode: +v maeste[08:54:26] *** ALR has left #jboss-as7[09:02:41] *** rawblem has joined #jboss-as7[09:05:47] *** magesh has joined #jboss-as7[09:08:45] *** tdiesler has quit IRC[09:16:33] *** pilhuhn has joined #jboss-as7[09:16:33] *** ChanServ sets mode: +v pilhuhn[09:18:43] *** hbraun has joined #jboss-as7[09:23:54] *** miclorb__ has quit IRC[09:31:36] <hbraun> maeste: fyi https://issues.jboss.org/browse/AS7-719[09:31:38] <jbossbot> jira [AS7-719] En/Disable datasource returns error although state change is applied [Open (Unresolved) Bug, Major, Stefano Maestri] https://issues.jboss.org/browse/AS7-719[09:37:32] <maeste> hbraun: oki thx[09:37:57] <maeste> hbraun: it seems a domain related problem. I'll work on this[09:38:13] <hbraun> i just wanted to let yo know[09:38:17] <hbraun> just discoverd it[09:39:29] *** emuckenhuber has joined #jboss-as7[09:39:29] *** emuckenhuber has joined #jboss-as7[09:39:29] *** ChanServ sets mode: +v emuckenhuber[09:40:36] <hbraun> maeste: another question[09:40:51] <hbraun> the schema for xa DS contains a driver element[09:41:22] <hbraun> is that correct[09:41:24] <hbraun> ?[09:41:37] <hbraun> can it have driver and dadasource-class ?[09:42:09] <hbraun> maeste: datasource-class[09:42:33] <maeste> hbraun: yup it have[09:42:58] <maeste> hbraun: at least in some case[09:43:04] <maeste> hbraun: let me find a sample[09:43:11] *** pilhuhn has quit IRC[09:44:18] *** tdiesler has joined #jboss-as7[09:44:18] *** ChanServ sets mode: +v tdiesler[09:45:56] <maeste> hbraun: hmm maybe it's a refuse...let me check better IJ code...I'll let you know in few mins[09:46:02] *** emuckenhuber has quit IRC[09:46:35] <hbraun> maeste: no rush, just wondering ...[09:52:00] *** liweinan has quit IRC[09:58:03] *** liweinan has joined #jboss-as7[10:00:11] *** emuckenhuber has joined #jboss-as7[10:00:11] *** emuckenhuber has joined #jboss-as7[10:00:11] *** ChanServ sets mode: +v emuckenhuber[10:05:29] *** kkhan has joined #jboss-as7[10:05:29] *** ChanServ sets mode: +v kkhan[10:07:30] *** AndyTaylor has joined #jboss-as7[10:07:30] *** ChanServ sets mode: +v AndyTaylor[10:44:13] *** kkhan has quit IRC[10:47:58] *** kkhan has joined #jboss-as7[10:47:58] *** ChanServ sets mode: +v kkhan[10:52:06] *** jcosta has joined #jboss-as7[10:52:06] *** ChanServ sets mode: +v jcosta[10:52:21] *** adietisheim has joined #jboss-as7[10:52:21] *** ChanServ sets mode: +v adietisheim[10:58:03] <maeste> hbraun: well, I've verified[10:58:10] <hbraun> shoot[10:58:26] <maeste> hbraun: driver is needed to express the dependency to jdbc driver servic[10:58:45] <maeste> hbraun: in AS7 which is responsble to load jar in the correct cl[10:59:02] <hbraun> in that case it should be mandatory, no?[10:59:18] <maeste> hbraun: and the jar (containing the driver) generally is containing also the xa class[10:59:22] *** galderz has joined #jboss-as7[10:59:22] *** ChanServ sets mode: +v galderz[10:59:32] <maeste> hbraun: only in AS7, not mandatory in standalone JCA[10:59:49] <maeste> hbraun: and that's tricky[11:00:42] <hbraun> strange, because this one works: https://github.com/maeste/IronJacamar/blob/master/adapters/src/test/resources/h2-xa-ds.xml[11:00:42] *** AndyTaylor has quit IRC[11:00:57] <hbraun> at least it doesn't choke on the config[11:01:06] <hbraun> havn't tried using the actual DS[11:01:13] <hbraun> but it doesn't contain a driver[11:01:52] <maeste> hbraun: I can't remember for sure, but it should trow a class loading exception if you try to use it[11:02:15] <maeste> hbraun: even if you have probably also a standard DS on h2 in your config[11:02:21] <hbraun> maybe you need a separate schema for AS7?[11:02:37] <maeste> hbraun: that's force jdbc driver service start before[11:02:49] <hbraun> forget my last question[11:03:25] <hbraun> i don't want to blame anyone, but this whole DS config thing is really hard to understand[11:03:36] <hbraun> this will come back to us[11:03:54] <hbraun> it's too comlpicated, too many edge cases[11:04:00] <maeste> hbraun: the hard to understand part is on driver[11:04:09] <maeste> hbraun: xa part is not up to us[11:04:17] <maeste> hbraun: as said there is no standard there[11:04:28] <hbraun> well, the driver required in xa is up to us[11:04:49] <maeste> hbraun: yup as said driver part is the hard part[11:05:06] <hbraun> yes, the remaining bits are fine[11:05:24] <hbraun> i hink a driver registry would help a lot[11:05:24] <maeste> hbraun: maybe it would be better to change the name and behavior[11:05:50] <maeste> hbraun: not calling it Driver at all[11:05:53] <hbraun> so people can simply reference a driver instead of providing string attributes[11:06:02] <maeste> hbraun: yup[11:06:17] <maeste> hbraun: referring to jar name or something like that[11:06:26] <hbraun> i would even go that far, that you cannot create a DS w/o referencing a driver[11:06:27] *** pilhuhn has joined #jboss-as7[11:06:28] *** pilhuhn has joined #jboss-as7[11:06:28] *** ChanServ sets mode: +v pilhuhn[11:06:36] <hbraun> which would force people to deploy the driver first[11:07:01] <maeste> hbraun: yuo in fact it is[11:07:33] <hbraun> i know, but currently it resolves the service dependency (driver) on demand[11:08:03] <hbraun> so you can actually create a DS, get an error back, then deploy the driver, and the DS becomes available[11:08:27] <maeste> yup[11:08:48] <hbraun> now add a wrong driver declaration to the mix and you get pending DS configurations waiting in vain[11:09:02] <hbraun> fail fast would be better[11:09:12] <hbraun> i.e. in the above case, not create the DS at all[11:09:19] <hbraun> if the driver is not installed[11:10:08] <hbraun> the question is: what can we do between now and CR1?[11:10:39] <maeste> hbraun: first of all changing a bit the config would help a lot[11:11:41] <hbraun> changing into what direction?[11:12:09] <maeste> hbraun: and the real question about services is: is this fast fail right in the AS7 landscape? Having services installed and waiting for deps are normal beahviour in AS7[11:12:23] <maeste> hbraun: ops link not paste :) https://gist.github.com/958669[11:12:33] <hbraun> right, but you are in charge of the DMR operation[11:12:49] <hbraun> it can be dealt on that level, no?[11:13:46] <maeste> yup, but it's a problem of what user expect. If in all case services waiting for deps are normal case it should be on ds too IMHO[11:13:59] <hbraun> i don't understand what the gist is trying to explain[11:14:22] *** cvasilak has joined #jboss-as7[11:14:29] <maeste> adding a name to driver is helping a lot[11:15:01] <hbraun> can you explain the gist?[11:15:39] <maeste> I've added a property called "name" to driver tag into drivers[11:16:04] <hbraun> driver-class belongs to driver as well[11:16:19] <hbraun> shouldn't it be relocated as well?[11:16:25] <maeste> it permit you to refer to that driver using this symbolic name instead of DriverName#version[11:16:43] <hbraun> that's better, agreed[11:16:51] <maeste> the second file[11:17:04] <maeste> is changing the name from driver to jdbc-client[11:17:04] *** jamezp_afk has quit IRC[11:17:13] <maeste> much more clear inxa case IMHO[11:17:17] <hbraun> bad choice[11:17:20] *** tdiesler has quit IRC[11:17:21] <hbraun> ;)[11:17:29] <hbraun> don't agree with this one[11:17:32] <maeste> why?[11:17:44] <hbraun> client is an overloaded term[11:17:54] <maeste> oki, let's find another name[11:18:00] <hbraun> a jdbc-client is a component using the data source[11:18:15] <maeste> I'm just saying that driver is not the best for xa case[11:18:30] <maeste> where driver class is not used at all[11:19:26] <hbraun> ok let#s stick to driver until we have abetter name[11:19:41] <hbraun> shouldn't the driver-class go into that element as well?[11:19:46] <maeste> driver class could be not specified with jdbc 4 compliant driver[11:19:59] <maeste> but is needed for jdbc 3[11:20:23] <hbraun> https://gist.github.com/958679[11:20:43] <maeste> it can be moved there, even if in jdbc 3 you can have more than one in the same driver jar if I well remember[11:20:43] *** tdiesler has joined #jboss-as7[11:20:43] *** ChanServ sets mode: +v tdiesler[11:20:58] <maeste> so you have to choice which one at ds level[11:21:02] <hbraun> yes, but you are just using one[11:21:08] <maeste> per ds[11:21:16] <maeste> not per driver[11:21:22] <hbraun> do you think people are using it?[11:21:33] <hbraun> i mean this whole effort is about simplification[11:22:07] <hbraun> a datasource requires a driver[11:22:11] <maeste> yup, so we can simplify a lot declaring that driver-class is not needed for jdbc 4 driver[11:22:25] <hbraun> a driver consists of: driver-class, name, etc[11:22:41] <hbraun> agreed, it might become optional[11:23:12] <maeste> I would not spend a lot of time in config needed only for jdbc 3[11:23:28] <hbraun> i would expect people to create one or two drivers, whcih then are referenced by several data sources[11:23:31] <maeste> almost vendor ave a 4 compliant driver nowdays[11:23:58] <maeste> anyway I've got your idea[11:24:04] <hbraun> typically companies run a single kind of DB[11:24:11] <hbraun> sometimes different versions[11:24:16] <maeste> to do that we need some changes to IJ[11:24:24] <maeste> I have to discuss with jesper[11:24:47] <maeste> because in these days management require asap a stable 1.0 for satellite projects[11:25:05] <hbraun> i see[11:25:17] <maeste> I'm not sure we have enough time quite frankly[11:25:24] <hbraun> what about creating a custom schema for AS7? then translating into IJ schema?[11:25:49] <maeste> yup it's possible, even if we have to rewrite the parser[11:25:54] <hbraun> re-using the same schema makes sense for migrating between AS7 and standalone IJ[11:26:15] <hbraun> but on the other hand it creates a tight coupling[11:26:20] <maeste> yup[11:26:43] <maeste> the point is also on the parser of this schema that is unique atm[11:27:02] <hbraun> what do you mean?[11:27:12] <maeste> having just one schema[11:27:21] <maeste> we have the same parser in IJ and AS7[11:29:59] <maeste> I think we could move the drivers part into IJ too...but as said I have to check the plan in IJ and time we have left[11:31:12] <maeste> moving drivers into <datasources> tag[11:31:37] <maeste> so datasources will have a list of drivers and a list of datasource using one of them[11:32:17] *** liweinan has quit IRC[11:33:41] <maeste> hbraun: but doing that would require an IJ release too before it can be merged into master....[11:34:39] <hbraun> i understand[11:35:05] <hbraun> but then DS is a highly visible configuration[11:35:16] <maeste> hbraun: anyway I'm putting this discussion in a mail to jesper[11:35:25] <maeste> hbraun: you on CC[11:35:27] *** liweinan has joined #jboss-as7[11:35:28] *** galderz has quit IRC[11:35:53] <hbraun> IMO it there is much benefit in cleaning it up before AS goes Final[11:36:13] *** galderz has joined #jboss-as7[11:36:13] *** ChanServ sets mode: +v galderz[11:36:18] <hbraun> otherwise this will become the #1 hit in FAQ[11:36:41] <hbraun> even I don't understand it[11:36:51] <hbraun> as you can tell ;)[11:36:56] <maeste> hbraun: I almost agree. As said extending the discussion to jesper too[11:37:01] <hbraun> sure[11:37:12] <hbraun> sounds like a reasonable plan[11:40:19] <maeste> hbraun: mail sent you on CC[11:40:27] <hbraun> tnx[11:42:30] <maeste> hbraun: if you have it on hand, could you paste the exception for AS7-720 please?[11:42:31] <jbossbot> jira [AS7-720] Removing and re-adding datasource with same name throws Exception (domain) [Open (Unresolved) Bug, Major, Stefano Maestri] https://issues.jboss.org/browse/AS7-720[11:43:22] <maeste> it would help me to understand the problem and organize the work and pulls (maybe it's good to fix it with some other bug)[11:43:37] <maeste> hbraun: if it require time to you, just ignore me, I'll do myself[11:43:38] <hbraun> hang on[11:46:04] <hbraun> i don't get a reasonable exception[11:46:10] <hbraun> it#s simple http 500[11:46:31] <hbraun> but you can reproduce it through the CLI+[11:55:25] *** galderz has quit IRC[11:57:01] <maeste> hbraun: oki thx[12:05:21] *** rawblem has quit IRC[12:05:47] *** rawbdor has quit IRC[12:13:21] *** stliu has quit IRC[12:20:00] *** kkhan_ has joined #jboss-as7[12:21:49] *** kkhan has quit IRC[12:34:01] *** magesh1 has joined #jboss-as7[12:35:50] *** magesh has quit IRC[12:37:54] *** kkhan_ has quit IRC[12:38:15] *** kkhan has joined #jboss-as7[12:38:15] *** ChanServ sets mode: +v kkhan[12:55:42] *** AndyTaylor has joined #jboss-as7[12:55:42] *** ChanServ sets mode: +v AndyTaylor[12:57:44] *** JimMa has quit IRC[13:04:15] *** galderz has joined #jboss-as7[13:04:15] *** ChanServ sets mode: +v galderz[13:18:12] *** miclorb has joined #jboss-as7[13:22:21] *** sebersole has quit IRC[13:52:26] *** bgeorges has joined #jboss-as7[13:55:00] *** stliu has joined #jboss-as7[13:55:00] *** ChanServ sets mode: +v stliu[14:02:01] *** bgeorges has joined #jboss-as7[14:03:47] *** bgeorges has quit IRC[14:09:37] *** jpederse has joined #jboss-as7[14:24:35] *** tdiesler has quit IRC[14:29:01] *** kkhan is now known as kabir_luncheon[14:32:51] *** mmoyses has joined #jboss-as7[14:32:51] *** ChanServ sets mode: +v mmoyses[14:35:13] *** miclorb has quit IRC[14:38:51] *** liweinan has quit IRC[14:39:20] *** bgeorges has joined #jboss-as7[14:41:58] *** magesh1 has left #jboss-as7[14:43:06] *** kabir_luncheon is now known as kkhan[14:43:41] *** hardy has joined #jboss-as7[14:47:36] *** pferraro has joined #jboss-as7[14:47:36] *** ChanServ sets mode: +v pferraro[14:51:05] <hbraun> maeste: ping[14:51:15] <maeste> hbraun: pong[14:51:24] <hbraun> maeste: i cannot set a xa property list in dmr[14:51:29] <hbraun> do yu have an example?[14:51:35] <hbraun> the data structure seems weird[14:51:48] <hbraun> when reading i get this: "xa-data-source-properties" => {"URL" => "jdbc:h2:mem:test"}[14:52:06] <hbraun> an object with a property URL[14:52:37] <hbraun> but when writing it seems to expect a list of ModelNode type property[14:53:05] <maeste> hbraun: hmm I have to check...I've not written it[14:53:09] <hbraun> anyway, i cannot get it to work[14:53:19] <maeste> hbraun: let me have a look[14:54:58] *** pferraro has quit IRC[14:55:54] <hbraun> i cannto find a test case for this[14:56:30] <maeste> hbraun: there isn't[14:56:51] <hbraun> maeste: that's what I expected[14:56:55] <maeste> hbraun: :)[14:57:04] <hbraun> otherwise it would work i guess[14:57:21] <maeste> hbraun: let me write one, it's the best way to verify what have been done[14:57:24] <hbraun> p i t a[15:02:34] *** frainone has joined #jboss-as7[15:02:34] *** ChanServ sets mode: +v frainone[15:03:08] *** bbrowning has joined #jboss-as7[15:03:08] *** ChanServ sets mode: +v bbrowning[15:04:05] <bbrowning> In AS7 does HornetQ clustering still need a separate configuration from the JGroups config for Infinispan / session replication?[15:04:12] *** pferraro has joined #jboss-as7[15:04:12] *** ChanServ sets mode: +v pferraro[15:04:40] <dmlloyd> bbrowning: the guy to ask just connected :)[15:04:48] <dmlloyd> pferraro: <bbrowning> In AS7 does HornetQ clustering still need a separate configuration from the JGroups config for Infinispan / session replication?[15:04:56] <bbrowning> whee nice timing :D[15:05:17] <dmlloyd> guess it depends on how we define a "cluster"[15:06:04] <dmlloyd> might make sense to be able to define N cluster "memberships" and then assign subsystem clustering config to specific ones?[15:07:09] <bbrowning> hmm[15:07:39] <bbrowning> In my case, I define a cluster as N nodes with "/queues/abc" where I can publish a message on one node and receive it on any of the other nodes[15:07:54] *** pferraro has quit IRC[15:08:29] <bbrowning> but without pointing all the other nodes at a specific HornetQ running on node X[15:09:17] <bbrowning> I'm just trying to grasp all the pieces needed to port ToruqeBox from AS6 to AS7 :)[15:16:59] <hbraun> emuckenhuber: ping[15:17:03] <emuckenhuber> hbraun: pong[15:17:27] <hbraun> what was the reason for having a custom "add-system-property" operation again?[15:18:06] <hbraun> i am trying to figure out how to set a property list on a model node[15:19:01] *** pferraro has joined #jboss-as7[15:19:01] *** ChanServ sets mode: +v pferraro[15:20:28] <emuckenhuber> hbraun: hmm, you mean "add-system-property" as opposed to an ":add(value=foo)" ?[15:20:29] *** pferraro has quit IRC[15:21:04] <hbraun> emuckenhuber: i remember you implemented it differently for server-groups and configs[15:21:29] <hbraun> emuckenhuber: wondering if there was a reason that might expplain the trouble i am having[15:24:16] <emuckenhuber> hbraun: really? i can't remember... i thought for system properties it's consistent...[15:25:50] *** lgao has quit IRC[15:26:02] <hbraun> it's consistent in that way that there is a custom operation to add/remove them[15:26:50] <hbraun> opposed to transmitting them as part of the model representation when add() is invoked[15:27:03] <hbraun> the later is what I am trying to do on datasources[15:27:25] <hbraun> emuckenhuber: let me rephrase my question[15:27:48] <hbraun> emuckenhuber: how to I add a list of properties to a ModelNode in Java?[15:28:09] <emuckenhuber> node.add("key", "value")[15:28:34] *** pferraro has joined #jboss-as7[15:28:34] *** ChanServ sets mode: +v pferraro[15:28:35] <emuckenhuber> at least that's what i usually do[15:28:53] <emuckenhuber> not sure if it applies to all the models though[15:28:57] <hbraun> and in a subsequent step parent.get("name").set(node) ?[15:29:03] <hbraun> sounds reasonable[15:30:56] *** pferraro has quit IRC[15:31:44] <hbraun> maybe the server side op is wrong instead[15:42:47] <hbraun> emuckenhuber: tnx. it was a bug in an operation handler instead[15:42:52] <hbraun> sorry for bothering[15:44:15] *** ccrouch has joined #jboss-as7[15:44:15] *** ChanServ sets mode: +v ccrouch[15:53:56] <hbraun> maeste: https://gist.github.com/958985[15:54:12] <hbraun> maeste: i don't get it. I thnk driver is optional?[15:55:30] <maeste> hbraun: it is in xsd, but there is a part of the code where it is considered mandatory. I was exactly on this part atm[15:55:43] <hbraun> ;)[15:55:48] <hbraun> gotcha[15:55:51] <maeste> hbraun: becasue as said before[15:56:08] <maeste> hbraun: it is used to get jdbc service dependency[15:59:10] *** opalka has quit IRC[16:03:41] <hbraun> maeste: fyi https://issues.jboss.org/browse/AS7-721[16:03:42] <jbossbot> jira [AS7-721] XA datasource properties not parsed correctly [Open (Unresolved) Bug, Major, Heiko Braun] https://issues.jboss.org/browse/AS7-721[16:04:04] <hbraun> that#s the piece I was struggling with the whole afternoon[16:04:18] <hbraun> the diff explains the issue[16:04:31] <hbraun> now i am left with the driver issue[16:05:48] <maeste> hbraun: tthanks for the fix[16:06:03] <hbraun> np[16:06:10] <hbraun> i hope it's correct[16:06:11] <hbraun> ;)[16:06:19] <maeste> hbraun: the driver issue is not an issue...I mean the error is to consider it not mandatory[16:06:28] <maeste> hbraun: at least in AS7[16:06:50] <hbraun> that#s how the UI is implemented atm[16:07:04] <hbraun> so we just need to patch the op handler?[16:07:14] <maeste> hbraun: at the moment is always needed to have a reference to jar containing the xa class normally part of driver jar[16:07:22] <maeste> yup[16:07:28] <maeste> hbraun: I'll do[16:07:29] *** fnasser has joined #jboss-as7[16:07:35] <maeste> hbraun: right now[16:07:40] <maeste> hbraun: give me a min[16:09:27] <hbraun> btw, here's how it looks atm: http://imageshack.us/photo/my-images/853/xasamples.png/[16:09:51] <hbraun> this one is actually better: http://imageshack.us/f/853/xasamples.png/[16:11:47] <maeste> cool[16:13:33] <hbraun> took me a while to figure out how to split the different possible attributes across the screen[16:13:43] <hbraun> but I think the tabs work quiet well[16:14:58] <hbraun> maeste: here#s the regular datasource: http://imageshack.us/f/857/regulardssample.png/[16:15:07] <hbraun> quiet similar[16:20:45] *** mmoyses has quit IRC[16:21:41] <maeste> hbraun: in the operation description driver is already set REQUIRED=true[16:22:12] <hbraun> maeste: what do you mean?[16:22:16] *** cvasilak has quit IRC[16:24:45] <maeste> hbraun: https://github.com/jbossas/jboss-as/blob/master/connector/src/main/java/org/jboss/as/connector/subsystems/datasources/DataSourcesSubsystemProviders.java#L356[16:24:58] <maeste> hbraun: and the property for driver is[16:25:28] <maeste> hbraun: https://github.com/jbossas/jboss-as/blob/master/connector/src/main/java/org/jboss/as/connector/subsystems/datasources/AttributeDefinition.java#L36[16:25:36] <maeste> so required is set to true[16:25:56] <hbraun> we just remove it, no?[16:26:41] <maeste> hbraun: perhaps I'm tired, haven't we said that driver is mandatory and the ui is implemented this way?[16:26:58] <hbraun> no it#s the other way around[16:27:08] <hbraun> for XA the driver is not mandatory, is it?[16:27:18] <maeste> hbraun: but driver is needed IMHO[16:27:19] <hbraun> maybe we should call it a day[16:27:25] <hbraun> ;)[16:27:31] <hbraun> time for the weekend[16:27:47] <maeste> hbraun: it is needed because it's a reference to the jar containing xs class too[16:27:50] <maeste> xa[16:28:14] <hbraun> ok, so you need the datasource class, the driver and the version[16:28:15] <maeste> hbraun: probably your add works because you have a local ds already installed[16:28:17] *** rawbdor has joined #jboss-as7[16:28:30] <maeste> hbraun: yup[16:28:37] <hbraun> ok, fine+[16:28:38] *** rawblem has joined #jboss-as7[16:28:41] <hbraun> then we are good[16:28:48] <hbraun> i'll update the UI on monday[16:28:54] <maeste> oki have a nice we[16:29:00] *** jpederse has quit IRC[16:29:07] *** mmoyses has joined #jboss-as7[16:29:07] *** ChanServ sets mode: +v mmoyses[16:29:08] <hbraun> the remaining issue would then be the r7w for attributes[16:29:18] <hbraun> but I'll get back to it on moday[16:29:23] <hbraun> yeah, you too[16:29:26] <hbraun> thanks for the help[16:29:28] *** blviper has joined #jboss-as7[16:29:49] *** blviper has left #jboss-as7[16:31:21] *** hbraun has quit IRC[16:31:32] *** jamezp has joined #jboss-as7[16:32:51] *** pferraro has joined #jboss-as7[16:32:52] *** ChanServ sets mode: +v pferraro[16:33:09] *** pferraro has left #jboss-as7[16:35:07] *** mmoyses has quit IRC[16:36:30] <dmlloyd> can't believe we're over 700 JIRA issues already.[16:36:43] <dmlloyd> is that really good or really bad? :)[16:43:40] <jamezp> Good question... ...I guess we'll be busy :-)[16:45:06] *** pferraro has joined #jboss-as7[16:45:06] *** ChanServ sets mode: +v pferraro[16:45:26] *** pferraro has left #jboss-as7[16:55:48] *** hardy has quit IRC[16:58:33] *** emuckenhuber has quit IRC[17:15:42] *** JimMa has joined #jboss-as7[17:17:22] *** pgier has joined #jboss-as7[17:17:22] *** ChanServ sets mode: +v pgier[17:17:43] <pilhuhn> I wonder if the resource description should supply the default in a parseable way - especially if an attribute is not available via :read-resource[17:20:11] *** adietisheim has quit IRC[17:21:04] *** mmoyses has joined #jboss-as7[17:21:04] *** ChanServ sets mode: +v mmoyses[17:24:05] *** clebert has joined #jboss-as7[17:24:05] *** ChanServ sets mode: +v clebert[17:30:43] *** JimMa has quit IRC[17:31:44] *** jamezp has quit IRC[17:35:04] *** emuckenhuber has joined #jboss-as7[17:35:04] *** emuckenhuber has joined #jboss-as7[17:35:04] *** ChanServ sets mode: +v emuckenhuber[17:36:47] *** Jaikiran has joined #jboss-as7[17:36:47] *** ChanServ sets mode: +v Jaikiran[17:37:31] <Jaikiran> dmlloyd: https://github.com/jaikiran/jboss-as/commits/new-ee-framework[17:37:43] <Jaikiran> as discussed yesterday, i got that implemented for SFSBs[17:37:57] <Jaikiran> there's one remaining EJB singleton bean failure in smoke-tests which I'm looking into now[17:38:03] <Jaikiran> the rest of them pass[17:38:51] <dmlloyd> awesome!!![17:38:55] <dmlloyd> I'll look in a few minutes[17:39:16] *** jamezp has joined #jboss-as7[17:39:58] *** sannegrinovero has joined #jboss-as7[17:39:59] *** sannegrinovero has quit IRC[17:39:59] *** sannegrinovero has joined #jboss-as7[17:39:59] *** ChanServ sets mode: +v sannegrinovero[17:40:00] *** frainone is now known as frainone_lunch[17:44:12] *** pferraro has joined #jboss-as7[17:44:13] *** ChanServ sets mode: +v pferraro[17:46:10] *** pferraro has left #jboss-as7[17:48:12] <dmlloyd> Jaikiran: added a couple line notes.[17:48:19] * Jaikiran checks[17:49:15] <dmlloyd> otherwise looks good[17:51:26] <Jaikiran> done https://github.com/jaikiran/jboss-as/commits/new-ee-framework[17:51:38] *** kcbabo has joined #jboss-as7[17:51:38] *** ChanServ sets mode: +v kcbabo[17:51:42] <dmlloyd> great, thanks![17:52:22] <dmlloyd> grabbed it.[17:52:32] <dmlloyd> this thing is gonna be ready to merge soon[17:52:37] <Jaikiran> yep.[17:52:44] * dmlloyd is very glad :)[17:52:51] <Jaikiran> once the smoke tests pass, we should be able to push it upstream[17:53:00] <Jaikiran> i'll fix/debug the singleton issue in the next few minutes[17:53:18] <dmlloyd> cool, I'll run it by stuartdouglas if we get that fixed[17:54:03] *** pilhuhn is now known as pil-afk[17:54:51] <dmlloyd> okay I have to head out for a few hours[17:55:20] <Jaikiran> sure[17:55:46] <dmlloyd> thanks again for your help on this refactor[17:58:54] *** sannegrinovero has quit IRC[18:00:32] *** jcosta has quit IRC[18:02:10] *** smcgowan has joined #jboss-as7[18:02:11] *** ChanServ sets mode: +v smcgowan[18:02:21] *** stliu has quit IRC[18:08:42] *** alexsmirnov has joined #jboss-as7[18:15:55] *** kcbabo has quit IRC[18:17:55] *** mmoyses is now known as mmoyses_[18:27:14] *** smarlow has joined #jboss-as7[18:27:14] *** ChanServ sets mode: +v smarlow[18:32:34] *** asaldhan has joined #jboss-as7[18:32:34] *** ChanServ sets mode: +v asaldhan[18:38:54] *** jfclere has quit IRC[18:44:14] *** mmoyses_ is now known as mmoyses[18:51:10] *** jamezp is now known as jamezp_afk[18:57:45] *** galderz has quit IRC[19:01:55] <jamezp_afk> dmlloyd: Is there a special way the classpath should be set for MODULES-85?[19:01:57] <jbossbot> jira [MODULES-85] Add support for class path bootstrap [Open (Unresolved) Enhancement, Major, James Perkins] https://issues.jboss.org/browse/MODULES-85[19:04:25] *** frainone_lunch is now known as frainone[19:05:30] *** Jaikiran has quit IRC[19:13:58] *** kkhan has quit IRC[19:15:29] *** galderz has joined #jboss-as7[19:15:29] *** ChanServ sets mode: +v galderz[19:16:24] *** galderz has quit IRC[19:16:41] *** jamezp_afk is now known as jamezp[19:20:50] *** emuckenhuber has quit IRC[19:23:52] *** ccrouch has left #jboss-as7[19:48:19] <smcgowan> mmoyses: what is the value of the usersProperties relative to as in this case: http://pastebin.com/hEhh9Yn8 - the .properties files currently reside in $JBOSS_HOME/standalone/configuration[19:50:02] <mmoyses> smcgowan: let me test that real quick[19:50:11] <smcgowan> even prefixing with ${jboss.server.config.dir} was also unsuccessful - the deployment recognizes to use the cts config but can't find the file[19:55:09] <smcgowan> jamezp: welcome to jboss as team. have a question for you while you're here. i was trying to bring in jbossts v 4.15.0 and it had a dependency on jboss-logging-processor and jboss-logging-generator. are those going to be included in as 7[19:55:17] <smcgowan> http://pastebin.com/KjmbLHnC[19:56:15] <jamezp> smcgowan: Thanks. They should be needed for compile time only.[19:56:43] <jamezp> It's just annotation processor that runs at compile time not runtime.[19:57:15] <smcgowan> which they are - ok[19:57:35] <mmoyses> smcgowan: i don't think standalone/configuration is in the classpath and we don't parse for system properties on the option[19:57:43] <mmoyses> you need a full path name[19:58:07] <mmoyses> like /opt/cts/cts-users.properties or something like that[19:58:26] <mmoyses> or add those files to the web app so they can be found in the classpath under WEB-INF/classes[19:58:53] <smcgowan> mmoyses: i tried that too: http://pastebin.com/fzqkCsXW[19:59:20] <smcgowan> adding the files the web app is successful[20:02:13] <smcgowan> i wonder if they'd be found in bin[20:02:22] <mmoyses> checking the source code...[20:05:17] <mmoyses> it uses the context classloader to try and load the file so setting the full path name doesn't work indeed[20:06:00] <mmoyses> dmlloyd: are there any plans to have a directory included in the classpath, like the conf/ directory in previous versions?[20:14:22] *** maeste has quit IRC[20:23:12] *** maeste has joined #jboss-as7[20:23:12] *** ChanServ sets mode: +v maeste[20:28:42] *** rawblem has quit IRC[20:28:44] *** jwulf has quit IRC[20:28:54] *** smarlow has quit IRC[20:28:57] *** jamezp has left #jboss-as7[20:32:42] *** jamezp has joined #jboss-as7[20:35:27] *** asaldhan has quit IRC[20:46:31] *** jamezp has left #jboss-as7[20:46:55] *** jamezp has joined #jboss-as7[20:50:08] *** bgeorges has quit IRC[20:50:12] *** asaldhan has joined #jboss-as7[21:00:00] *** jamezp has left #jboss-as7[21:03:38] *** jamezp has joined #jboss-as7[21:08:00] *** jamezp has quit IRC[21:08:33] *** opalka has joined #jboss-as7[21:08:33] *** opalka has joined #jboss-as7[21:08:33] *** ChanServ sets mode: +v opalka[21:09:22] *** jamezp has joined #jboss-as7[21:15:42] *** maeste has quit IRC[21:15:49] *** jamezp has quit IRC[21:16:06] *** jamezp has joined #jboss-as7[21:17:15] *** mgoldmann has joined #jboss-as7[21:17:15] *** ChanServ sets mode: +v mgoldmann[21:17:56] *** pferraro has joined #jboss-as7[21:17:57] *** ChanServ sets mode: +v pferraro[21:20:20] *** pferraro has quit IRC[21:29:09] *** maeste has joined #jboss-as7[21:29:09] *** ChanServ sets mode: +v maeste[21:30:22] *** baileyje has joined #jboss-as7[21:30:22] *** ChanServ sets mode: +v baileyje[21:33:31] <dmlloyd> jamezp: yeah basically we're using the -cp/-classpath argument to set up a list of resource roots for a boot module. The boot module should be assembled using a JarModuleLoader like the -jar option does, but instead of using the MANIFEST to configure it uses CLI options[21:33:49] <dmlloyd> mmoyses: AS7 doesn't have a class path :)[21:34:04] <dmlloyd> mmoyses: you should not have any external config files, as a rule.[21:35:26] <mmoyses> right[21:37:50] <baileyje> dmlloyd: How goes it? I just got home.[21:38:42] <dmlloyd> hey, not bad. Just got home myself.[21:39:01] <dmlloyd> took the baby and the ladies out for some lunch and shopping[21:39:08] <baileyje> nice..[21:41:20] <baileyje> So. That interceptor bit. Did you see my comments?[21:41:26] *** pferraro has joined #jboss-as7[21:41:27] *** ChanServ sets mode: +v pferraro[21:41:44] <dmlloyd> yeah I merged your thing[21:41:48] <dmlloyd> looks fine[21:42:08] <dmlloyd> someday I"ll use that evil trick for something evil though :)[21:42:21] <baileyje> You have any idea on when we will get a jboss-invocation update?[21:42:39] <dmlloyd> I'll probably tag one when we merge the ejb branch back in[21:42:45] <dmlloyd> looks really close now, just a couple smoke tests[21:43:40] <baileyje> I had it in there, but I realized an interceptor isn't allowed to throw a throwable, and if a target method does, the invoking interceptor already converts to an undeclared anyway, so the nasty hack never really got invoked[21:46:44] *** pferraro has quit IRC[21:51:15] <dmlloyd> yeah we'd have to put the nasty hack in at the start of the chain too, and let's face it, that's a lot of nastiness[21:56:54] <baileyje> Yeah. And if someone really declared and then actually threw a raw throwable, they can deal with it.[21:57:01] <baileyje> since that is dumb..[21:58:03] <jamezp> dmlloyd: Thanks. That makes sense. I'll dig in now. Just about have everything configured finally :-)[21:58:34] *** jamezp has quit IRC[21:59:26] *** pferraro has joined #jboss-as7[21:59:27] *** ChanServ sets mode: +v pferraro[22:01:56] *** pferraro has quit IRC[22:12:13] *** pferraro has joined #jboss-as7[22:12:13] *** ChanServ sets mode: +v pferraro[22:14:37] *** pferraro has quit IRC[22:17:36] *** jamezp has joined #jboss-as7[22:31:00] *** pferraro has joined #jboss-as7[22:31:00] *** ChanServ sets mode: +v pferraro[22:31:20] *** pferraro has quit IRC[22:33:41] *** asoldano has joined #jboss-as7[22:33:41] *** ChanServ sets mode: +v asoldano[22:40:38] <asoldano> opalka, can you join #jbossws if you're online?[22:40:53] <opalka> asoldano, sure[22:42:43] *** asaldhan has left #jboss-as7[22:51:40] *** kkhan has joined #jboss-as7[22:51:41] *** ChanServ sets mode: +v kkhan[22:54:39] *** mmoyses has quit IRC[22:55:51] *** pferraro has joined #jboss-as7[22:55:51] *** ChanServ sets mode: +v pferraro[22:58:14] *** pferraro has quit IRC[23:01:44] *** pferraro has joined #jboss-as7[23:01:44] *** ChanServ sets mode: +v pferraro[23:03:27] *** pferraro has quit IRC[23:05:52] *** kkhan has quit IRC[23:40:04] *** clebert has quit IRC[23:44:32] *** mgoldmann has quit IRC[23:44:37] <stuartdouglas> morning[23:44:51] <jamezp> morning[23:45:34] *** maeste has quit IRC[23:48:23] <asoldano> stuartdouglas, still feels weird to me to see "morning" in the middle of the nigh :-P[23:48:30] <asoldano> s/nigh/night/[23:48:56] <stuartdouglas> which tz are you in?[23:56:10] <asoldano> stuartdouglas, CEST[23:56:21] <asoldano> stuartdouglas, basically, central europe (italy)[23:57:28] <opalka> asoldano, stuartdouglas now it's morning in Prague 00:00 ;)[23:57:56] <asoldano> opalka, lol[23:58:07] <stuartdouglas> It's saturday morning here[23:58:17] * stuartdouglas lives in the future[23:58:31] <asoldano> ;)[23:59:26] <opalka> ;)