14:00:50 #startmeeting Cross Community CI 14:00:50 Meeting started Wed Mar 21 14:00:50 2018 UTC. The chair is hwoarang. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:50 Useful Commands: #action #agreed #help #info #idea #link #topic. 14:00:50 The meeting name has been set to 'cross_community_ci' 14:00:56 #topic Rollcall 14:00:59 identify yourselves 14:01:05 #info Jack Morgan 14:01:06 #info Dimitrios Markou 14:01:09 #info Tianwei Wu 14:01:16 fdegir is away so i will chair the meeting this week. Hopefully it will be short 14:01:21 #info Markos Chandras 14:01:30 #info Victor Morales 14:01:40 agenda is on the usual place 14:01:42 #link https://etherpad.opnfv.org/p/xci-meetings 14:01:57 please add anything that you think it's missing 14:02:05 i will wait 1 min before we begin 14:03:19 #info David Blaisonneau 14:03:42 ok moving on 14:04:13 #topic High Priority Items 14:04:36 #info CI verification jobs for both internal and external scenario have been enabled 14:04:50 which basically means that CI will test the right thing based on what your patchset modifies 14:05:15 #info centos is not voting and needs love 14:05:47 #info merge jobs has also been enabled for internal and external scenarios 14:06:08 so the taking point here is that CI is doing well, but if you care about centos it's time to step up :) 14:06:38 #info functest - openstack healthcheck is used in verify and merge jobs 14:06:49 #info functest - smoke test integration is WIP 14:07:04 #info functest - k8s healthcheck integration is WIP 14:07:42 #info baremetal - patches at https://gerrit.opnfv.org/gerrit/#/q/status:open+project:releng-xci+branch:master+topic:stabilization 14:07:46 #link https://gerrit.opnfv.org/gerrit/#/q/status:open+project:releng-xci+branch:master+topic:stabilization 14:07:56 so anything that you would like to add, discuss etc? 14:08:37 I encourage you to actually try the BM patches and review them. 14:08:39 #info Manuel Buil 14:09:10 and feel free to contact me for any support on PDF/IDF creation for your infra 14:09:14 hwoarang: those jobs are not voting right? 14:09:20 #info Periyasamy Palanisamy 14:09:22 merge jobs I mean 14:10:46 mbuil: probably not 14:10:51 but i am not too sure 14:11:24 ok moving on 14:11:27 #topic Quick Status Check for the Scenarios 14:11:37 scenario people anything that you would like to say? 14:12:44 no update from sfc. Still working on patches 14:12:53 hwoarang: I am try to create the k8-canal-nofeature scenario 14:13:05 #info do not use ansible <2.4.3 14:13:17 #info will fail with SSL error 14:13:18 mardim: what is the problem? 14:13:52 #info hw_wutianwei is working on the k8-canal-nofeature scenario 14:13:53 mbuil, if we use ansible lower that 2.4.3 it will give SSL error 14:14:05 is a bug in Ansible if I am not mistaken 14:14:11 mardim: something else is missing it's very unlikely for ansible to have broken ssl for all these years ;p 14:14:40 hwoarang, when I changed the ansible version 14:14:40 maybe 2.4.3 pulled just a pip dependency that you were missing before 14:14:50 and fixed it by 'accident' 14:14:50 SFC pike deployed succesfully 14:15:04 like python-cryptography 14:15:35 hwoarang, Yes Maybe that too 14:15:41 hwoarang, To be honest 14:15:59 hwoarang, I din't search it furhter after it deployed succesfully 14:16:06 :) 14:16:10 so lets figure it out the next time it happens 14:16:16 ok scenarios are good then moving on 14:16:18 hwoarang, cool 14:16:23 The upstream reviews for bgpvpn scenario are merged in osa and ansible-opendaylight. will work with fdegir in updating xci sha's 14:16:53 #info bgpvpn scenarion merged in osa/queens ansible-opendaylight. Need to bump shas soon 14:17:03 #undo 14:17:03 Removing item from minutes: 14:17:06 can't type 14:17:13 #info bgpvpn scenario merged in osa/queens and ansible-opendaylight. Need to bump shas soon 14:17:17 ok thanks epalper 14:17:35 hwoarang, mbuil more info here https://github.com/ansible/ansible/issues/25402 14:17:57 i think we will do one more queens bump before moving to rocky to be honest 14:18:50 #info ansible -2.3 is having issues with ssl - https://github.com/ansible/ansible/issues/25402 14:19:06 #info possibly affecting OSA pins < Queens 14:19:23 thanks mardim 14:19:34 ok nothing else from scenarios i suppose 14:19:38 #topic - AoB 14:19:59 anything for ONS next week worth mentioning here? 14:20:39 mardim and I are showing a demo in the LFN booth on Wednesday and Thursday :) 14:20:55 demo about XCI. Pass by and talk to us! We like people ;) 14:21:11 #info ONS - mardim and mbuil are showing an XCI demo in the LFN booth on Wednesday and Thursday 14:21:16 I will be giving demo on BGPVPN 14:21:27 I will give candies for free :) 14:21:34 I forgot epalper is also in the demo :) 14:21:41 epalper: based on XCI righ? :D 14:21:58 yes hwoarang, from XCI :) 14:22:02 #info ONS - epalper will have a demo about XCI BGPVPN 14:22:04 cool! 14:22:19 better not break anything by then :p 14:22:27 *until then! 14:22:47 I just had trial run with latest xci, it works :) 14:23:01 great 14:23:08 good luck with your demos then! 14:24:15 anything else to discuss today? 14:25:24 one thing 14:25:41 yes 14:25:45 just about PDF/IDF introduction for installer/scenarios 14:26:22 i will propose some atomic patches (after ONS :)), but they will be based on https://gerrit.opnfv.org/gerrit/#/c/53893 14:26:52 this is PDF/IDF for deployment in VM, with the same config we are using today 14:27:36 the idea is to continue deploying on VM as today but more and more getting IPs and other POD values from PDF 14:27:56 then introduce PDF/IDF from BM 14:28:19 is there any comment/proposition about this subject or the way to do it ? 14:28:31 #info David to introduce pdf/idf for installer/scenarios for VM and BM deployments 14:28:41 David_Orange: how does that relate to your existing patches? 14:28:45 because i got confused a bit 14:29:25 the next patches will source those config files 14:29:38 your patches do not contain IDF/PDF right? so i guess this is a continuation 14:30:16 the stabilisation patches ? yes, they contains the same, and it is a continuation 14:30:45 but with separate patches, to introducte them smothly 14:30:52 smoothly 14:31:21 ok so for me, it would look like this 1) get idf/pdf files 2) make bifrost deploy using idf/pdf so we keep XCI working 3) bring BM 14:31:23 am i more clear ? 14:31:23 something like that 14:31:37 would that make sense? 14:32:06 for now 1/2 is working with the proto 14:32:31 yeah so i guess that also means start moving bits from proto to real xci correct? 14:32:50 but not the 3, for the installers we need to introduce more template configs 14:33:34 after review, yes, if it get the agreement 14:33:54 Manuel Buil proposed releng-xci: Connect ODL CI-CD to XCI https://gerrit.opnfv.org/gerrit/53173 14:33:56 ok then 14:34:24 getting things in proto is easy for people to try. we will see how to bring them to xci after that 14:34:48 thank you for your work btw 14:35:13 hwoarang: yes, that is why it is better to keep it in proto to let reviewers review this large patches bundle 14:35:23 hwoarang: thx 14:35:54 and now this will be to introduce PDF/IDF in installers/proto 14:36:05 ok thank you for the update 14:36:06 anything else 14:36:13 i know the PDF/IDF is a bit obscure, i hope i am clear 14:37:23 hwoarang: from my side 14:37:26 no 14:37:27 yes it takes a bit to digest that 14:37:41 anybody else? 14:37:48 hwoarang: i understand that :) 14:38:00 I have two patches that are not triggering jenkins jobs https://gerrit.opnfv.org/gerrit/#/c/54125/ https://gerrit.opnfv.org/gerrit/#/c/54123/ 14:38:30 should I need to resubmit them? 14:38:58 electrocucaracha: no because changes in the VM script are being ignored 14:39:12 for security reasons. if you mess with the VM script you can damage the host 14:39:19 :O 14:39:33 so need manual verification 14:40:08 got it, in that case can you make the honors to take a look? 14:40:14 hehehe 14:40:22 yes I will 14:41:01 anything else? 14:41:13 and regarding the patch that renames bridges, we decided to offer different approach 14:41:25 https://gerrit.opnfv.org/gerrit/#/c/52841/ 14:42:20 David_Orange: suggested to have a variable for selected bridged or non-bridged networks 14:42:39 i will review that too 14:43:06 anything else? 14:43:06 I offered the alternative to use based on the installer selected 14:43:08 the fact is that kolla prefers to work on interfaces, not on bridges (like osa) 14:43:48 electrocucaracha patch is to set or not the bridges 14:44:01 electrocucaracha: i will do it tomorrow, thx :) 14:45:00 ok understood 14:45:04 yeah, the thing is that maybe we should consider to configure the networks differently for every installer 14:45:32 pros more flexibility, cons we'll be loosing a reference 14:46:24 let me review that patch to understand the big picture 14:46:32 +1 14:46:40 then we can discuss so i dont say stupid stuff right now 14:47:38 ok unless there is something else to discuss we can end here for today 14:47:56 thx hwoarang 14:48:03 #endmeeting