14:00:08 <[1]JonasB> #startmeeting BGS Fuel status call 14:00:08 Meeting started Tue May 26 14:00:08 2015 UTC. The chair is [1]JonasB. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:08 Useful Commands: #action #agreed #help #info #idea #link #topic. 14:00:08 The meeting name has been set to 'bgs_fuel_status_call' 14:00:15 <[1]JonasB> #info Jonas Bjurel 14:00:17 #info Szilard Cserey 14:00:25 #info Michal Skalski 14:00:28 <[1]JonasB> Hi Szilard 14:00:31 Hi Jonas 14:00:47 #info Jose Lausuch 14:02:59 <[1]JonasB> Lets wake Stefan up 14:03:29 My shirt pocket is making sounds! :) 14:03:34 #info Stefan Berg 14:03:46 <[1]JonasB> Always works:-) 14:03:55 <[1]JonasB> Lets get started 14:04:04 * stefan_berg has an IRC -> Jabber bridge :) 14:04:04 <[1]JonasB> #topic agenda bashing 14:04:21 <[1]JonasB> So what do we have for this call 14:04:44 <[1]JonasB> #info ODL script and testing status for Arno 14:05:11 <[1]JonasB> #info ODL Fuel plugin status 14:05:33 <[1]JonasB> #info Status on Autodeploy 14:05:41 <[1]JonasB> What more? 14:05:51 which Autodeploy ? 14:06:13 <[1]JonasB> Your one, when yo can get LF access etc. 14:06:32 which topic should we start 14:06:48 #info I'm actually working on the Reaping functionality 14:07:21 <[1]JonasB> #info testing status 14:07:29 <[1]JonasB> Anything more? 14:07:47 nope 14:07:55 <[1]JonasB> Lets start with testing status (ODL excluded) 14:07:56 A small question on Fuel 6.1, but that's AoB. 14:08:03 <[1]JonasB> #topic Testing status 14:08:18 <[1]JonasB> Jose - can you give some updates? 14:08:27 yes 14:08:44 maybe I'll copy the same as we wanna say in the TSC meeting 14:09:07 <[1]JonasB> Yep - a lot of copy and paste:-/ 14:09:38 #info 1) heavy refactoring on CI (to be consistent, wget test files not trustable) 14:09:47 #info 2) No full CI sequence successful so far neither on POD1 (Tempest + bench + vPing) nor POD2 (Tempest + bench + ODL) assuming that there were errors in Tempest and Bench 14:09:57 #info 3) Tests in progress on both PODs 14:10:26 #info 4) working with pbandzi to get ODL tests working 14:10:37 but maybe that is for the other topic :) 14:10:47 <[1]JonasB> #info Stefan has changed the quota setings, so Tempest should now pass or? 14:10:59 #info Changed, but haven't merged yet. 14:11:09 once I fix the problems of automation from Jenkins we will see 14:11:14 #link https://gerrit.opnfv.org/gerrit/#/c/646/ 14:11:18 <[1]JonasB> Any reason to not merge it? 14:11:27 there were a lot of changes in ours scripts to do things more consistent 14:11:42 maybe you can merge :) 14:12:01 Wanted to give trozet the courtesy of reviewing it should he want to. But sure, I'm merging... now! 14:12:22 <[1]JonasB> Ok 14:12:32 #info Quota changes in Fuel has been merged. 14:12:44 <[1]JonasB> Ok, great. 14:13:10 thanks! :) 14:13:15 <[1]JonasB> Jose, apart from ODL. Do you have any problems specific to fuel? 14:13:15 done 14:13:34 yes, but nothing that I could fix myself 14:13:37 well 14:13:38 actually 14:13:43 POD1 is meesed up 14:13:49 creating/deleting glance images doesnt work 14:13:54 I guess there is a problem with Ceph 14:14:00 glance api hangs 14:14:09 dont know how to troubleshoot 14:14:15 <[1]JonasB> Or an issue with ODL and the bridges? 14:14:19 That's a bit surprising as the deploy health checks does these kind of things. 14:14:19 logs shows no errors 14:14:39 maybe the tests screwed up something 14:14:47 so I would like to redeploy this evening 14:14:51 I know that lmcdasm was planning to do some ODL work there, any chance the storage network has been hit? 14:14:59 <[1]JonasB> Ask Daniel to help with trouble shooting 14:15:02 yes 14:15:06 strange though 14:15:22 the jenkins jobs also hangs because of glance 14:15:23 jose_lausuch: It has normally worked? 14:15:26 not good 14:15:33 On pod1 I mean. 14:15:39 on this deployment Im not sure, I havent tried 14:15:47 but before of course 14:15:50 always worked 14:15:53 OK. 14:15:59 but this weekedn I ddint try it 14:16:11 <[1]JonasB> I would suspect the ODL scripts 14:16:24 could be, but they dont do anything to the storage, but who knows 14:16:32 maybe the san network is affected 14:16:35 need to check with Daniel 14:16:58 <[1]JonasB> Exactly, if Daniel by mistake has hit the SAN bridges... 14:17:19 but anyway, I can perform other tests 14:17:27 but will need redeploy soon 14:17:40 <[1]JonasB> Ok Thanks Jose, can we move to ODL? 14:17:43 yes 14:17:47 pbandzi: ping 14:17:48 <[1]JonasB> #topic ODL 14:18:19 for ODL, I need to discuss something with you guys 14:18:37 <[1]JonasB> #info Daniel has commited a patch for the ODL scripts - not final though 14:18:44 <[1]JonasB> Yes shoot Jose 14:18:46 Daniel and I kind of agreed of having a script to create networks/ports and then do the "curl" command 14:18:57 but that is exactly what the robot test from pbandzi are doing 14:19:05 need clarification there 14:19:20 is not Tempest more appropriate for this ? 14:19:23 i am on tsc nbow on behalf of frank 14:19:36 <[1]JonasB> Cant we use pbanzis robot suite then? 14:19:41 I don't understand why use Robot in these types of tests 14:20:30 I would say so, but we need to make them work and/or combine them with Daniel's idea 14:20:43 this is what ODL tests do : https://wiki.opnfv.org/r1_odl_suite 14:20:50 <[1]JonasB> szilard: Not on top of this but I thing ODL is using robot, and we want to reuse what is already done in ODL? 14:21:59 aha ok, so that's why 14:22:43 <[1]JonasB> jose_lausuch: So how do we proceed here, any proposal? 14:22:55 need to wait for Daniel to see what makes sense 14:23:15 and also need to dig in the odl testing that peter wrote 14:23:25 to understand it better to see if its what Daniel suggested 14:23:36 but I presume it is 14:23:43 so no need to write yet another test for that 14:23:45 <[1]JonasB> jose_lausuch: Do you have access to the robot suite? 14:23:50 yes 14:23:53 or what do you mean? 14:24:04 <[1]JonasB> Cool, so lets wait for Daniel then. 14:25:05 at least, creating stuff manually and running the curl command works 14:25:28 I get the right output from D' script 14:25:53 <[1]JonasB> #info Jose have done the ODL tests manually and it seems to work. 14:26:07 <[1]JonasB> #info Jose will look into robot. 14:26:30 <[1]JonasB> Jose: Do you know if D's scripts survive a restart? 14:27:02 dont know 14:27:08 <[1]JonasB> Ok 14:27:18 <[1]JonasB> Should we move on? 14:27:34 yes 14:27:39 yep 14:27:47 <[1]JonasB> #topic Autodeploy rebasing status for SR1 14:28:01 <[1]JonasB> Szilard - youre the master now 14:28:04 ok 14:28:27 #info Szilard is working on the missing part, that is Reaping from an existing Fuel configuration 14:28:33 #info 30% 14:28:58 and that's it 14:29:24 <[1]JonasB> So my proposal is that we give the LF1 POD to Szilard the minute after we have released the 4th of June 14:29:36 sounds good :) 14:29:54 ok 14:29:56 I agree. 14:30:10 And until then, use all blades you can in Montreal. 14:30:28 sure, I will do :) 14:30:43 The logic should be same (Fuel libvirt, Compute/Controller IPMI) 14:30:54 <[1]JonasB> #info Szilard will get access to LF1 to test out the SR1 rebase of the autodeploy function immediately after the Arno release 14:31:06 brb (1min) 14:31:25 <[1]JonasB> brb=afk or?? 14:32:04 Apparently. :) 14:32:04 <[1]JonasB> Move forward? 14:32:11 * stefan_berg nods 14:32:25 Short for brb, which some people use literally, and are gone anywhere from 3 seconds to 5 minutes 14:32:28 google :)) 14:32:40 <[1]JonasB> :-) 14:33:04 <[1]JonasB> #topic ODL Fuel plugin status 14:33:11 <[1]JonasB> Michal - go 14:33:13 #info created skeleton of the plugin which I keep for now in my git repository. 14:33:25 #info started to use fpm ruby gem to create deb package with opendaylight karaf during plugin build. Plugin will contain this package, and if user choose this plugin opendaylight will be installed on primary controller 14:33:32 back! :p 14:33:37 #info also working on adding some custom fields to fuel UI which allow end user to choose for example on which port opendaylight rest api will be available 14:33:51 first build can be downloaded from here 14:33:57 #link http://devel.opt.im/opendaylight-1.0-1.0.0-1.noarch.rpm 14:34:08 and installed on fuel master with: fuel plugins --install opendaylight-1.0-1.0.0-1.noarch.rpm 14:34:18 for now when activated in settings tab it will only install opendaylight files on primary controller node 14:34:33 and it is all for me 14:34:37 Cool stuff! 14:34:56 <[1]JonasB> Great, must try that! 14:35:20 Started playing with plugins today (6.0, as my 6.1 isn't installing) and it's really easy to work with! 14:35:30 <[1]JonasB> mskalski: any issues worth mentioning? 14:35:48 btw, if you want to 6.1 iso you try dowload from here 14:36:00 |#link https://ci.fuel-infra.org/view/ISO/ 14:36:26 Ah, great, exactly what I wanted to ask about! :) Thanks, will try that. 14:36:30 [1]JonasB: for now no, but this part is easy 14:37:36 <[1]JonasB> Question - should the ODL plugin be in the OPNFV repo or should we rather have it in the Fuel repo and expose it as a fuel plug-in in the "Fuel App store"? 14:38:34 Goes towards maintability I guess? 14:38:38 most of plugins are avaiable here: https://github.com/stackforge?utf8=%E2%9C%93&query=fuel-plugin 14:39:25 but for now our plugin is not ready to be published on stackforge 14:40:39 <[1]JonasB> Lets decide later in the design cycle 14:40:47 ok 14:42:09 <[1]JonasB> Anything more we need to discuss? 14:42:44 I'm good. 14:43:19 <[1]JonasB> Take that as a no 14:43:25 mskalski: All ISOs seeded? Haven't seen any Torrent dl yet... 14:44:03 <[1]JonasB> Thanks everyone 14:44:08 <[1]JonasB> #endmeeting