14:30:49 <edwarnicke> #startmeeting Karaf Happy Hour 14:30:49 <odl_meetbot> Meeting started Sun Aug 31 14:30:49 2014 UTC. The chair is edwarnicke. Information about MeetBot at http://ci.openstack.org/meetbot.html. 14:30:49 <odl_meetbot> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:30:49 <odl_meetbot> The meeting name has been set to 'karaf_happy_hour' 14:30:53 <edwarnicke> #topic Roll Call 14:30:58 <edwarnicke> #info edwarnicke 14:31:03 <edwarnicke> Please #info in :) 14:31:10 <vjanandr> #info vjanandr for snbi 14:31:29 <xsited> #info thomas for pcmm 14:32:06 <edwarnicke> Anybody else? 14:32:50 <edwarnicke> I'll leave rollcall open till x:35 and then we can go to status 14:35:32 <edwarnicke> #topic Status 14:35:41 <edwarnicke> OK... so where is everyone? 14:36:39 <edwarnicke> vjanandr: xsited Where are you wrt your karaf feature files? 14:38:18 <vjanandr> edwarnicke: able to add a test CLI to karaf.. 14:38:33 <edwarnicke> vjanandr: Cool :) 14:38:38 <vjanandr> now extending the shell to support some snbi test clis.. 14:38:41 <edwarnicke> So you have a working feature file then? 14:38:45 <vjanandr> yes 14:39:30 <edwarnicke> Excellent :) 14:39:39 <edwarnicke> Is it merged up in your master repo at ODL? 14:39:58 <vjanandr> not yet.. still testing.. once I am done with testing.. I will send it for review.. 14:40:18 <edwarnicke> vjanandr: Cool :) 14:40:18 <xsited> working, checked in, features-packetcable .............................. SUCCESS 14:40:34 <edwarnicke> vjanandr: Feel free to ping me on IRC if you'd like another set of eyes :) (but don't block on me ;) ) 14:40:45 <edwarnicke> xsited: Awesome! 14:40:46 <vjanandr> edwarnicke: sure.. will do.. 14:40:53 <edwarnicke> xsited: Have you gotten your feature into integration? 14:41:09 <xsited> meaning? 14:42:24 <edwarnicke> #link https://wiki.opendaylight.org/view/Karaf:Step_by_Step_Guide#Prepare_your_feature_for_addition_to_the_integration_features_index 14:43:11 <xsited> lemme read 14:45:23 <edwarnicke> xsited: :) 14:47:06 <xsited> Review 5.4 not and I could stand to clean up the comments in features/pom.xml. 14:47:47 <edwarnicke> xsited: I don't understand "Review 5.4" 14:48:06 <xsited> in generally yes, but I am open to review. I do however feel a little karaf smarter today. 14:48:54 <edwarnicke> karaf smarter is good :) 14:49:04 <edwarnicke> Bug I still am not following "Review 5.4 not and I could stand to clean up the comments in features/pom.xml." 14:49:07 <xsited> in the "what to review" section — my commit message did not say exactly that. 14:49:26 <edwarnicke> Oh... you mean your commit message to integration ? 14:49:40 <edwarnicke> Or your commit message for your commit to packetcable ? 14:49:54 <xsited> it said more like "damn it's done and so am i" 14:51:23 <xsited> commit to repo. does that mean a courtesy email to the integration team with status? If I would be happy to do that. 14:52:36 <edwarnicke> xsited: You need to push a patch to integration to bring your feature in there as described in https://wiki.opendaylight.org/view/Karaf:Step_by_Step_Guide#Prepare_your_feature_for_addition_to_the_integration_features_index 14:54:22 <edwarnicke> vjanandr: Are you set for today, or do you need other assistance? 14:57:14 <vjanandr> edwarnicke: yes I think I should be done by today.. 14:57:45 <edwarnicke> vjanandr: My question was if you needed anymore help right now :) 14:57:57 <edwarnicke> vjanandr: Want to make sure you get the help you need :) 14:58:43 <vjanandr> edwarnicke: nope.. I think I understand karaf and features better now :) 14:58:49 <vjanandr> google was lot of help :) 14:59:07 <vjanandr> and of course you and Madhu :) 14:59:28 <vjanandr> will try to get it done by today.. 14:59:39 <vjanandr> will reach out to you on IRC.. if I run into some issues.. 15:00:04 <edwarnicke> vjanandr: Also fee free to text me at 919-621-0958 if you don't find me on IRC :) 15:00:15 <vjanandr> sure will do thanks.. 15:04:18 <edwarnicke> xsited: Do the directions about getting into integration make sense? 15:10:07 <xsited> i will work on that and ask more questions if I get stuck. 15:13:59 <edwarnicke> xsited: Cool :) 17:03:06 <Madhu> edwarnicke: u there > 17:03:07 <Madhu> ? 18:42:42 <edwarnicke> Madhu: I am 18:42:46 <edwarnicke> Madhu: How can I help you? 18:46:33 <Madhu> edwarnicke: its okay :) i was looking to talk to you before commenting on a Gerrit. 18:46:47 <edwarnicke> Madhu: Which Gerrit? 18:46:48 <Madhu> but i did it anyways .. since I couldn't talk to u in the morning 18:46:54 <edwarnicke> No worries... 18:46:57 <edwarnicke> Looking for the gerrit 18:47:48 <Madhu> its the brand fragment stuff 18:47:48 <edwarnicke> Was it the removing the branding fragment one? 18:47:53 <Madhu> yep 18:47:54 <edwarnicke> Yeah... glad you were reviewing :) 18:48:12 <Madhu> lol. i wrote them in my previous life 18:48:20 <edwarnicke> I *did* try to make sure that I didn't zap stuff being built, or that looked like it could be currently important 18:48:22 <edwarnicke> Madhu: I know :) 18:48:35 <Madhu> it is not important for sure :) 18:48:35 <edwarnicke> Madhu: But I also know my vision on these things is not perfect 18:48:47 <Madhu> but we never know how folks use the product 18:49:04 <Madhu> i know for a fact that the brand fragment is being used in one of the commercial ODL products 18:49:07 <Madhu> so ... u know :) 18:49:16 <edwarnicke> Madhu: Dead code is a problem... so my initial instinct is to delete dead code (it can always be gotten back from git)... but if you have a reason to retain it I'm fine 18:49:31 <Madhu> edwarnicke: very thin reason... 18:49:39 <Madhu> don't have a strong feeling either way. 18:49:44 <edwarnicke> Madhu: LOL... yeah... but not interested in arguing about it 18:49:50 <Madhu> perfect :) 18:49:53 <Madhu> me 2 . 18:49:53 <edwarnicke> Madhu:Those code cleanups were intended to be slam dunks 18:49:59 <edwarnicke> Madhu: If they are not... they are not :) 18:50:06 <Madhu> :) 18:50:07 <Madhu> btw 18:50:13 <Madhu> i had another question for you 18:50:18 <edwarnicke> Madhu: Ask :) 18:50:25 <Madhu> with regards to Inventory listener events 18:50:54 <Madhu> i managed to make it work... but wanted to make sure I am in the right track or not 18:51:07 <Madhu> so...my requirement is to listen to Inventory update. 18:51:13 <Madhu> from MD-SAL 18:51:35 <Madhu> edwarnicke: so i used OpendaylightInventoryListener 18:51:36 <Madhu> to my surprise, not many use this. 18:51:44 <Madhu> so i was wondering... if am doing it right or not :) 18:52:03 <Madhu> it does work. but just curious to know if the world has moved away to use more cooler listeners ? 18:53:51 <edwarnicke> Madhu: Ah... so here's why 18:53:56 <edwarnicke> You have two choices 18:54:01 <edwarnicke> 1) Listen to Notifications 18:54:06 <edwarnicke> 2) Listen with a DataChange 18:54:13 <edwarnicke> #1 has some risks to it 18:54:17 <edwarnicke> For example 18:54:26 <edwarnicke> #1 does *not* mean there is inventory in the datatree yet 18:54:34 <edwarnicke> You could race condition with the InventoryManager 18:54:36 <edwarnicke> For example 18:54:47 <edwarnicke> A DataChangeListener lets you know it *is* in inventory 18:54:50 <edwarnicke> Does that make sense? 18:55:19 <Madhu> yes. 18:55:29 <Madhu> so those who are using it are legacy ? 18:55:41 <edwarnicke> (I have a bunch of personal nits about how I wish this was all done... I don't *like* Notifications for communicating that state has updated... but not everyone agrees with my tastes) 18:55:48 <edwarnicke> Legacy would be a bit strong 18:55:52 <edwarnicke> InventoryManager uses it 18:55:54 <Madhu> :) 18:56:11 <edwarnicke> (though I honestly wish that the OFplugin just wrote the state data directly) 18:56:11 <Madhu> oka 18:56:26 <Madhu> lemme poke around with data change listener 18:56:28 <edwarnicke> Basically: I would recommend against it personally except in very limited circumstances 18:56:36 <edwarnicke> Not everyone agrees with me (nor do they need to) 18:57:07 * edwarnicke tries to keep clarity abotu the distinction between 'You are doing something differently than I would do it' and 'You are doing it wrong' 18:57:50 <edwarnicke> Listening for the Notification is *not* wrong. It does produce some additional opportunities to screw things up... you *can* manage those manually if you want. 18:58:34 <edwarnicke> (and there are probably times and places where it makes sense to pay the manual management tax in exchange for other benefits... I just can't think of any ;) ) 19:14:11 <Madhu> understood... having a working option in the pocket is always preferred :) 19:14:38 <Madhu> but will try poking around with dataChange Listener. I see most of the apps using that 23:15:19 <xsited> ed 14:25:12 <edwarnicke> Good morning :) 14:25:25 <edwarnicke> 5 minutes till Karaf Happy Hour :) 14:25:53 <xsited> morning, why waste time … 14:27:13 <xsited> for the integration procedure … i believe i need something in my root pom to get artifacts to show up in integration area? or how does that work? 14:27:38 <edwarnicke> Its not in your root pom... directions are here: https://wiki.opendaylight.org/view/Karaf:Step_by_Step_Guide#Prepare_your_feature_for_addition_to_the_integration_features_index 14:28:26 <xsited> to that end i believe I had a bad group id according to another wiki page, so i changed that and broke something in jenkins 14:29:35 <edwarnicke> Link ? 14:30:02 <xsited> F68FAA0F-69F2-4BFF-BD1A-337B7451BE8C-59036-00039ED7D7238ED3https://jenkins.opendaylight.org/packetcable/job/packetcable-nightly/44/consoleText 14:32:47 <xsited> i did the first few steps of that integration procedure and mvn clean install failed to pull bundles from nexus 14:35:10 <edwarnicke> You think this is the source of your troubles: https://git.opendaylight.org/gerrit/#/c/10564/1/features-packetcable/src/main/resources/features.xml 14:35:11 <edwarnicke> ? 14:35:46 <JYOTI_13> I am working with karaf integration with plugin2oc module.When I am trying to install feature within karaf container it is giving me following error: Invalid manifest header Import-Package: "com.sun.jdmk.comm;resolution:="optional"" : Cannot import a package more than once "com.sun.jdmk.comm" 14:36:16 <edwarnicke> xsited: In Jenkins you have no packetcable-merge job 14:36:29 <edwarnicke> #link https://wiki.opendaylight.org/view/GettingStarted:Project_Main 14:36:39 <edwarnicke> #link https://wiki.opendaylight.org/view/GettingStarted:Jenkins_Merge_Job_Setup 14:36:49 <edwarnicke> Without one,your artifacts are never published to nexus 14:38:08 <edwarnicke> JYOTI_13: Did the error tell you anything about which bundle it was complaining about? 14:38:20 <edwarnicke> Or could you pastebin the full error from the build? 14:39:00 <JYOTI_13> Right now i dont have full error message.. 14:39:13 <edwarnicke> JYOTI_13: Can you get it? 14:39:42 <xsited> working it … 14:39:50 <JYOTI_13> Not right now.. 14:39:53 <JYOTI_13> sorry 14:39:57 <xsited> #info thomas for pcmm 14:40:01 <xsited> btw 14:40:03 <edwarnicke> LOL 14:40:06 <edwarnicke> I should start the meeting 14:40:11 <odl_meetbot> edwarnicke: Error: Can't start another meeting, one is in progress. Use #endmeeting first. 14:40:17 <edwarnicke> #topic Roll Call 14:40:19 <edwarnicke> #info edwarnicke 14:40:28 <xsited> irc://irc.freenode.net:6667/#info thomas for pcmm 14:40:34 <edwarnicke> JYOTI_13: So... I can make a few suggestions without the full error, but its going to be hard 14:40:50 <xsited> #info thomas for pcmm 14:40:58 <edwarnicke> JYOTI_13: Basically, it looks like the manifest header of one of your bundles has repetitions Import-Package info 14:41:07 <edwarnicke> I can't be sure which one wihtout the full error 14:41:29 <edwarnicke> And the next step from figuring out which one would be to look at the MANIFEST.MF file for it and see what you can figure out 14:41:53 <edwarnicke> JYOTI_13: My best guess is that you are manually using <Import-Package> in your pom.xml and accidently listed something twice 14:42:00 <edwarnicke> JYOTI_13: Does any of that make sense? 14:43:32 <JYOTI_13> there is one log4j manifest file which where package has been used twice.Is it possible to edit manifest file manually?? 14:43:56 <edwarnicke> JYOTI_13: It depends on how your MANIFEST.MF is being managed 14:44:02 <JYOTI_13> ok 14:44:04 <edwarnicke> If you are managing it manually, sure :) 14:44:11 <JYOTI_13> ok 14:44:12 <edwarnicke> If you are managing it with maven-bundle-plugin 14:44:13 <edwarnicke> No 14:44:24 <edwarnicke> It kind of depends what you are doing... 14:44:41 <edwarnicke> xsited: I think you need the following two steps: 14:44:44 <edwarnicke> 1) Set up a merge job 14:45:03 <edwarnicke> 2) In your root pom.xml, change <groupId>org.opendaylight.controller.packetcable</groupId> to <groupId>org.opendaylight.packetcable</groupId> 14:45:46 <edwarnicke> xsited: Because only the controller project can write into the org.opendaylight.controller subsection of the groupId tree (just as only packet cable can publish artifacts for org.opendaylight.packetcable ) 14:46:09 <edwarnicke> xsited: Does that make sense? 14:46:19 <edwarnicke> alagalah_: Back from the happiest breakfast place on earth? ;) 14:46:55 <alagalah_> edwarnicke: Yeppas 14:47:21 <alagalah_> edwarnicke: Back to circuit breaker failure and 6' freezer defrosted all our food :( 14:47:38 <edwarnicke> alagalah_: That sounds like unhappy breakfast 14:48:44 <JYOTI_13> is it mandatory to include <configfile> tag within feature.xml?Kindly let me know its significance. 14:49:01 <edwarnicke> JYOTI_13: Not mandatory at all... it depends on whether you need a configfile or not 14:49:17 <JYOTI_13> ok 14:49:28 <edwarnicke> JYOTI_13: If you are using .cfg files for ConfigAdmin or using the config subsystem you will likely need one, otherwise not 14:56:26 <edwarnicke> JYOTI_13: Need anything else? 14:57:37 <JYOTI_13> I am not managing MANIFEST.MF file manually..managing it with maven-bundle-plugin..so how to resolve the issue now? 15:02:02 <xsited> ed: merge job created and groupid already in 15:18:54 <JYOTI_13> will again check maven-bundle-plugin within pom.xml to identify what is there inside import package tag.. 15:52:05 <tbachman> join opendaylight-opflex 15:52:16 <tbachman> har 14:25:42 <edwarnicke> Good morning :) 14:25:46 <edwarnicke> 5 Minutes to Karaf Happy Hour :) 14:31:53 <odl_meetbot> edwarnicke: Error: Can't start another meeting, one is in progress. Use #endmeeting first. 14:32:00 <edwarnicke> #endmeeting