16:00:46 #startmeeting openstack_ansible_meeting 16:00:47 Meeting started Tue Feb 12 16:00:46 2019 UTC and is due to finish in 60 minutes. The chair is mnaser. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:00:48 #topic rollcall 16:00:49 o/ 16:00:49 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 16:00:50 o/ 16:00:51 The meeting name has been set to 'openstack_ansible_meeting' 16:01:16 @fnpanic: patch basically adds environment variable REQUESTS_CA_BUNDLE to heat systemd service 16:01:45 o/ 16:01:45 p/ 16:01:47 rpsene something to try - clear out the fact cache, then set the fact gathering subset to 'all', then try again, ie rm -rf /etc/openstack_deploy/ansible_facts; export ANSIBLE_GATHER_SUBSET='all'; ... execute the playbooks ... 16:01:51 o/ 16:02:08 o/ 16:02:31 #topic Last week highlights 16:02:40 anyone gots some fun highlights they want to share from last week? 16:02:52 FOSDEM was interesting :) 16:03:29 i hear it was a mini ptg :) 16:03:53 mnaser haha, not quite - but there was much beer, and evrardjp and I managed to have a short chat about config_template 16:04:00 o/ 16:04:17 fun times 16:04:28 more progress on that in https://review.openstack.org/#/q/topic:bug/1791258 and https://etherpad.openstack.org/p/osa-config-template-migration 16:05:16 and releases too ! :p 16:05:41 sounds like a fun time 16:05:45 (I was merely pushing my laziness agenda I guess?) 16:06:27 i support that agenda, hah 16:06:36 oh yes, he told me about https://review.openstack.org/#/q/topic:easier_osa_releases which is awesome :) 16:06:49 awesome, we're almost landing that 16:06:51 https://bugs.launchpad.net/openstack-ansible/+bug/1815631 16:06:52 Launchpad bug 1815631 in openstack-ansible "[Docs] all examples lacking correct horizon image upload mode" [Undecided,New] 16:07:04 with that merged, we'll never have to update openstack_release, ever again (from stein onwards) 16:07:04 I will backport those 16:07:12 🔥 16:07:23 i'm excited :) 16:08:45 oh, and we chatted a bit about the python build simplification - I got confirmation from evrardjp that my thinking was right - I'll do some updates to the existing patches on friday and will hopefully be able to spread the pattern across other roles if everyone's happy with them... so we should be able to chat about that next week in more detail 16:09:25 with that done, I'll circle back to using the integrated repo for all role tests - I think we should be able to get it done for stein 16:09:49 w00t, odyssey4me please let us know if you need any help with all of this :) 16:10:02 During that conversation I pointed out the fact that we should branch when we feel this is ready 16:10:19 slightly tangential to that, I need another review for https://review.openstack.org/636297 to unblock the os_heat repo's integrated build tests: https://review.openstack.org/636297 16:10:36 urgh, sorry - I got a little over excited with copy/paste there 16:11:29 how can i use the ansible inventory groups in my plays? 16:11:42 and this patch unblocked this https://review.openstack.org/#/c/634032/3 odyssey4me :) 16:11:52 it is still finishing but I see all green 16:12:19 yep, great - thanks guilhermesp 16:12:33 fnpanic we can discuss that after the meeting if you don't mind? 16:12:42 ok sorry 16:12:46 coolio 16:12:54 that's some awesome updates over the weekends :) 16:12:56 was not awarre thar is it that late 16:13:09 or the week rather ;) 16:13:15 For the record, I filed a bug for the horizon -> cinder issue I had with my AIO https://bugs.launchpad.net/horizon/+bug/1815635 16:13:16 Launchpad bug 1815635 in OpenStack Dashboard (Horizon) "Horizon is unable to retrieve Cinder API versions when it has a self-signed SSL certificate" [Undecided,New] 16:13:29 aaaand some clues about this https://review.openstack.org/#/c/627782/6 the workspace move maybe is not working as expected 16:14:14 oh, I see the placement role patch merged - is that all good to go? if so, then part of my work to use the integrated build for role testing will be to work on the upgrade path.... but someone will still need to change the nova role up to use the new placement role 16:14:39 odyssey4me: yes, my next step is doing os_nova to use os_placement 16:14:48 thanks dmsimard - we'll move to triage shortly and that'll likely come up 16:15:14 odyssey4me: I filed the bug in horizon, but it looks like there are improvement opportunities for self-signed certificates in AIO in general 16:15:28 mnaser ok cool - just greenfield will be fine, we'll work on the upgrade path later and test it using the integrated build, hopefully in the gate :) 16:15:57 Georgina Shippey proposed openstack/openstack-ansible-os_neutron stable/rocky: Add support for dns_domain_ports api extension https://review.openstack.org/636368 16:16:24 apologies, I wasn't aware a meeting was ongoing 16:17:13 we need to have a job that uses in-repo placement in os_nova 16:17:38 and then another one that uses the new os_placement role 16:17:45 (because nova wants to test in-repo placement using our jobs) 16:18:50 but yes indeed odyssey4me we'll figure it out for greenfield and then upgrade will be the tricky one 16:19:46 #topic bug squash day 16:19:53 mnaser heh, sounds like a bit of fun then 16:20:35 * mnaser #link https://etherpad.openstack.org/p/osa-bug-squash-q1 16:20:38 #link https://etherpad.openstack.org/p/osa-bug-squash-q1 16:20:45 dmsimard most definitely, that's an itch which no-one has really scratched yet unfortunately... although some reasonable work was done in the galera_server/galera_client roles which may be worth looking at 16:21:38 i came up with a little link to share and work on some of our progress 16:21:47 for tomororw hopeuflly 16:22:15 right on 16:22:17 I'm working weird hours tomorrow but I'll work through the review list when I can for sure 16:23:04 is removing old facts, # in the config and using inventroy-manage --remove to delete enough to remove services .. like i want to not use aodh/gnocchi anymore 16:23:07 how does everyone feel about a high bandwidth way of chatting 16:23:19 zoom? hangouts? good o'l conference call? :P 16:23:28 I unfortunately won't be able to actively participate in the bug triage, other than to be available to answer questions and advise where I can via IRC. 16:23:40 I vote zoom 16:24:37 mnaser I would personally prefer IRC, because it's all logged and open - but it may make sense to use vid conf for small groups who're working together... 16:24:57 odyssey4me: yeah im thinking maybe like a subgroup to hack on a specific task if they feel like but *leaving* everything in irc 16:25:05 admin0 we can discuss that after the meeting 16:25:40 sorry .. did not realized we are in a meeting ( just walked in) 16:25:45 mnaser maybe create 2-3 zooms and have links in the ether that folks can check in and out? 16:25:47 and log all that stuff in the etherpad to track our progress 16:26:10 yeah I think would be good to divide and conquer in this way...main topics on irc and small groups discussing in a conference, then registering things on etherpad 16:26:11 i'll leave it as on-demand 16:26:15 we can get a good room anytime :) 16:26:28 mnaser yep, makes sense - and spotz' suggestion to have a few rooms to use for particular focus groups is a good one, I think. 16:26:32 hangouts or whatever can be really on demand 16:26:39 yeppers 16:26:58 does anyone have any other suggestions for that etherpad? 16:27:05 it's my first time running something like this so all advice is welcome 16:27:26 Maybe just a note when something is merged so those of us focusing on reviews don't keep opening things? 16:27:29 zoom's a great way to get more eyes on the same console to see the same problem as the person doing the triage/fix 16:28:00 spotz I think that's what gerritbot is for. ;) 16:28:28 odyssey4me: Bah, you know I tend to squirrel onto things in front of me!:) 16:28:53 o/ 16:29:21 i have fully carved out my day to dedicate for this :) 16:29:42 so it'll be fun except for the fact we can't go out after we're done some good work D: 16:30:01 o/ 16:30:01 I see what you did there 16:31:56 Not that you managed to go out with us in Berlin mnaser.....:) 16:32:23 hahahah 16:32:26 hey cloudnull and cjloader 16:32:28 mnaser True. I think I'd like to facilitate a virtual mid-cycle in Train for OSA where we can focus on achieving our goals, so I'm interested to see how this goes. I'm thinking it may also be useful to facilitate a regular bug/review smash for the Ansible OpenStack modules too, given that those reviews and bugs aren't getting enough attention either. 16:32:57 odyssey4me: i agree! this will be a good attempt at seeing that 16:33:08 +1 odyssey4me 16:33:15 i would love for us to take a day or two to sprint and get all those super awesome things we discucsed at the ptg 16:33:16 we've done great progress 16:34:29 * redrobot sneaks into the meeting 16:34:51 so if anyone is around does anyone have a show of hands of who might be there tomorrow? :) 16:35:02 i'll do my best 16:37:07 we will be there 16:37:08 mnaser: o/ 16:37:27 o/ but at weird times 16:37:39 coolio 16:37:50 given we'll be doing a bunch of bug triage tomorrow i'll skip that :) 16:37:53 #topic open discussion 16:38:32 o/ 16:38:45 hi redrobot :) 16:38:54 Hi mnaser! 16:39:13 something fun to share for our open discussion ? :P 16:39:15 looks like I snuck in here just in time :D 16:39:20 yes! 16:39:44 So, both HSM ansible roles are part of openstack-ansible now! 16:39:58 🎉🎉🎉 16:40:09 Thanks to everyone who +1/+2 patches 16:40:23 There's only one outstanding thing in my todo 16:40:48 and that's adding myself and ade_lee (or barbican-core) to the review teams for those two roles 16:41:01 #link https://review.openstack.org/#/admin/groups/2005,members 16:41:10 redrobot personally, I'd be happy to see barbican-core in there 16:41:11 #link https://review.openstack.org/#/admin/groups/2005,members 16:41:46 mnaser, I believe you own both groups... so you'd be the one to add new members 16:41:46 +1 16:42:16 I think we're all sensible enough to review on the basis of the things we understand. 16:42:29 odyssey4me, spotz 😁 16:42:30 I'd suggest also adding the openstack-ansible-core group there too. 16:42:38 indeed 16:43:12 i agree with odyssey4me. 16:43:14 odyssey4me, +1 ... I agree sharing duties between our two teams seems ideal 16:43:23 redrobot: i added barbican-core and osa-core to that group 16:43:37 good job mnaser 16:43:43 please don't shy from sending reviews as you start 16:43:58 mnaser, how about the other one? 😬 16:44:04 i see the same link twice D: 16:44:08 oh oops 16:44:09 my bad 16:44:21 #link https://review.openstack.org/#/admin/groups/2006,members 16:44:27 ^^ that's the second one 16:44:28 redrobot I'm not really all that familiar with the subject matter, so hopefully someone in the community will adopt them for their environment and start working on getting them working with our integrated build - or creating an appropriate user story and some openstack-ansible-ops playbooks to help people understand how to use them. 16:44:32 bad redrobot! 16:44:57 voila redrobot :) 16:45:30 odyssey4me, agreed, the challenge there is to have both HSMs readily available for test. We're working on downstream tests for sure, but I'll try to push for upstream testing as well. 16:46:02 I'm soliciting feedback on https://docs.openstack.org/openstack-ansible/latest/user/network-arch/example.html and https://docs.openstack.org/openstack-ansible/latest/user/prod/provnet_groups.html for those interested. Both new docs. 16:46:06 redrobot well, if there's some sort of downstream testing via external CI to provide active feedback to changes on the role, that'd be good enough in my books 16:46:11 (additionally, to complicate things some HSM client software may not have open source licenses. 😔 16:46:22 tbh I am not too familiar with those, but I guess if we are a little trained with them we could review more efficiently. I can only review the ansible bits to see if they match the latest standards 16:46:24 you can still have 3rd party voting 16:46:25 redrobot: I see 16:46:28 odyssey4me, ack, I will definitely keep that in mind. 16:46:33 s/3rd party voting/3rd party ci/ 16:46:39 mnaser, much appreciated! :D 16:47:04 ooh jamesdenton you drew purdy pics :) 16:47:15 ok, that's all I have 16:47:39 just for you, odyssey4me 16:47:58 hehe 16:48:19 Adri2000: plan on backporting I2a3e266e3066d6c271a29038e814ff8e3090fafb ? looks like nova has the same change in rocky (and queens, pike) 16:49:22 jamesdenton: i like those docs! 16:49:41 anyone else has any other proposed solutions? 16:49:47 thanks! 16:49:50 jamesdenton at a cursory glance, that's really nice work - great example configs and all 16:50:27 prometheanfire anyone can propose the backport :) 16:50:55 mnaser proposed solutions? I'm confused 16:51:04 mnaser: i'll be there tomorrow -- had to check w my manager 16:51:08 `ER 16:51:10 proposed topics 16:51:12 im sorry 16:51:15 :< 16:51:21 odyssey4me: there are other issues it looks like, how did both https://github.com/openstack/openstack-ansible-os_nova/commit/f67e610609df9f99778de00630c7e0db9b4295ed and https://github.com/openstack/openstack-ansible-os_nova/commit/a383124336dfc74b4d652be594f7bba0197c24b1 make it in, they both deleted the same section of the template 16:52:48 prometheanfire heh, I have no idea 16:53:19 ya, need to fix that up too 16:53:22 mnaser I have nothing for now. 16:53:41 cool, anyone else? :) 16:53:50 mnaser: I just have the upgrade jobs that we need some feedback of tempest guys 16:53:55 we already discussed about it too 16:53:56 backports would be incomplete so I might make a combined (correct) one. 16:54:35 also AFIK https://review.openstack.org/#/c/635579/3 depends on a fix that I don't fully remember now 16:54:39 guilhermesp oh, that workspace thing is odd - so running the 'move' command doesn't actually move it? perhaps we should just move it with 'mv' after the 'move' command? 16:54:54 prometheanfire yep, sounds good 16:55:10 maybe odyssey4me ... what I saw in the hold is that the old workspace is in the same place and the new one is empty 16:55:18 it's like nothing was moved, just in the api point of view 16:55:22 prometheanfire: I think I didn't backport that because it's quite easy to override in the config, but feel free to submit a backport if needed 16:55:48 that's why we could be experiencing these nested venv issues when running tempest run -l between the two releases 16:55:52 guilhermesp I see other os_nova patches passing CI today - so perhaps that just needs a recheck now 16:56:04 odyssey4me: will do that now 16:56:45 shall i end? :> 16:57:29 for me it's ok mnaser 16:57:34 coolio 16:57:36 #endmeeting