14:36:38 #startmeeting #opnfnv-fuel 14:36:38 Meeting started Tue May 31 14:36:38 2016 UTC. The chair is Greg_E_. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:36:38 Useful Commands: #action #agreed #help #info #idea #link #topic. 14:36:38 The meeting name has been set to '_opnfnv_fuel' 14:37:16 mskalski: thanks for the reviews! 14:38:35 hi guys 14:38:39 roll call 14:39:03 #info Stefan Berg 14:39:10 #info Josep Puigdemont 14:39:14 hi 14:39:17 #info Peter Barabas 14:39:34 #info Billy O'Mahony 14:39:51 #info Jack Morgan 14:39:59 #info Michal Skalski 14:40:19 Peter how goes the port 14:40:47 Greg_E_: we can make a deploy on 'real' hardware 14:41:01 have we run FuncTest 14:41:04 s_berg: has your commits been merged? 14:41:31 z0d: I was waiting for this meeting to announce that I will merge https://gerrit.opnfv.org/gerrit/#/c/13919/ afterwards. 14:41:37 I'm testing a nested deployment. right now we're troubleshooting with DanSmithEric 14:41:40 what scenario can we deploy? 14:41:43 no-sdn? 14:41:46 This will break things, obviously, but that's a necessary stepping stone. 14:41:56 what plugins are ready? 14:42:04 try without plugins first 14:42:32 Greg_E_: The usual procedure we've followed before is to disable all plugins, merge the build system and deploy changes, then let the plugin "re-certify" themselves. 14:42:50 ok 14:43:08 That is, enable the plugin and verify that it works, then commit. 14:43:45 there is a change for odl to be build for mitaka https://gerrit.opnfv.org/gerrit/#/c/14385/, but also scenarios needs to be updated 14:43:52 So with the build system fix we, obviously, can build. The cache is disabled for now (meaning that deployment requires Internet connection). 14:44:28 My suggestion is that I merge, z0d and all of us make sure that the deployer works both on real iron and virtually. 14:44:28 #info Daniel Smith 14:44:41 Next we re-enable the cache and update scenarios. 14:44:43 peter are you at the point where any testing can be done on the deployment? 14:45:06 or do we need scenario update before functest can run 14:45:39 I'm not too familiar with FuncTest 14:46:10 the CI for both Baremetal and NEsted are back online now 14:46:18 so as soon as Sberg does his commit 14:46:34 the nodhs-ha-scneario (and otehr jobs in the batcH) will run 14:46:47 i believe that functest and ys, etc are part fo that.. Faith - can you remind me? 14:47:02 And fail I assume, until we update the "Fuel common" part of the scenarios, right? 14:47:07 fdegir that is :) 14:47:23 (I.e. reap a Fuel 9 and paste accordingly into the dea fragments.) :) 14:47:24 correct.. i would imagine that we start the integration uplift to match the testing suite now as well 14:47:27 correct 14:47:45 im working with Peter now on your latest iso bu its being petulant 14:47:58 so i would really like that merged patch and a nice new ISO :) 14:48:12 since what we are doing will be obsolete as soon asyou lick submit :p 14:48:14 click* 14:48:16 As long as we're all in agreement I'm happy to pull the merge trigger now. 14:48:26 i say ja! 14:48:40 +1 14:49:00 +1 14:49:07 s_berg: you already has 10 points in review ;) 14:49:18 s_berg goes for the maximum points :-P 14:49:23 mskalski: :) 14:49:32 Eurovision Song Contest style. 14:49:36 :D 14:49:38 :-D 14:49:41 Merged! 14:50:07 HAHHA 14:50:10 Look at all that smoke! ;) 14:50:29 <-: 14:50:37 looking now it seems that functest is running on a CI no feature 14:50:44 and that is of master (before the merge just now of course) 14:50:45 https://build.opnfv.org/ci/ 14:51:00 z0d, sending you a fresh dea.yaml I've just reaped from my virtual setup for reference. 14:51:05 so i would imagine that it will get picked up in the next build cycles and we will get a nice report of everything that is dead 14:51:13 Let's use the ISO being built off of this merge for testing. 14:51:21 s_berg: thanks 14:51:27 ack 14:52:27 peter can you push a copy to the blade we are testing on (should i push one from y build server) or grab it from jenkins? 14:52:51 when do we expect FT to start passing? 14:53:08 the .iso? sure, I grab it 14:53:09 Greg_E_ can you be more specific 14:53:16 when you say "passing" which part? 14:53:32 no sdn and basic ha should pass right away (maybe some things missing) for basic smoke tests 14:53:50 but any other scenario you would expect to fail 14:53:51 Greg_E_: Speaking from past experience it will take a while for us to stabilize here. 14:54:18 ^^ 14:54:19 release meeting is in 10 minutes 14:54:28 well.. for status 14:54:32 as usual they are looking for dates from us 14:54:36 we have just pushed our uplift (merge) to mitaka 14:54:40 and have starting integratino testing phase 14:54:46 so that plugins can start their uplift 14:54:56 I'd like to see us focus on verifying deploy on virtual and non-virtual basic deploy functionality this week. 14:54:57 espect 1-2 weeks iterations for stability across the board 14:55:06 and then development can start on newer stuf 14:55:16 ask always +1 to Herr Berg 14:55:19 I agree with that. 14:55:20 so mid june 14:55:24 (the 1-2 weeks that is) 14:56:30 If we something bites us it will take a while longer. It's as always more about resources than rocket science. 14:57:21 if we need help in any area let me and dklenov know 14:57:49 we have additional resources from Mira side we can bring 14:58:29 ZoD Any updated on the scenario handling? 14:58:37 That's great. Stay tuned to this IRC to assess the height of the flames during the week. 14:59:21 ^^z0d 15:00:09 jonasbj: not yet, sorry. I was busy troubleshooting 15:01:18 z0d: But you were able to auto deploy on bare metal from a reaped config? 15:03:09 Release meeting starting, so I am a pumkin, but you guys keep going I can close this meeting later 15:05:01 s_berg: I only tried nested, 15:05:20 s_berg: and it's still in progress, but it should work 15:05:22 z0d: Ah, I suspected I got something backwards. :) 15:05:26 what is nested? 15:05:46 joseppc: virutal nodes 15:05:49 virtual* 15:05:56 virtual Fuel, controllers, computes 15:06:04 aah, ok 15:06:06 a.k.a. libvirtd 15:06:28 thanks 15:06:58 does scenarios files will change now, or base on what we have now I can modify them to support newer version of the pluigin? 15:07:49 they won't change soon I think 15:08:05 ok 15:12:15 About the patches I sent earlier, do you have any more comments? 15:12:54 specially the comment z0d left on the first one: https://gerrit.opnfv.org/gerrit/#/c/14703/ 15:13:34 would that be acceptable? I kind of rely on the functionality implemented here on the other patches 15:14:00 if not, I'll have to find a workaround 15:19:46 i would suggest *unless its absolutely necessary) that for the no sdn HA scneario, ceilo be removed.. its such a hog - atkes up 45% of the deployment time 15:20:01 and we only run the systems for a 4 hour period - hardly long enough to get data. 15:20:05 (just a thought) 15:22:13 I’ll check with other projects to see what they are doing 15:35:51 this will be where you want to look once this build is done and tests run 15:35:52 http://testresults.opnfv.org/reporting/functest/release/master/index-status-fuel.html 15:35:58 Doctor needs ceilometer 15:36:15 but they also need congress and other bits we don’t support 15:36:21 well thats fine 15:36:28 but that is in a doctor scenario 15:36:35 do we need to smoke test it for everyone other one? 15:36:43 will sync up with them and see if they plan to install it themselves 15:36:54 no we don’t 15:36:55 this idea of matrix'ing everything with everything is going to kill our system 15:36:59 ok 15:37:50 #info C stable branch date Aug 15 15:43:58 build of that mege is done.. pulling iso to nsted setup now peter 15:44:04 merge* 15:44:09 nested* 15:48:25 #endmeeting