[00:01:53] *** stuartdouglas has joined #jbosstesting
[00:34:25] *** bobmcw has quit IRC
[00:39:19] *** galderz has quit IRC
[00:41:45] *** galderz has joined #jbosstesting
[00:41:55] *** galderz has quit IRC
[01:00:09] *** tdiesler has quit IRC
[01:16:30] *** adriancole has joined #jbosstesting
[01:17:39] <adriancole> anybody know the as7 equiv of -b 0.0.0.0
[01:18:01] <adriancole> trying to switch jclouds tests to use as7, but I need to expose on the default interface, not loopback
[01:50:00] <adriancole> seems replacing the interface contents with any-address works..
[01:50:09] <adriancole> would be nice to know how to do this via a cli param
[02:01:55] *** gastaldi has joined #jbosstesting
[02:06:44] *** gastaldi has quit IRC
[02:21:40] *** PeteRoyle has quit IRC
[02:29:23] *** lightguard_jp has quit IRC
[02:40:18] *** PeteRoyle has joined #jbosstesting
[03:39:26] *** k4ffee has quit IRC
[03:39:45] *** stuartdouglas has quit IRC
[03:43:34] *** k4ffee has joined #jbosstesting
[03:56:09] *** stuartdouglas has joined #jbosstesting
[04:06:10] *** OndraZizka has quit IRC
[04:06:42] *** OndraZizka has joined #jbosstesting
[05:29:15] *** gastaldi has joined #jbosstesting
[05:29:27] <gastaldi> hey
[05:59:48] *** lightguard_jp has joined #jbosstesting
[06:24:00] *** gastaldi has left #jbosstesting
[06:27:18] *** PeteRoyle has quit IRC
[07:37:37] *** jose_freitas_aw has quit IRC
[07:38:37] *** jose_freitas_aw has joined #jbosstesting
[07:46:59] *** ge0ffrey has joined #jbosstesting
[08:04:21] *** kpiwko has joined #jbosstesting
[08:12:46] *** jharting has joined #jbosstesting
[08:53:45] *** torben has joined #jbosstesting
[08:53:45] *** torben has joined #jbosstesting
[08:56:14] *** OndraZizka has quit IRC
[09:02:08] *** jharting has quit IRC
[09:02:10] *** jharting1 has joined #jbosstesting
[09:04:51] *** jharting1 has quit IRC
[09:05:21] *** maeste has joined #jbosstesting
[09:07:17] *** jharting has joined #jbosstesting
[09:07:58] *** maschmid has joined #jbosstesting
[09:10:00] *** kpiwko1 has joined #jbosstesting
[09:10:30] *** OndraZizka has joined #jbosstesting
[09:11:32] *** kpiwko has quit IRC
[09:16:17] *** jeand has joined #jbosstesting
[09:28:33] *** lfryc has joined #jbosstesting
[09:32:19] *** aslak has joined #jbosstesting
[09:39:57] *** galderz has joined #jbosstesting
[09:41:07] *** Jaikiran has joined #jbosstesting
[09:55:04] *** tdiesler has joined #jbosstesting
[10:06:08] <kpiwko1> aslak: tdiesler: hi guys
[10:06:22] <tdiesler> hi
[10:06:43] <tdiesler> kpiwko1: ^
[10:06:43] <kpiwko1> aslak: tdiesler: we are just discussing if arquillian-domain based as7 container will make it to as7 final
[10:07:14] <tdiesler> kpiwko1, who is 'we'?
[10:07:41] <kpiwko1> tdiesler: qe guys and Heiko
[10:08:43] *** kpiwko1 is now known as kpiwko
[10:09:05] <tdiesler> kpiwko1, ok. I don't know the status of the top of my head - could you remind me? what are the missing features?
[10:10:08] <tdiesler> kpiwko, it would help if we could talk about a set of jiras
[10:10:32] *** galderz has quit IRC
[10:11:21] <kpiwko> tdiesler: well, I don't know neither...but I see there are some sub modules in arquillian in AS7 named *domain, but I have no information how these are supported
[10:13:02] <kpiwko> tdiesler: or how they've been tested
[10:15:27] <tdiesler> kpiwko, I guess you want to talk to brian https://github.com/tdiesler/jboss-as/commits/master/arquillian/container-managed-domain
[10:16:29] <kpiwko> tdiesler: ok, thanks, I'll ask him then
[10:27:31] *** PeteRoyle has joined #jbosstesting
[10:28:37] *** adinn has joined #jbosstesting
[10:28:46] *** adinn has left #jbosstesting
[11:03:41] *** maschmid is now known as maschmid_afk
[11:06:16] *** tdiesler is now known as tdiesler_afk
[11:26:18] *** galderz has joined #jbosstesting
[11:27:25] *** galderz has quit IRC
[11:39:26] <maeste> aslak: may I ask your help to understand an ARQ exception?
[11:41:30] <aslak> maeste, sure
[11:42:12] <maeste> aslak: https://gist.github.com/1058178
[11:42:54] <maeste> aslak: I'm getting this exception, but I'm creating the archive from a real jar having correct manifest
[11:43:00] <maeste> aslak: what am I missing?
[11:43:06] <aslak> tdiesler_afk, ^
[11:44:30] <aslak> maeste, it seems to not find META-INF/MANIFEST.MF
[11:44:48] <maeste> aslak: yup, but the jar have it
[11:45:31] <maeste> aslak: let me see how Archive appear with a debugger
[11:51:19] <maeste> aslak: archive.get("META-INF"); got a correct children list containing also MANIFES.MF
[11:51:46] <maeste> *MANIFEST.MF
[11:57:45] *** tdiesler_afk is now known as tdiesler
[11:58:23] *** galderz has joined #jbosstesting
[11:59:43] *** lightguard_jp is now known as lightguard_jp_aw
[12:00:08] <tdiesler> maeste, is your intention to package an osgi bundle?
[12:00:09] <aslak> maeste, can you debug the AbstractOSGiApplicationArchiveProcessor
[12:02:03] <tdiesler> aslak, I'm looking at http://maven.apache.org/guides/mini/guide-attached-tests.html Have you tried that? I get "No tests to run"
[12:03:54] <aslak> tdiesler, no i haven't tested it yet, tho it's been in my plans to do so.. :)
[12:04:22] <aslak> tdiesler, was planning to convert the examples in core to a basic test suite for the contianers, following that same route
[12:04:35] <maeste> tdiesler: no, it's a normal jar for me, but it is mysql latest driver that is an OSGi bundle AFAIK
[12:05:01] *** galderz has quit IRC
[12:07:52] <tdiesler> maeste, the content of the manifest defines what it is - not what you want it to be ;-) So it is an osgi bundle
[12:09:37] * tdiesler digging up the code
[12:10:06] <maeste> tdiesler: yup sure, I'm verifying. I meant that I want to use a normal jar, and I haven't verified if it is an OSGi.
[12:10:20] *** maeste is now known as maeste_afk
[12:12:16] <tdiesler> maeste, the code that this is running through means that the archive is an osgi bundle that contains the arq test.
[12:12:27] <maeste_afk> tdiesler: oki removing OSGi infos, it works. So the problem is creating Archive from rela jar that is OSGi bundle
[12:12:57] <maeste_afk> tdiesler: yup, I've just changed the MANIFEST.MF
[12:13:48] <tdiesler> maeste, if you wnat to deploy this as non testable jar you need to use @Deployment(testable=false)
[12:14:23] <tdiesler> maeste, the jar your deploying should not get enhanced as arq test deployment I think
[12:21:02] *** maschmid_afk is now known as maschmid
[12:21:45] <maeste_afk> tdiesler: oki, I'll try this too. Thanks
[12:33:13] *** aslak is now known as aslak_afk
[12:35:16] *** mgoldmann has joined #jbosstesting
[12:35:56] *** pmuir has joined #jbosstesting
[12:36:00] <pmuir> aslak_afk: ping
[12:49:50] *** maeste_afk has quit IRC
[12:51:16] *** kpiwko has quit IRC
[13:02:08] *** maeste_afk has joined #jbosstesting
[13:06:03] *** kpiwko has joined #jbosstesting
[13:27:21] *** galderz has joined #jbosstesting
[13:38:37] <tdiesler> aslak, I need an osgi container release for jbosgi. how can we go about it. https://github.com/arquillian/arquillian-container-osgi/pull/7
[13:41:21] *** Jaikiran is now known as Jaikiran|AFK
[14:06:26] *** jose_freitas_aw has quit IRC
[14:13:36] *** jose_freitas_aw has joined #jbosstesting
[14:18:44] *** alesj has joined #jbosstesting
[14:23:39] *** vtunka has joined #jbosstesting
[15:17:45] *** torben has quit IRC
[15:30:53] *** k4ffee has quit IRC
[15:31:23] *** kpiwko has quit IRC
[15:31:26] *** alesj has quit IRC
[15:38:28] *** maeste_afk has quit IRC
[15:42:32] *** jharting has quit IRC
[15:59:50] *** maschmid has quit IRC
[16:03:47] *** pilhuhn has joined #jbosstesting
[16:04:38] <pilhuhn> Hey,
[16:05:05] <pilhuhn> I am currently trying to set up Arquillian to start AS7 for some integration tests with the RHQ as7 plugin
[16:05:51] <pilhuhn> I am basically following the example of the kitchensink (which works), but AS7 does not start in my case. I suspect that our use of TestNG instead of Junit may cause this
[16:05:58] <pilhuhn> any ideas how to tackle this?
[16:08:10] <pmuir> got any code?
[16:08:14] <pmuir> to show
[16:09:20] <pilhuhn> not sure .. my tries are on my local box right now
[16:09:38] <pilhuhn> I found a
[16:09:39] <pilhuhn> <groupId>org.jboss.arquillian</groupId>
[16:09:39] <pilhuhn>             <artifactId>arquillian-testng</artifactId>
[16:09:54] <pilhuhn> thingy meanwhile, but this is still Alpha5
[16:10:05] <pilhuhn> and it is missing the @RunWith from junit4
[16:10:17] <pilhuhn> (or better testng is missing it)
[16:10:55] <pmuir> try org.jboss.arquillian.testng:arquillian-testng-container
[16:10:58] <pilhuhn> and when I convert my whole test class to use junit's @Test, I get no coverage at all (but also no test failures :-> )
[16:11:05] <pmuir> @RunWith is a junit annotation
[16:11:14] <pmuir> if you are using testng you need to extend a base class
[16:13:23] <pilhuhn> *sigh*
[16:13:30] <pmuir> extends Arquillian
[16:13:41] <pilhuhn> you see ... that is why I said "simple" at the very start
[16:13:42] <pmuir> yeah, testng is not as good as junit unfortunately
[16:14:01] <pmuir> pilhuhn: it depends on what simple means for you
[16:14:15] <pilhuhn> sure
[16:14:36] <pmuir> this is simple but I can see if you have an stick with an existing set of tests it's not so straightforward
[16:14:37] <pilhuhn> I think I'll tell you over a beer ...
[16:15:00] <pmuir> :-)
[16:15:15] <pmuir> talk to alr, he may know what you want or kkhan
[16:15:15] <pilhuhn> Don't get me wrong - I don't want to bash Arq in any way
[16:16:44] <pilhuhn> Don't get me wrong - I don't want to bash Arq in any way
[16:17:20] <pmuir> :-)
[16:17:48] <pmuir> it's also tougher to use with testng due to the somewhat less powerful extension points it exposes compared to junit
[16:18:25] <jose_freitas_aw> imho testng made sense before junit4
[16:18:35] *** jose_freitas_aw is now known as jose_freitas
[16:18:44] <pilhuhn> jose_freitas I think so too
[16:19:01] <pilhuhn> pmuir Ok, I got it to start now. Not in the 1.2s range but anyway :-)
[16:19:16] <pmuir> haha
[16:19:21] <pilhuhn> Is there a way to start as7 in domain instead of standalone mode?
[16:19:49] <pmuir> not for arquillian - OndrejZizka/OndraZizka is maybe working on this?
[16:20:43] <pmuir> jose_freitas: pilhuhn: yes unfortunately testng has a lot of features that junit does not, because junit exposes much more powerful hooks so you can just extend it, also testng internally is, um, a little bit like a popular italian dish
[16:21:29] <pilhuhn> totally flat with tomato, mozarella and anchovis? :->
[16:22:18] <jose_freitas> hahaha
[16:23:55] <jose_freitas> I think it's more like spaguethi, but anyway, testng has some neat features, but I don't miss it at all
[16:28:16] <pilhuhn> java.lang.NullPointerException
[16:28:16] <pilhuhn> 	at org.jboss.arquillian.container.impl.client.container.ContainerDeployController.forEachDeployedDeployment(ContainerDeployController.java:228)
[16:28:24] *** nilian has joined #jbosstesting
[16:28:28] <pilhuhn> probably as I did not deploy anything before
[16:28:35] * pilhuhn is writing a bug report
[16:39:52] *** vtunka has quit IRC
[17:02:10] * pilhuhn is considering to setup junit testing for this one module
[17:18:21] *** oskutka has joined #jbosstesting
[17:18:50] *** jharting has joined #jbosstesting
[17:22:40] *** tdiesler has quit IRC
[17:33:39] *** galderz has quit IRC
[17:43:27] *** lfryc has quit IRC
[17:44:16] *** galderz has joined #jbosstesting
[17:47:20] *** nilian has quit IRC
[17:51:21] *** vtunka has joined #jbosstesting
[17:52:13] <aslak_afk> pmuir, delayed home from doctor pong
[17:52:17] <aslak_afk> :)
[17:52:19] *** aslak_afk is now known as aslak
[17:53:46] *** oskutka has quit IRC
[18:00:49] *** vtunka has quit IRC
[18:05:54] *** pilhuhn is now known as pil-gone
[18:08:52] *** vtunka has joined #jbosstesting
[18:09:18] *** vtunka has quit IRC
[18:20:21] *** ALR has joined #jbosstesting
[18:28:26] <pmuir> aslak: it's ok
[18:28:30] <pmuir> all over now baby blue
[18:29:48] <aslak> pmuir, what's over? :)
[18:30:19] <pmuir> the ping
[18:31:15] <aslak> hehe.. i meant the original reason for the ping.. but that's ok, i don't have to know. :)
[18:40:41] *** ge0ffrey has quit IRC
[18:53:28] *** lightguard_jp_aw is now known as lightguard_jp
[18:55:53] *** pmuir has quit IRC
[18:56:29] <OndrejZizka> pil-gone:  There's a chance I will work on that. I'd like to.
[18:59:22] *** aslak has quit IRC
[18:59:59] *** aslak has joined #jbosstesting
[19:00:44] *** aslak has quit IRC
[19:33:05] *** galderz has quit IRC
[20:07:03] *** Jaikiran|AFK has quit IRC
[20:33:15] *** lfryc has joined #jbosstesting
[20:34:50] *** Tashtego has joined #jbosstesting
[20:36:49] *** aslak has joined #jbosstesting
[20:48:05] *** lfryc has quit IRC
[21:19:13] <adriancole> repost from yesterday... anybody know the as7 equiv of -b 0.0.0.0
[21:19:46] <adriancole> currently, I'm using sed to replace standalone.xml interface contents with any-address, albeit hacky
[21:34:16] <aslak> adriancole, not sure.. :)
[21:40:48] *** bobmcw has joined #jbosstesting
[21:42:55] <jose_freitas> hey aslak :)
[21:42:57] <adriancole> aslak no biggie right now, as sed works
[21:45:14] <aslak> jose_freitas, heya
[21:48:36] <aslak> jose_freitas, http://community.jboss.org/message/612986#612986
[21:49:36] <jose_freitas> I'm working on it right now
[21:49:50] <jose_freitas> I downloaded the code from jacoco and I'm digging it
[21:50:37] <jose_freitas> there's something wrong with th way the data is stored on our byteArray
[21:51:23] <jose_freitas> I'm making simple tests outside arquillian and it might be a problem on jacoco
[21:51:38] <jose_freitas> I could reproduce the same behaviour outside it
[21:52:07] <jose_freitas> I hope to get it working soon
[21:52:50] <jose_freitas> :)
[21:53:38] <jose_freitas> So, since on the branch we're already bind it to arq-final, maybe we could hope to have within arquillian final version. when do you plan to release it?
[21:57:39] <aslak> jose_freitas, sunday
[21:58:21] <jose_freitas> really?
[21:58:31] <jose_freitas> 1.0.0.Final?
[21:58:35] <aslak> yea
[21:58:40] <jose_freitas> july 3th?
[21:58:44] <aslak> yea
[21:59:06] <jose_freitas> :D
[21:59:15] <aslak> some small bugs, some small exception improvements
[21:59:25] <jose_freitas> so maybe a couple of days after
[21:59:37] <jose_freitas> I won't have much time this weekend
[21:59:42] *** bobmcw has quit IRC
[22:00:16] <jose_freitas> but I feel that the success finish of the extension is near
[22:00:46] <aslak> jose_freitas, no rush with the extension, probably only Core out in Final on sunday
[22:00:52] <aslak> in/on
[22:01:56] <jose_freitas> :)
[22:08:07] <aslak> jose_freitas, what's the current issue?
[22:08:33] <aslak> jose_freitas, keep in mind that the original instrumenter etc was written for the .3 or .4 release, some stuff might have changed in .5
[22:10:41] <jose_freitas> well, it's instrumenting right. but when we're going to write the data on the buffer
[22:11:15] 
[22:11:35] <jose_freitas> maybe there's a step missing
[22:11:40] <jose_freitas> or even a bug on jacoco
[22:11:48] <jose_freitas> I made a simple example
[22:11:59] <jose_freitas> working with executinDataStore and executionDataWritter
[22:12:05] <jose_freitas> with datastore it worked
[22:12:14] <jose_freitas> with dataWritter dont
[22:12:23] <jose_freitas> I contact jacoco guys
[22:12:25] <jose_freitas> contacted
[22:12:46] <jose_freitas> maybe they can give us a hint of what we're missing
[22:13:03] <jose_freitas> this is the last issue
[22:13:12] <jose_freitas> or at least, I believe so
[22:13:17] <aslak> yea, it's only instrumenting pr method it seems
[22:15:08] <jose_freitas> I guess that instrumentation is right
[22:15:23] *** mgoldmann has quit IRC
[22:15:43] <jose_freitas> I decompiled CoverageTestBean.class that you creat for testing purposes
[22:15:56] <jose_freitas> and there's code like $jacocoInit();
[22:16:02] <jose_freitas> on public methods
[22:16:55] <jose_freitas> the good thing is that I could isolate the behaviour outside the extension
[22:17:13] <jose_freitas> so, on my simple tests, I'm not using our InstrumenterAsset
[22:18:09] <aslak> But your using ArquillianRuntime?
[22:18:10] <jose_freitas> and since with the same instrumentation I could see a report using executionDataStore
[22:18:13] <jose_freitas> nope
[22:18:26] <jose_freitas> using a jacoco abstract
[22:18:30] <jose_freitas> http://pastebin.com/y5PvJNCq
[22:18:55] <jose_freitas> abstractRuntime
[22:19:43] <jose_freitas> then my report work outside the extension
[22:19:49] <jose_freitas> I'll replicate it inside
[22:19:59] <jose_freitas> sorry
[22:20:06] <jose_freitas> rewritting
[22:20:19] <jose_freitas> then, when my report work outside the extension
[22:23:49] <aslak> mm
[22:26:38] <aslak> oh my.. i had forgotten how slow svn is
[22:27:10] <jose_freitas> hehehe
[22:27:19] <jose_freitas> downloading jacoco too?
[22:27:24] <aslak> hehe yea
[22:28:02] <jose_freitas> :)
[22:28:45] <jose_freitas> I based my example on http://www.eclemma.org/jacoco/trunk/doc/examples/java/CoreTutorial.java
[22:28:57] <jose_freitas> on this example it runs a report
[22:29:31] <jose_freitas> what I did is too replace runtime.collect(executionData, null, false);
[22:29:31] <jose_freitas> 		runtime.shutdown();
[22:29:53] <jose_freitas> with our collect(executionDataWritter);
[22:30:24] <jose_freitas> the data I got was the same
[22:30:35] <jose_freitas> that we got with arquillian extension
[22:30:40] <jose_freitas> and debugging it
[22:30:43] <jose_freitas> the data seems ok
[22:30:51] <aslak> so still no line data ?
[22:30:57] <jose_freitas> no
[22:31:16] <jose_freitas> so, or we're missing a step
[22:31:18] <aslak> i see there is a 0.53 out..
[22:31:39] <jose_freitas> or it could be a bug when using dataWritter
[22:32:49] <jose_freitas> one thing that I'm trying now
[22:33:16] <jose_freitas> is to use the analyze tool
[22:33:24] <jose_freitas> I think it's the missing link
[22:33:49] <jose_freitas> maybe it's not used to build the report but the data itself
[22:33:56] <jose_freitas> An Analyzer instance processes a set of Java class files and calculates coverage data for them. For each class file the result is reported to a given ICoverageVisitor instance. In addition the Analyzer requires a ExecutionDataStore instance that holds the execution data for the classes to analyze. The Analyzer offers several methods to analyze classes from a variety of sources.
[22:34:16] <jose_freitas> but the api don't help very much when working with executinDataWritter
[22:36:22] <jose_freitas> http://www.eclemma.org/jacoco/trunk/doc/api/index.html
[22:39:26] *** pgmjsd has left #jbosstesting
[22:43:42] *** nilian has joined #jbosstesting
[22:45:05] *** ALR has quit IRC
[22:53:16] *** Tashtego has quit IRC
[22:54:07] <jose_freitas> aslak ?
[22:54:13] <aslak> jose_freitas, reading .. :)
[22:54:30] <jose_freitas> http://pastebin.com/Tpuj3JhW
[22:55:07] <jose_freitas> I made some cleaning
[22:55:38] <jose_freitas> so, this is what basically the extension do
[22:55:43] <jose_freitas> with a different runtime
[22:58:28] <jose_freitas> just put on a project with jacoco libs and you have it running
[22:58:51] <jose_freitas> I need to go home
[22:58:55] <jose_freitas> catch you up later
[22:59:07] *** jose_freitas has quit IRC
[23:15:00] *** danielbremer-ton has joined #jbosstesting
[23:15:12] *** nilian has quit IRC
[23:18:08] <adriancole> aslak I tested several clouds cheapie instances... first-time startup of as7 until port 8080 is listening is 4-9seconds
[23:18:34] <adriancole> this is the time inclusive of starting the jvm, not what the logs say
[23:21:27] <adriancole> fastest complete bootstrap from base os (including install) is 1:08
[23:21:29] <adriancole> which is amazon micro/amazon linux
[23:22:05] <adriancole> 2nd place rackspace at 1:22
[23:24:27] <aslak> adriancole, doable.. :)
[23:24:34] <adriancole> 3rd elastichosts 2:48
[23:24:43] <aslak> adriancole, do you remember the as6 times?
[23:24:48] <adriancole> well
[23:24:54] <adriancole> my changes aren't pushed
[23:24:58] <aslak> i have 5 min in my mind, but i'm not sure if that is correct
[23:25:00] <adriancole> I can run the same test on my laptop
[23:25:22] <adriancole> most clouds we are at 3-5m
[23:25:28] <adriancole> now.. I mean
[23:25:41] <adriancole> and with some tuning could get a bit better across the board
[23:25:53] <adriancole> I'll push my changes soon, but run all tests before/after
[23:25:54] <aslak> adriancole, that is from cold start right? what about hibernation?
[23:26:04] <adriancole> not tested right now..
[23:26:14] <adriancole> I suspect it is less
[23:26:24] <adriancole> we can start with apples/apples
[23:26:30] <adriancole> cold 7 vs cold 6
[23:26:33] <aslak> yea
[23:26:34] <adriancole> then start playing
[23:26:43] <adriancole> also.. I might switch to bz2
[23:26:59] <adriancole> a lot of the time is extracting the 85m dist
[23:27:02] <aslak> adriancole, by cold, you mean from you do the initial start call, or jsut when ever the cloud gets around to starting it until it's up?
[23:27:15] <adriancole> no machine to port 80080
[23:27:20] <adriancole> this is the race
[23:27:24] <aslak> right
[23:27:39] <adriancole> should be much better anyway
[23:27:42] <adriancole> the instances are puny
[23:27:44] <aslak> using google storage for dist distribution?
[23:27:50] <adriancole> no amazon cloudfront
[23:27:57] <adriancole> paying for it, but worth for speed
[23:28:01] <aslak> :)
[23:28:03] <adriancole> esp in EU
[23:28:17] <aslak> where you at btw?
[23:28:20] <adriancole> SF
[23:28:26] <aslak> aa :)
[23:28:32] <adriancole> :) also.. thinkning
[23:28:37] <adriancole> for basic jboss as7
[23:28:43] <adriancole> can we trim the heap a bit?
[23:28:55] <aslak> down up?
[23:29:00] <adriancole> this is the settings now: -Xms64m -Xmx512m -XX:MaxPermSize=256m
[23:29:21] <adriancole> if just doing a small http test it could move faster w/ less mem resources
[23:29:31] <adriancole> some of these instances are only 640m total ram
[23:29:35] <adriancole> or less
[23:29:48] <adriancole> thinking if we can safely reduce ...
[23:30:02] <adriancole> I guess only min heap..
[23:30:09] <aslak> sure, but does the allocation take that long?
[23:30:20] <adriancole> well.. if in swap it would
[23:30:31] *** danielbremer-ton has quit IRC
[23:30:33] <adriancole> but this is shooting in the dark
[23:30:51] <adriancole> we can tune later.. unfair anyway
[23:30:51] *** nilian has joined #jbosstesting
[23:31:10] <aslak> adriancole, have you looked at the speed showdown page?
[23:31:55] <aslak> adriancole, a few vm args there that will take the boot down a about 1 sec (2 being the 'normal')
[23:32:19] <adriancole> aslak can you resend me the link?
[23:33:11] <aslak> adriancole, seems the current record is .97
[23:33:14] <aslak> adriancole, http://community.jboss.org/wiki/AS7StartupTimeShowdown
[23:35:34] *** jose_freitas has joined #jbosstesting
[23:35:38] <jose_freitas> hey aslak
[23:35:42] <adriancole> should put flags used in that wiki
[23:35:43] <aslak> hey
[23:36:00] 
[23:36:10] 
[23:36:15] <jose_freitas> the dummy.data
[23:36:25] <jose_freitas> this is the data that we need to work
[23:36:30] <adriancole> oh ic
[23:36:59] <jose_freitas> sorry, should have cleaned that up too
[23:37:05] <aslak> jose_freitas, sure, but you have this: final Analyzer analyzer = new Analyzer(executionData, coverageBuilder);
[23:37:24] <aslak> jose_freitas, executionData is just = new ExecutionDataStore()
[23:37:28] <jose_freitas> yes
[23:37:55] <jose_freitas> I believe that this is used to create the report data
[23:38:11] <aslak> it basically 'proves' that using ArquillianRuntime and outputting the bytes and reading them again contains the correct data
[23:38:11] <adriancole> aslak thanks.. I think we could change a few things moving forward including switch to openjdk 32 bit
[23:38:26] <aslak> adriancole, yea.. :)
[23:38:58] 
[23:39:21] <jose_freitas> the analyze is not analysing the ExecutionDataWritter
[23:39:32] <jose_freitas> which is where we wrote the data
[23:40:21] <aslak> jose_freitas, sure, incontainer we use the Writer to output to our ByteArrayOutputStream, for then to send the data back to the Client. the Client then just writes tohse to disk
[23:40:34] <jose_freitas> exactly
[23:40:35] <aslak> jose_freitas, the next step is for whom ever is going to make the report, to read those data
[23:40:46] <jose_freitas> yes
[23:40:54] <jose_freitas> but to report work
[23:40:55] <aslak> so, i ntheory, if you write the data to where the sonar-jacoco plugin is reading the data.. it should work
[23:41:06] <jose_freitas> the dummy.data has to keep some data
[23:41:14] <jose_freitas> and it does not
[23:41:50] <jose_freitas> the dummy.data is almost empty
[23:41:54] <jose_freitas> it has just the class name
[23:42:08] <jose_freitas> and some binary data
[23:42:17] 
[23:42:45] <jose_freitas> but I was looking in the java agent
[23:42:47] <aslak> the storage format is classanme+ method name + a boolean array of line number hits
[23:42:52] <jose_freitas> yes
[23:43:04] <jose_freitas> right
[23:43:12] <adriancole> aslak as of last night a couple startups were between 22-24s quicker.. by tonight I'll have the real diff
[23:43:24] <jose_freitas> but that data should be transformed to something else
[23:43:29] <aslak> jose_freitas, have you tried Reading and Reporting on the dummy.data ?
[23:43:47] <jose_freitas> yes
[23:43:49] <aslak> jose_freitas, alla what the test we havve here does?
[23:43:58] <jose_freitas> empty
[23:44:50] <jose_freitas> Try to generate using this
[23:44:50] <jose_freitas> http://www.eclemma.org/jacoco/trunk/doc/examples/java/ReportGenerator.java
[23:45:08] <jose_freitas> set it to read dummy.data
[23:45:26] <jose_freitas> actually I think I renamed it too jacoco.exec
[23:45:35] <jose_freitas> and passed the dir
[23:45:40] <jose_freitas> as main arg
[23:47:28] <adriancole> aslak does as7 have a deploy api now? or is it still the case you need to scp, twiddle, etc
[23:48:44] <aslak> jose_freitas, yea, just tried the same.. hmm
[23:49:02] <jose_freitas> I think the dummy.data has to pass to one step more
[23:49:04] <aslak> adriancole, no, Rest based .. :)
[23:49:20] <jose_freitas> if you look to a jacoco.exec file (generated by sonar for example)
[23:49:26] <jose_freitas> it has hit line information
[23:49:33] 
[23:49:46] 
[23:49:48] <jose_freitas> hehehe
[23:50:46] <adriancole> aslak is there a rest spec handy?  I see isseus and such
[23:53:26] <aslak> adriancole, not that i know of..
[23:57:19] *** jharting has quit IRC





top