13:28:32 #startmeeting jquery_xpath_for_md_sal_intern 13:28:32 Meeting started Fri Aug 22 13:28:32 2014 UTC. The chair is devinavery. Information about MeetBot at http://ci.openstack.org/meetbot.html. 13:28:32 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 13:28:32 The meeting name has been set to 'jquery_xpath_for_md_sal_intern' 13:28:41 #chair regXboi Neel_B 13:28:41 Current chairs: Neel_B devinavery regXboi 13:28:57 morning - this is the last one, so let's button this up 13:29:05 #info Neel_B is having a bunch of compile errors when trying to get code running in controller. 13:29:18 do we have a paste bin? 13:29:25 #info devinavery asked Neel_B to push a draft so we can pull down to help debug 13:29:37 http://pastebin.com/jF1LeqtA 13:29:47 https://git.opendaylight.org/gerrit/10170 13:29:55 it went to drafts i guess 13:30:10 Perfect. Neel_B can you add me and Ryan as reviewers. 13:30:30 your emails right ? 13:30:30 You can find me under "Devin Avery" 13:30:33 that will be necessary for us to see the patch 13:30:40 That should give us permissions to see the draft 13:31:00 and I'm rmoats@us.ibm.com 13:31:09 regXboi: my guess on the error is that there is a scope issue with the dependencies, most likely because the pom.xml was intended to be a test-it bundle only. 13:31:25 it sure looks like it from reading 13:31:40 but there are some other possibilities 13:32:06 done 13:32:52 Ok. Pulling in 13:34:10 ditto and starting a build to watch it go boom 13:35:36 well... I think I want to change the snapshot version 13:37:03 ok, it looks to me like we are missing at least one dependency 13:37:07 and probably more 13:39:34 have we added the dependency on yang-data-api? 13:39:45 That is the one tha tis missing 13:39:46 13:39:46 org.opendaylight.yangtools 13:39:47 yang-data-api 13:39:49 13:40:47 just added 13:41:03 and that will fix the compile and get us to the tests 13:41:21 one other things - let's add the following to the pom.xml: 13:41:41 after the description: 13:41:51 0.1-SNAPSHOT 13:41:55 compiled and tunning tests 13:42:04 willbe failing 13:42:20 Also - the artifactId should match the folder name. In this case we have sal-xpath so that should be the artifact id 13:42:42 yes 13:43:08 that also will help the tests 13:43:09 (some) 13:43:52 ok... now the problem with the test is that bundle we just built isn't being included 13:44:47 i guess artifact id is correct ? 13:45:05 artifact id of the parent is sal-parent 13:45:16 but artifact id of the bundle should be sal-xpath, not sample-path 13:45:39 changed sal-xpath 13:47:22 should we change some parts of test code too ? 13:49:01 well... it looks like we may have to break the tests into a separate bundle 13:49:06 which I really don't want to do 13:49:32 okay but the bundle runs like an it 13:49:37 i guess 13:50:49 ok 13:50:56 yes I think we can make this work 13:51:03 can one of you try the following 13:51:28 Tin 13:51:37 in toastertest.java 13:51:42 lines 133 and 134 13:51:49 uncomment and replace with the following: 13:51:54 mavenBundle( "org.opendaylight.controller", "sal-xpath" ).versionAsInProject(), 13:52:08 that should pull in our bundle for the test and then the test should run 13:52:11 Was just looking to see if we included our project. :) 13:52:26 I'm getting a versioning error on my setup - but that was what I've been getting for a while now 13:52:44 I am getting a version error now 13:52:53 Missing a dependency for sal-xpath in your maven project 13:52:54 ... 13:53:23 did you set the version in the pom.xml directly? 13:53:24 4wait.. it is sample-xpath in my environment still.. one sec 13:53:39 I changed it to 0.1-SNAPSHOT in mine 13:53:48 and so I'm failing on 13:53:49 java.lang.RuntimeException: Could not resolve version. Do you have a dependency for org.opendaylight.controller.samples/sample-toaster-provider in your maven project? 13:53:52 The problem was that I did not change my artifact id 13:53:56 :) 13:54:01 that can be painful 13:54:07 When I used "sample-xpath" for the dependency it ran successfully 13:54:33 So Neel_B: change the artifact ID in the pom.xml, then add regXboi's line above into the ToasterTest 13:54:38 ok, then I think we make these changes, and we can change this from draft to visible and I am happy 13:54:52 that should get it so the test runs failing on a verification now instead of a compilation / runtime error 13:54:53 but the tests are failing right ? 13:55:40 Neel_B: I would comment out the verification looking for ChildToasterList - right now the OpendayLightToaster that is in source ( that we use) doesn't publish that. So commenting out those path validations will allow the test to pass 13:55:48 Just put a TODO in there for anyone who comes in the future 13:57:00 where to put TODO ? 13:57:38 org.sonatype.aether.transfer.ArtifactNotFoundException: Could not find artifact org.opendaylight.controller:sal-xpath:jar:0.1-SNAPSHOT in central (http://repo1.maven.org/maven2/) 13:58:16 first, the todo goes where the verification for child toaster list is commented out 13:58:29 Neel_B: have you recompiled the toaster project after modifying the pom and ToasterTest? ToDo go where regXboi said 13:58:31 second, how are you running the code to get that error? 13:58:53 cd sal-xpath 13:58:58 mvn clean install 14:00:40 that error says you aren't looking at your local install 14:06:38 devinavery: unfortunately I'm blocked on this sample-toaster-provider version bug that my install seems to have 14:06:47 can you see if you can make the tests work? 14:07:58 i am trying 14:08:25 regXboi: Ok I saw your change worked just fine. 14:08:54 Neel_B: if you are stuck, please upload your code again (push another draft) so I can pull it down again to see if the problem shows up 14:09:47 just a sec i am tryin to clear out 14:09:59 let me try again and will push it 14:10:14 devinavery: I'm at neel's point now - I expect that the options need to be looked at 14:13:30 strange that I am not seeing that error... 14:13:50 I will pull Neel's code when it is pushed and retry 14:14:50 I think it may have to do with aether version and or configuration 14:15:09 https://git.opendaylight.org/gerrit/10172 14:16:11 um 14:16:18 you shouldn't have pushed it as a new patch 14:16:26 you should have pushed it as a revision of the old patch 14:17:35 okay 14:17:37 added 14:17:40 you both 14:18:36 well first of all 14:18:49 Neel - did you follow regXboi's comment about pushing a new patch to the same draft? 14:19:28 how to do that 14:20:02 I believe we have to comment out the child toaster tests as well as that isn't part of the standard toaster eithre 14:20:30 when you push the patch, you have to change the commit message to include the patch identifier of the original patch 14:20:31 You edit the previous commit using " commit --amend" --- that will basically edit the previous commit you pushed 14:20:43 devinavery has the easy way :) 14:21:34 okay i thought i pushed it so it won't work 14:21:50 you did 14:21:51 from now i will do that 14:22:03 oh n/m my comment :) 14:22:06 wrong window 14:22:17 The problem Neel is that your changes are now spread out over two gerrits 14:22:51 So what we should do is collapse the patches back together and just push a second patch to the first git 14:23:01 yes 14:23:12 devinavery: can you walk through that? I have to get on a phone call 14:23:27 regXboi... I can try. I am relatively new to git. :) 14:23:31 ugh 14:23:33 But let me see if I can remember how to do that 14:23:45 ( I think I can dig up my notes) 14:23:52 ok 14:24:00 I'll be back when I'm done 14:24:06 ok 14:24:31 git rest --soft HEAD~2 14:24:36 reset* 14:24:48 is it that ? 14:24:59 Ahh... almost... try this 14:25:15 FIRST - copy your code to another folder outside of the control r- don't want to loose it. ;0 14:25:46 Next , run "git log". This should print out the history of all commits in your build. Your last two commits should be at the top - correct? 14:25:50 actually did it already 14:25:55 Oh.. alright. 14:26:27 last commits weren't mine 14:26:28 now 14:26:50 so what do i do 14:27:03 will those gerrits be gone 14:27:03 ? 14:27:05 Ok, so I think you went to far back 14:27:11 No - the gerrits are still there. 14:27:19 if you do a git status, what shows up? 14:30:59 devinavery: the whole sal-xpth in git status 14:31:38 Ok. Do another git commit. 14:31:58 But this time, in your commit message, make sure that you have the change ID for the first git 14:34:03 okay 14:34:05 then 14:34:21 Ok. SO what does your commit message look like? 14:34:29 how to remove the second gerrit ? 14:34:44 You can just discard it from the gerrit UI in a little bit 14:34:56 [DO NOT MERGE] Initial code prototype for XPath query capabilities on MD-SAL Change-Id: Ica3e660c14b834e42e5cfe5cb9ca17b01e3aa8f9 Signed-off-by: Neel 14:35:01 For now lets just push the new patch to the exist gerrit and see if it works 14:35:03 Ok that looks good 14:35:17 go ahead and push it again. It should show up as a new patch on the existing gerrit 14:36:31 new gerrit 14:37:27 Can you post link? 14:37:39 If it has everything then we will just go with that one and discard the other two 14:38:09 https://git.opendaylight.org/gerrit/#/c/10173/ 14:38:38 Ok. Please add me to the reviewers. 14:39:09 When you modify changes again use "git commit --amend -s" that should add the change ID and your sign off into the same commit 14:39:42 added 14:39:45 okay 14:39:52 pulling in the new code now 14:39:54 removing is delete draft change ? 14:40:16 There should be an option to "abandon" the draft 14:41:38 Ok. I built your new code and it compiled and the tests ran with no problem 14:43:41 deleted other two 14:44:15 i was getting errors :\ 14:44:56 short call - I'm back 14:45:02 I see there is a new commit 14:45:03 cool 14:45:19 Ok - regXboi: we ended up generating a third commit - just using that one now 14:45:26 Neel_B has removed the other two 14:45:34 I was able to pull in the new code compile and run tests with no problem 14:45:57 Neel_B: when you run "mvn clean install -DskipTests" my first question is does that run successfully? 14:46:08 catching up now 14:46:18 if so, then lease run mvn clean install and paste the output and failure to a paste bin for us to review 14:46:28 I need to step away for 2 minutes 14:47:00 http://pastebin.com/xZGLfdzM 14:47:22 so that is the pastebin from sal-xpath 14:47:44 can you go back to controller and run "mvn clean install -DskipTests" and does it pass? 14:48:54 no need to change sal-parent's pom to add sal-xpath 14:48:56 ? 14:49:52 running 14:50:10 back 14:51:17 Neel_B: eventually we will want to change sal-parent but for right now you do not need to change it to get this test running 14:52:28 devinavery 14:52:40 I'm pretty sure the problem with the tests is in the paxexam configuration 14:53:07 I'm trying to find where/how I can make sure paxexam is looking at the local maven repo 14:53:53 I would have guessed (and perhaps I am wrong here) that it uses your local repo as a proxy / cache 14:53:59 So I would have thought it only looks there 14:54:16 Probably worth checking first that the artifact is actually installed there... 14:54:50 Neel_B: can you check your local repository (~/.m2/repository/org/opendaylight/controller/ etc and make sure you can find your newly compiled bundle? 14:55:08 devinavery: I can and have and I get the same error 14:56:00 Ok. Let me recompile my entire controller... just incase I have something bogus in my cache 14:56:08 I'm doing that now 14:56:37 I think the question is do you have a settings.xml that is coming into play? 14:57:21 if you run "mvn help:effective-settings" that will print out what your settings are 14:57:51 now *that* is interesting 14:58:13 This is what my output is: basically points to my local repo 14:58:14 http://pastebin.com/86Kd8Vb9 14:58:54 aha 14:58:57 ok I have it working now 14:59:06 Oh. Great. What was the trick? 14:59:22 so we may need to rethink the pom 14:59:34 what I found was that I *didn't* have the jar file in the local m2 14:59:40 because the tests weren't passing 14:59:51 once I ran with skipTests and got the jar in 14:59:57 then all is better 15:00:08 Oh.. chicken and the egg?? :) let me try that then 15:00:14 and we lost Neel 15:00:40 and I get a comparison Failure (but I think that's on the child) 15:00:50 Bingo 15:00:56 I reproduced the error 15:01:03 :) 15:01:10 So I guess Pax-Exam is not smart enough to take the jar from the current bundle... 15:01:18 That seems odd... 15:01:52 ok... so I'm commenting out a few tests 15:01:59 to get back to the original toaster model 15:02:02 and... 15:02:20 oh yeah, I had to hardcode dependencies as well :( 15:02:34 and all is good 15:02:48 so I can say now that we can make this available 15:02:53 with some caveats 15:03:51 and I think we can link it into sal-parent as well 15:03:57 and add that to the next version of the patch 15:04:26 Ok. So wait.. did you just comment out the tests in the toaster so that it would basically skip the pax-exam test? 15:04:40 no, I commented out the tests on the child 15:05:09 because those aren't part of the standard toaster model 15:05:33 in the version of ToasterTest.java I'm looking at I'm talking lines 223-227 15:06:02 and that's because I've not caught up to the new patch 15:06:13 Neel did that already for me in the latest patch 15:06:54 ahh. Ok. Understood 15:07:21 So we still have the problem where if someone is building this bundle for the first time with "mvn clean install" it will fail... right? 15:08:00 hey devinavery 15:08:10 regXboi my net got disconnected 15:08:20 we notices 15:08:23 er noticed 15:08:31 but I'm pretty sure we've run everything down 15:08:34 devinavery: yes 15:08:44 I'd like to verify that the new patch compiles 15:08:50 and then make that as a note and carry on 15:09:29 I'm getting a compile error on the new branch 15:10:29 is anybody else seeing this? 15:11:04 On the code from 10173? 15:11:08 yes 15:11:10 I am not seeing any compile errors there 15:11:23 odd... I'm getting a compile error on line 121 of ToasterTest.java 15:11:39 about not being able to find a symbol, but no symbol name 15:13:40 you know what 15:13:42 let's move on 15:13:53 because it's now working for me 15:14:05 so I'm going to assume I was doing something stupid in the file system 15:14:18 regXboi: running -DskipTests 15:14:26 on the main controller? 15:14:30 yeah 15:14:35 ok 15:14:57 already halfway 15:15:32 so... I think you can solve your problem in sal-xpath by doing skipTests there too 15:15:42 and then when you run it again, without skipTests it will work 15:16:03 at least that's how I solved the same error code on my system 15:16:04 isn't it an issue for someone new using it ? 15:16:19 regXboi: this issue must be why the integration tests were split out into different bundles. Pax-exam can load the source code from the same bundle it is running in 15:16:26 may be because we are calling the bundle inside itselp 15:16:38 devinavery: that is very likely true 15:16:42 itself* 15:16:54 devinavery: *but* I'm not sure we have time to fix that anymore 15:17:03 as the project period is ending today 15:17:18 I mean, folks can work on it going forward of course :) 15:17:23 regXboi: agree - just means we need to put a note in there and that we can enable this project to compile by default (since it won't) 15:17:41 i will finish it 15:17:48 regXboi for sure.. :) 15:18:01 well, I think the right idea is to split out the test into sal-xpath-it 15:22:18 and yes, I've already done it :) 15:22:25 so I can tell you how 15:22:51 chapter and verse: 15:23:04 1) in md-sal, mkdir sal-xpath-it 15:23:15 regXboi: 15:23:19 passed 15:23:28 cool! 2) in md-sal, cp -r sal-xpath/* sal-xpath-it 15:23:38 3) in sal-xpath, rm -rf src/test 15:23:51 4) in sal-xpath, edit pom.xml and remove the pax exam plugin 15:23:58 5) in sal-xpath-it, rm -rf src/java 15:24:11 okay 15:24:13 6) in sal-xpath-it, edit pom.xml and add dependency to sal-xpath 15:24:18 let me try those things 15:24:23 and then you can build sal-xpath without failing 15:24:30 and sal-xpath-it will run the tests 15:24:58 and 7) edit md-sal/pom.xml and add sal-xpath and sal-xpath-it in it's own block under xsql (testing that one now) 15:25:25 Hey Guys. I need to step away for a bit. Neel_B I am happy to bring your code in and compile it once you have the -it test broken out etc 15:26:09 Just up load a new patch (this time you can upload it public by pushing to HEAD:refs/for/master ) and I will check it out 15:28:27 regXboi: are there any dependencies for the sal-xpath 15:28:28 ? 15:28:37 or i should remove all ? 15:28:39 regXboir and Neel_B - you guys should both be chairs on the meet bot. Please update it with final comments before 15:28:50 no... leave the dependencies alone 15:28:50 okay sure 15:28:56 just remove the pax-exam plugin 15:29:13 * devinavery is walking away for a while 15:33:21 okay xpath working fine 15:33:27 changed xpath-it tooo 15:33:41 try that one now 15:34:07 we have to change name in pom too ? 15:34:14 for xpath-it ? 15:34:42 remove packaging bundle? 15:35:27 regXboi: ? 15:35:42 give me a sec 15:36:01 you can try it 15:36:03 I didn't 15:36:06 but I probably could have 15:38:39 it should work 15:40:22 regXboi: same error as you got 15:40:23 :\ 15:40:30 which error is that? 15:40:30 line 121 15:40:59 did you add the dependency to sal-xpath in the pom.xml? 15:41:32 yeah i did 15:41:32 add 15:42:00 so, I found the problem was that I switched windows and was in a stale file system 15:42:27 so I changed up a directory and back down and got to the correct i-node 15:43:08 lemme try that 15:47:11 i guess its working 15:47:15 running tests 15:47:26 it should run two of them 15:49:29 1 test isfailing 15:49:32 lemme check 15:50:58 okay it is taking my toaster 15:51:07 so it should fail 15:51:18 if it takes the old taoster it passes 15:51:32 regXboi: how to get the old taoster 15:51:54 how is it taken your toaster? 15:53:15 i haven't changed anything in toaster 16:08:39 ah ok 16:12:27 so now you should be able to change md-sal/pom.xml 16:12:33 and have it all build 16:14:18 regXboi: 16:14:24 i am not able to ammend ? 16:14:33 it is creating new gerrit 16:15:10 but when i changed the commit -id to the same 10173 it says rejected 16:15:16 i mean changeid 16:15:49 when you did commit did you do commit --apend as devinavery suggested? 16:16:46 er I mean --ament 16:16:50 ACK 16:16:57 er I mean git commit --amend 16:19:30 Neel_B, Neel_B_ did you get that? 16:19:40 i am trying 16:19:48 will ping you 16:20:00 first thing 16:20:01 so when I do a git commit --amend, I find a bunch of files need to be added removed 16:20:10 yes? 16:21:03 okay lets keep that aside for a second 16:21:10 what should i add in md-sal pom 16:21:31 sal-xpath ? 16:21:34 right ? 16:21:36 no 16:21:53 down on line 81 (after sal-dom-xsql and the blank line 16:21:56 I'd add the following: 16:22:04 sal-xpath sal-xpath-it 16:22:07 that's four lines 16:22:10 the comment 16:22:14 two module lines 16:22:16 and a blank line 16:22:41 and then you are hooked into the whole build process 16:25:04 [ERROR] Failed to execute goal org.apache.maven.plugins:maven-checkstyle-plugin:2.12:check (default) on project sal-parent: Failed during checkstyle execution: There are 1 checkstyle errors. -> [Help 1] 16:25:19 did you use tab? 16:25:38 does a line end with whitespace? 16:27:18 retrying 16:28:46 that error gone let it compile 16:28:55 ok, it may fail 16:28:57 depending 16:29:11 sometimes I need to add -DskipTests for md-sal to build 16:29:30 but... so long as the refactor shows you in the list, you are good 16:33:11 regXboi: can you sum up all at once ? 16:33:19 mean while its compiling 16:33:31 so... if md-sal succeeds, great! 16:33:49 but if md-sal fails, then you'll need to go back and look at the error 16:33:58 it'snot going to on my system 16:34:09 because it takes the modified toaster 16:34:15 and it returns error 16:34:24 if run on normal toaster it passes 16:34:33 use the normal toaster 16:34:36 don't change that 16:35:25 i already changed a lot for the testting right 16:35:27 : 16:36:04 ok, let's back up 16:36:14 when you say the modified toaster, what files are you talking about? 16:41:12 opendaylight toaster.java 16:41:17 i will change it back 16:41:25 yes, that needs to be left alone 16:42:35 now, on my system without -DskipTests, building under md-sal fails at sal-rest-connector 16:43:10 but if I look at the reactor summary, I see that sal-xpath and sal-xpath-it are SKIPPED, which is good, that means they are part of the build 16:43:11 still running 16:43:21 so then if I build with -DskipTests they pass 16:58:24 regXboi: devinavery 16:58:41 both sal-xpath and sal-xpath -itare passing 16:58:49