| *** rstarmer has joined #openstack-ansible | 00:15 | |
| *** dave-mccowan has joined #openstack-ansible | 00:18 | |
| *** cjloader has joined #openstack-ansible | 00:19 | |
| *** cjloader has quit IRC | 00:24 | |
| *** rstarmer has quit IRC | 00:32 | |
| *** cjloader has joined #openstack-ansible | 01:21 | |
| *** cjloader has quit IRC | 01:25 | |
| *** markvoelker has joined #openstack-ansible | 01:47 | |
| *** markvoelker has quit IRC | 02:21 | |
| hw_wutianwei_ | hi everyone, when I try to use ceph as gnocchi_storage_driver, I got a error http://paste.openstack.org/show/649120/. it seem to miss installing the radoi in gnocchi container | 03:05 |
|---|---|---|
| hw_wutianwei_ | ^rados | 03:16 |
| *** markvoelker has joined #openstack-ansible | 03:17 | |
| *** cjloader has joined #openstack-ansible | 03:21 | |
| *** cjloader has quit IRC | 03:25 | |
| *** markvoelker has quit IRC | 03:51 | |
| *** chyka has joined #openstack-ansible | 03:56 | |
| *** chhavi__ has joined #openstack-ansible | 04:05 | |
| *** chyka has quit IRC | 04:09 | |
| *** rstarmer has joined #openstack-ansible | 04:17 | |
| *** Ankita has joined #openstack-ansible | 04:20 | |
| *** dave-mccowan has quit IRC | 04:21 | |
| *** chyka has joined #openstack-ansible | 04:26 | |
| *** chyka has quit IRC | 04:30 | |
| *** bhujay has joined #openstack-ansible | 04:32 | |
| *** ykarel|away has joined #openstack-ansible | 04:34 | |
| *** markvoelker has joined #openstack-ansible | 04:48 | |
| *** armaan has quit IRC | 04:56 | |
| *** armaan has joined #openstack-ansible | 04:57 | |
| *** rstarmer has quit IRC | 05:03 | |
| *** rstarmer has joined #openstack-ansible | 05:17 | |
| *** hybridpollo has quit IRC | 05:18 | |
| *** cjloader has joined #openstack-ansible | 05:21 | |
| *** markvoelker has quit IRC | 05:21 | |
| *** cjloader has quit IRC | 05:25 | |
| *** chyka has joined #openstack-ansible | 05:30 | |
| *** ykarel|away is now known as ykarel | 05:34 | |
| *** chyka has quit IRC | 05:34 | |
| *** threestrands has quit IRC | 05:40 | |
| *** chyka has joined #openstack-ansible | 05:51 | |
| *** lihi has quit IRC | 05:55 | |
| *** oanson has quit IRC | 05:55 | |
| *** chyka has quit IRC | 05:56 | |
| *** rstarmer has quit IRC | 05:57 | |
| *** armaan has quit IRC | 06:14 | |
| *** armaan has joined #openstack-ansible | 06:14 | |
| *** markvoelker has joined #openstack-ansible | 06:18 | |
| *** rstarmer has joined #openstack-ansible | 06:26 | |
| *** Ankita has quit IRC | 06:28 | |
| *** andries has quit IRC | 06:38 | |
| *** andries has joined #openstack-ansible | 06:40 | |
| *** markvoelker has quit IRC | 06:51 | |
| *** oanson has joined #openstack-ansible | 07:02 | |
| *** lihi has joined #openstack-ansible | 07:04 | |
| *** ThomasS has joined #openstack-ansible | 07:08 | |
| *** cjloader has joined #openstack-ansible | 07:21 | |
| *** cjloader has quit IRC | 07:25 | |
| *** chyka has joined #openstack-ansible | 07:27 | |
| *** mardim has joined #openstack-ansible | 07:34 | |
| *** chyka has quit IRC | 07:36 | |
| *** ThomasS has quit IRC | 07:38 | |
| *** ThomasS has joined #openstack-ansible | 07:38 | |
| *** ykarel is now known as ykarel|lunch | 07:41 | |
| *** markvoelker has joined #openstack-ansible | 07:48 | |
| *** pcaruana has joined #openstack-ansible | 07:52 | |
| *** rstarmer has quit IRC | 07:53 | |
| *** mbuil has joined #openstack-ansible | 07:55 | |
| *** rstarmer has joined #openstack-ansible | 08:07 | |
| *** chyka has joined #openstack-ansible | 08:08 | |
| *** chyka has quit IRC | 08:13 | |
| evrardjp | hw_wutianwei_: I am pretty sure nobody worked on that yet. Maybe you could propose a patch to fix it? | 08:14 |
| evrardjp | and good morning everyone :) | 08:14 |
| *** ykarel|lunch is now known as ykarel | 08:20 | |
| *** cjloader has joined #openstack-ansible | 08:21 | |
| *** markvoelker has quit IRC | 08:21 | |
| *** m|y|k has joined #openstack-ansible | 08:25 | |
| *** m|y|k has quit IRC | 08:25 | |
| *** cjloader has quit IRC | 08:25 | |
| *** hamza21 has joined #openstack-ansible | 08:32 | |
| *** pbandark has joined #openstack-ansible | 08:41 | |
| *** openstackgerrit has joined #openstack-ansible | 08:42 | |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-os_cinder master: Fix upgrade test https://review.openstack.org/535791 | 08:42 |
| openstackgerrit | Merged openstack/openstack-ansible-os_glance master: Fix upgrade test https://review.openstack.org/535787 | 08:46 |
| *** sxc731 has joined #openstack-ansible | 08:49 | |
| *** shardy has joined #openstack-ansible | 08:54 | |
| *** shardy has quit IRC | 08:59 | |
| *** shardy has joined #openstack-ansible | 09:00 | |
| *** markvoelker has joined #openstack-ansible | 09:19 | |
| *** jafeha has joined #openstack-ansible | 09:19 | |
| jafeha | good morning | 09:19 |
| *** markvoelker has quit IRC | 09:23 | |
| *** jafeha_ has joined #openstack-ansible | 09:25 | |
| *** jafeha has quit IRC | 09:28 | |
| evrardjp | good morning jafeha__ | 09:32 |
| evrardjp | Taseer: here is the cause of the issue: https://github.com/evrardjp/ansible-keepalived/compare/3.0.0...3.1.0 | 09:33 |
| evrardjp | woops | 09:33 |
| evrardjp | wrong link | 09:33 |
| evrardjp | http://logs.openstack.org/91/522491/39/check/openstack-ansible-functional-ubuntu-xenial/59ce37d/job-output.txt.gz#_2018-01-17_10_56_38_828314 | 09:34 |
| evrardjp | so you have to define this var. | 09:34 |
| evrardjp | check how we do db configuration in other roles to get inspiration. It's not a gating issue, it's a role issue there. | 09:35 |
| *** rstarmer has quit IRC | 09:37 | |
| openstackgerrit | Taseer Ahmed proposed openstack/openstack-ansible-os_congress master: Introduce os_congress role in gerrit https://review.openstack.org/522491 | 09:41 |
| ivveh | evrardjp, cloudnull: just fyi, the patch that was submitted for the lbaas v2 for centos/rhel does work, i had an error in my heat template that screwed me over. should have done manual tests :P lesson learned anyways (for me) | 09:42 |
| *** jafeha has joined #openstack-ansible | 09:44 | |
| *** jafeha_ has quit IRC | 09:46 | |
| *** SmearedBeard has joined #openstack-ansible | 09:52 | |
| *** rstarmer has joined #openstack-ansible | 09:54 | |
| hwoarang | good morning | 09:55 |
| evrardjp | ivveh: could you give me a link? | 09:56 |
| evrardjp | ivveh: how can we improve this testing? Tempest maybe? | 09:56 |
| evrardjp | hwoarang: good morning! | 09:56 |
| evrardjp | hwoarang: could you give me a vote for this: https://review.openstack.org/#/c/531850/4 ? Tell me if you think that's normal, or just my brain :p | 09:57 |
| evrardjp | if you don't mind ofc, it's not the most urgent thing in the planet. | 09:57 |
| hwoarang | ah yes the guide fix let me have a look | 09:57 |
| evrardjp | if you have time* | 09:57 |
| *** chyka has joined #openstack-ansible | 09:58 | |
| evrardjp | yeah I'd like to shape it properly before we start translating it with eumel | 09:58 |
| *** chhagarw has joined #openstack-ansible | 09:58 | |
| *** chhavi__ has quit IRC | 09:58 | |
| odyssey4me | o/ | 09:59 |
| *** chyka has quit IRC | 10:02 | |
| *** szaher has joined #openstack-ansible | 10:02 | |
| *** rstarmer has quit IRC | 10:04 | |
| *** chhagarw has quit IRC | 10:04 | |
| *** yolanda has joined #openstack-ansible | 10:17 | |
| *** electrofelix has joined #openstack-ansible | 10:23 | |
| evrardjp | good morning odyssey4me | 10:24 |
| evrardjp | my last mac os x update seem to randomly break my ssh-agent. That's so fun. | 10:24 |
| evrardjp | randomly is probably when I close my laptop, or open a tmux session, I don't know yet, didn't want to spend the time debugging this. But it's fun times! | 10:25 |
| *** hamza21 has quit IRC | 10:25 | |
| *** dcdamien has quit IRC | 10:31 | |
| *** dcdamien has joined #openstack-ansible | 10:32 | |
| Taseer | evrardjp: do I need to set the db_password in the default variables to overcome this => http://logs.openstack.org/91/522491/40/check/openstack-ansible-functional-ubuntu-xenial/f6ee580/job-output.txt.gz#_2018-01-22_10_06_02_751319 | 10:33 |
| evrardjp | Taseer: you can add these in your os_congress-overrides.yml | 10:39 |
| evrardjp | if you think it will be used by multiple roles, then it's maybe wise to add those in our tests repo later: https://github.com/openstack/openstack-ansible-tests/blob/master/test-vars.yml | 10:40 |
| evrardjp | but in the meantime, let's put it in the role tests | 10:40 |
| *** gillesMo has joined #openstack-ansible | 10:41 | |
| Taseer | evrardjp: okay | 10:43 |
| *** stuartgr has joined #openstack-ansible | 10:43 | |
| openstackgerrit | Taseer Ahmed proposed openstack/openstack-ansible-os_congress master: Introduce os_congress role in gerrit https://review.openstack.org/522491 | 10:47 |
| gillesMo | hello ! I'm fighting with neutron config, and I wonder if what I try to do is possible : Using Linux bridge on hosts and for some VLAN provider networks, and using openvSwitch for VXLAN tenant networks ? If think it cannot because it need neutron-linuxbridge-agent and neutron-openvswitch-agent to run, inside the same container (there's only one neutron-agent container) | 10:50 |
| *** taseer1 has joined #openstack-ansible | 10:54 | |
| *** Taseer has quit IRC | 10:55 | |
| *** taseer2 has joined #openstack-ansible | 10:55 | |
| *** taseer1 has quit IRC | 10:59 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_keystone master: Updated from global requirements https://review.openstack.org/535060 | 10:59 |
| *** taseer2 is now known as Taseer | 11:16 | |
| *** yolanda has quit IRC | 11:17 | |
| *** markvoelker has joined #openstack-ansible | 11:19 | |
| *** chyka has joined #openstack-ansible | 11:20 | |
| *** chyka has quit IRC | 11:21 | |
| *** cjloader has joined #openstack-ansible | 11:21 | |
| jafeha | gillesMo, i'm not totally sure, but i don't think it is possible to mix those ml2 drivers (lbridges and ovs) in neutron. tbh i don't see why you should do such a thing. what is it you're trying to accomplish? | 11:23 |
| hwoarang | evrardjp: re https://review.openstack.org/#/c/535252/ your comment about other arches | 11:25 |
| hwoarang | current code is only deployed on x86_64 | 11:25 |
| *** cjloader has quit IRC | 11:26 | |
| jafeha | gillesMo, as you can use vlan and vxlan networks with both drivers, i do not see any advantage in mixing them. | 11:26 |
| evrardjp | hwoarang: power? | 11:26 |
| hwoarang | mariadb repo for centos is hardcoded to x86_64 and the percona repos do not provide anything else but x86_64 | 11:26 |
| hwoarang | the percona downloads on https://www.percona.com/downloads/XtraBackup/LATEST/ only list x86_64 | 11:26 |
| evrardjp | that is true, but IIRC, the IBM guys builded their own packages | 11:27 |
| evrardjp | if we don't have that, it would be due to a regression | 11:27 |
| hwoarang | oh so it's for external usage? | 11:28 |
| evrardjp | https://github.com/openstack/openstack-ansible-galera_server/blob/stable/pike/vars/ubuntu-16.04.yml#L87 | 11:28 |
| evrardjp | they had a deb for power | 11:28 |
| hwoarang | :/ | 11:28 |
| evrardjp | sorry to be pain in the _ | 11:29 |
| hwoarang | ok so it's only for ubuntu | 11:29 |
| evrardjp | yeah | 11:29 |
| evrardjp | to what I am aware | 11:29 |
| hwoarang | well ok at least we can clean the rest | 11:29 |
| evrardjp | on the package: we can do a deb: "{{ deb_url | default(omit) }}" | 11:30 |
| evrardjp | replace deb_url with what's relevant that gets only defined for power under ubuntu | 11:31 |
| evrardjp | haha no that wouldn't be nice | 11:31 |
| hwoarang | i was hoping to drop all that code from downloading packages and simply use the repos | 11:31 |
| evrardjp | yes | 11:31 |
| evrardjp | we should do that | 11:31 |
| hwoarang | maybe move all that stuff to the ubuntu install file instead | 11:31 |
| hwoarang | i will have a look | 11:31 |
| evrardjp | not sure what you meant there | 11:32 |
| evrardjp | so many ways to do it | 11:32 |
| evrardjp | we can even use openstack_hosts for it | 11:32 |
| hwoarang | true | 11:33 |
| *** openstackgerrit has quit IRC | 11:33 | |
| evrardjp | but I don't think it's a good idea to use openstack_hosts either :p | 11:33 |
| evrardjp | odyssey4me: andymccr do you know if the IBM crew could have an apt mirror for these galera packages, instead of having files? | 11:34 |
| evrardjp | that would simplify stuff | 11:34 |
| hwoarang | that^ | 11:34 |
| evrardjp | hwoarang: is OBS working for all the distros? | 11:35 |
| evrardjp | I mean we could maybe use this kind of things | 11:35 |
| hwoarang | yeah | 11:35 |
| evrardjp | or we build a ppa for those packages. | 11:35 |
| hwoarang | good idea | 11:36 |
| hwoarang | i can actually do it pretty easily | 11:37 |
| evrardjp | oh? | 11:37 |
| evrardjp | wait there is maybe a licensing thing to sort out... | 11:38 |
| evrardjp | I am pretty sure that's the kind of issues we're gonna get. | 11:38 |
| hwoarang | yeah and ubuntu+ppc64el doesn't seem to work on obs maybe there is no machine to build such thing | 11:39 |
| evrardjp | maybe mhayden knows a little more to that story too | 11:39 |
| evrardjp | yeah would make sense | 11:39 |
| evrardjp | but for ppa building it would pretty much depend on the licensing I guess | 11:39 |
| hwoarang | ok anyway we shall see | 11:39 |
| evrardjp | except if we can pull stuff live during package install | 11:39 |
| odyssey4me | evrardjp IIRC their longer term plan was to use the packages in xenial, but I don't know if the packages are available in xenial for them - if not there, then if there are other mirrors available cc adreznec | 11:42 |
| evrardjp | I think first step would be to find out whether it's still used, why, and how... Then we can think about PPA | 11:42 |
| evrardjp | odyssey4me: that sounds like a solid plan | 11:43 |
| evrardjp | seing how it will happen for 18.04 would be important too | 11:44 |
| hwoarang | xenial has everything | 11:46 |
| hwoarang | except qpress :( | 11:46 |
| evrardjp | I don't know what qpress is :p | 11:47 |
| evrardjp | we are carrying this around for ages, I never really asked. | 11:47 |
| evrardjp | should we carry it? | 11:47 |
| hwoarang | no clue it's some compressing library it seems | 11:48 |
| hwoarang | i guess percona can work without it | 11:48 |
| evrardjp | makes sense to optimise per arch though :p | 11:49 |
| evrardjp | I foresee more breakages soon | 11:49 |
| hwoarang | let me check | 11:49 |
| odyssey4me | I have no idea. cloudnull added it some time ago. | 11:49 |
| evrardjp | hwoarang: I think it's better to wait for adreznec 's comments now | 11:49 |
| odyssey4me | During the great kilo rewrite I think. | 11:50 |
| hwoarang | sure | 11:51 |
| *** openstackgerrit has joined #openstack-ansible | 11:51 | |
| openstackgerrit | Taseer Ahmed proposed openstack/openstack-ansible-os_congress master: Introduce os_congress role in gerrit https://review.openstack.org/522491 | 11:51 |
| evrardjp | odyssey4me: is that normal that everything that start with "the great ..." is causing me an irrationale fear? | 11:52 |
| evrardjp | probably due to the great purge :p | 11:53 |
| openstackgerrit | Markos Chandras (hwoarang) proposed openstack/openstack-ansible-galera_server master: tasks: Drop use_percona_upstream variable variable https://review.openstack.org/535252 | 11:53 |
| odyssey4me | or the great fire ;) | 11:53 |
| evrardjp | or the great plague ! | 11:54 |
| *** markvoelker has quit IRC | 11:54 | |
| evrardjp | the great alexander? | 11:54 |
| evrardjp | oh no that's Alexander, the great. Less scary. | 11:55 |
| evrardjp | on that fun stories.... hwoarang I am fully behind you on the simplification on this role | 11:55 |
| evrardjp | well, that was a french way of saying "you have my full support" | 11:55 |
| hwoarang | hahaha | 11:56 |
| *** Smeared_Beard has joined #openstack-ansible | 11:57 | |
| *** chyka has joined #openstack-ansible | 11:57 | |
| *** SmearedBeard has quit IRC | 11:58 | |
| odyssey4me | I tried doing that in https://review.openstack.org/#/c/466827/2/files/mariadb.fact by using local facts instead of tasks to figure out the state... but alas that way of doing things never made it through to the end of the patches for that change. | 11:58 |
| odyssey4me | I didn't have the time and understanding to get it through to the end. @cloudnull did a reasonable job of simplifying based on what was there though. master is better than pike, I think. | 12:00 |
| *** chyka has quit IRC | 12:02 | |
| *** dave-mccowan has joined #openstack-ansible | 12:12 | |
| *** bhujay has quit IRC | 12:18 | |
| *** bhujay has joined #openstack-ansible | 12:18 | |
| *** sxc731 has quit IRC | 12:20 | |
| *** dave-mcc_ has joined #openstack-ansible | 12:21 | |
| *** dave-mccowan has quit IRC | 12:22 | |
| hwoarang | yeah master is somewhat cleaner | 12:25 |
| openstackgerrit | Merged openstack/openstack-ansible master: Include P-Q changes into the upgrade script https://review.openstack.org/531123 | 12:28 |
| evrardjp | wow | 12:29 |
| evrardjp | it merged | 12:29 |
| evrardjp | let's hope I haven't forgotten anything there, and let's try it! | 12:29 |
| *** ThomasS has quit IRC | 12:32 | |
| *** strobelight has joined #openstack-ansible | 12:33 | |
| *** smatzek has joined #openstack-ansible | 12:33 | |
| *** gameon has joined #openstack-ansible | 12:33 | |
| gameon | Hi all - I've just seen that the 16.06 release has been cut, is there any advantage in running it against a fresh 16.05 deployment? | 12:34 |
| *** bhujay has quit IRC | 12:39 | |
| jafeha | gameon, depends on your setup, you can check the release notes here: https://docs.openstack.org/releasenotes/openstack-ansible/pike.html | 12:39 |
| gameon | Thanks jafeha : It looks as though I may well leave it, though I would imagine there are other improvments for each O/S component | 12:40 |
| gameon | I am a little bit hesitatnt to run it again against a working deployment... | 12:41 |
| jafeha | gameon, i guess that decision depends on your level of production / integration. i probably would not upgrade every minor in production unless there are serious security patches included. during an integration period i'd probably do a couple of minors to learn / test upgrade paths. | 12:43 |
| *** smatzek has left #openstack-ansible | 12:44 | |
| gameon | jafeha: Thank you - in theory I should be able to get the latest tag and re-run using my existing inventory against the current containers? | 12:45 |
| jafeha | gameon, have a look here https://docs.openstack.org/openstack-ansible/pike/user/minor-upgrade.html | 12:49 |
| gameon | Awesome, thank you | 12:49 |
| jrosser | evrardjp: so we got to the bottom of our deploy slowness problem, this line https://github.com/openstack/openstack-ansible/blob/bab51e2967bfa060576d85efaab52ce63d7cfaea/etc/openstack_deploy/user_variables.yml#L138 is evaluated per task, rather than once as you'd maybe hope | 12:50 |
| *** markvoelker has joined #openstack-ansible | 12:51 | |
| jrosser | groups[] and hostvars[] are huge stuctures so it's really slow. do you have a suggestion for where to put an "evaluate once" expression like that, as there are likley many ways to do it | 12:51 |
| jrosser | ^ odyssey4me advice welcome on this | 12:51 |
| *** woodard has quit IRC | 12:52 | |
| *** woodard has joined #openstack-ansible | 12:52 | |
| odyssey4me | jrosser I think logan- and evrardjp have done some digging around that sort of thing. Maybe they can suggest some things off the cuff. I'm tied up in some other work atm and nothing springs to mind. | 12:54 |
| jrosser | ok, np | 12:54 |
| odyssey4me | My only thought is perhaps to do something like set a local fact with the result of that eval on the deploy host, then use that? | 12:54 |
| hwoarang | would someone like to get https://review.openstack.org/#/c/534764/ in which also fixes linters? :/ | 12:56 |
| *** sxc731 has joined #openstack-ansible | 12:56 | |
| evrardjp | hwoarang: done | 12:57 |
| hwoarang | awesome | 12:57 |
| evrardjp | jrosser: mmm , I am not sure how we can do that | 12:57 |
| evrardjp | because basically what happens here is two fold | 12:57 |
| evrardjp | first the user_variables location for this is better, because it gets loaded once instead of for every node of the inventory | 12:58 |
| evrardjp | but where is used is a different story | 12:58 |
| evrardjp | env is passed to every task | 12:58 |
| evrardjp | (almost) | 12:59 |
| evrardjp | so maybe our salvation there could be to simplify, and be explicit on which tasks we need to set those vars | 13:00 |
| evrardjp | alternatively do not do pass proxy related on any tasks (setup things for respecting proxy settings on container creation) AND ensuring all tasks that need external connection go through the deploy node which is itself passing through the proxy | 13:02 |
| *** markvoelker has quit IRC | 13:02 | |
| *** markvoelker has joined #openstack-ansible | 13:02 | |
| evrardjp | let me check in more details real quick | 13:03 |
| openstackgerrit | Markos Chandras (hwoarang) proposed openstack/openstack-ansible-openstack_hosts master: tests: Add missing pip_install role dependency https://review.openstack.org/536358 | 13:03 |
| evrardjp | we could also generate it once, to avoid looping over vars. | 13:05 |
| evrardjp | the problem is indeed the size of the var | 13:05 |
| evrardjp | if the content of the var could be simplified, that would make things run faster | 13:06 |
| evrardjp | jrosser: could you do some benchmarks for us? | 13:06 |
| evrardjp | like, generate a static directive whose content is the same as your current one, see if it's better (I don't expect a large performance increase) | 13:07 |
| *** chyka has joined #openstack-ansible | 13:15 | |
| odyssey4me | evrardjp so the file may be loaded once, but it's only evaluated when the variable in question is used | 13:15 |
| odyssey4me | and that evaluation is done every time that var is used, which is why using extra-vars is slow | 13:16 |
| gillesMo | jafeha: I thought it was simpler to user Linux Bridge for flat and VLAN networks. For VXLAN, I thought I have to use OpenvSwitch, but no. I will try to use only OpenvSwitch, to be able to use DVR too. | 13:16 |
| odyssey4me | jrosser evrardjp of course a specific way to work around the issue for OSA only (not ansible in general) would be to make the repo container/host a proxy and only ever apply the proxy settings for upstream proxies to it | 13:18 |
| odyssey4me | but that would seem a messy way of going about it - not a nice way to do things for a speed boost | 13:18 |
| evrardjp | odyssey4me: yes that's exactly what I said | 13:18 |
| jafeha | gillesMo, yeah thats pretty much what we've decided to go for. if you have questions, i might be able to help. | 13:19 |
| evrardjp | odyssey4me: the user_var is the best place to avoid multiplication, but then it still gets used all the time | 13:19 |
| evrardjp | at every single play/task | 13:19 |
| *** chyka has quit IRC | 13:19 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_barbican master: Reinitialize the virtualenv https://review.openstack.org/535823 | 13:19 |
| evrardjp | odyssey4me: tracking tasks that requires env var to be set would be a good exercise | 13:20 |
| evrardjp | jrosser: ^ | 13:20 |
| evrardjp | because we are maybe setting it where we don't need | 13:20 |
| * jrosser returns from computer room and reads scrollback..... | 13:20 | |
| evrardjp | odyssey4me: hwoarang could you vote on https://review.openstack.org/#/c/535788/ ? | 13:20 |
| jrosser | evrardjp: sure we can benchmark things, it's all up running and we've been poking it for a few days | 13:22 |
| jrosser | 13:07 < evrardjp> like, generate a static directive whose content is the same as your current one | 13:22 |
| jrosser | ^ not sure i'm quote following that | 13:22 |
| jrosser | and yes it would be entirely feasable to go through and apply environment: only to to the tasks which require it | 13:23 |
| jrosser | i figured asking here first was a good idea as there are many different ways to slice it | 13:23 |
| evrardjp | what I meant is that, for now, for every play/task, you hit this single globally scoped var, but its content is generated based on host vars. | 13:25 |
| evrardjp | if you'd generate this var in advance, once and for all, for your inventory, into a file user_proxied.yml | 13:26 |
| evrardjp | I wonder how fast would you get by NOT loading host vars every time | 13:26 |
| evrardjp | I guess the change would be not that big, but it's worth testing imo | 13:26 |
| evrardjp | On a scale of 1 to 10, how much clear am I? -5? | 13:27 |
| jrosser | ah ok, well we effectively have tested this as we set no_proxy to a big string and it went quick | 13:27 |
| evrardjp | with the same content? | 13:28 |
| evrardjp | same size of content I mean | 13:28 |
| jrosser | well actually it was longer | 13:28 |
| evrardjp | you mean the string was longer, but the runtime was shorter? | 13:28 |
| jrosser | yeah, the issue is not the length of the string but the complexity of evaluating whats in the {{ }} | 13:29 |
| evrardjp | sounds like we are hitting the hostvars discovery big dict | 13:29 |
| evrardjp | yeah | 13:29 |
| evrardjp | so I think we should do what I just said | 13:29 |
| evrardjp | generate a user_variable with that content | 13:30 |
| evrardjp | I mean a variable into a user_*.yml file, let's call it user_proxies.yml or something | 13:30 |
| jrosser | ok, and the content of that is dependant on the inventory | 13:30 |
| evrardjp | yeah, at first run the inventory is generated | 13:31 |
| evrardjp | we could probably have something in this file that would contain the upstream proxy, and based on that, it would generate the content of the user_proxies.yml file | 13:32 |
| evrardjp | whatever the file name is | 13:32 |
| evrardjp | the only thing is that we shouldn't overlap variable names or we should be smart in file naming to avoid overriding. | 13:33 |
| evrardjp | but I think it would be still smarter to avoid using those vars everywhere | 13:34 |
| jrosser | point me to where the inventory is generated - i'm a bit lost in the code.... | 13:35 |
| evrardjp | I am not sure we should do that in the inventory generation | 13:36 |
| evrardjp | https://github.com/openstack/openstack-ansible/blob/master/inventory/dynamic_inventory.py | 13:36 |
| evrardjp | that is our dynamic inventory | 13:36 |
| evrardjp | jrosser: for my understanding, you have a var that holds your proxy url, and you have other vars, like the deployment env var, whose content has no proxy on all containers, right? | 13:38 |
| jrosser | we use this bit https://github.com/openstack/openstack-ansible/blob/bab51e2967bfa060576d85efaab52ce63d7cfaea/etc/openstack_deploy/user_variables.yml#L134-L138 | 13:39 |
| jrosser | and that is then set everywhere in the playbooks via environment: | 13:39 |
| evrardjp | yeah I know that part :) | 13:39 |
| jrosser | :) | 13:39 |
| evrardjp | so... | 13:40 |
| evrardjp | I don;t like the idea of including this into the inventory because: 1) it makes the inventory more of a black box | 13:41 |
| evrardjp | 2) we are running it every time, which doesn't allow the user to override it for its case. | 13:41 |
| evrardjp | I like the idea of keeping this in a user_*.yml file, because it won't get loaded for all the hosts | 13:41 |
| evrardjp | I mean multiplied by the amount of groups for examples | 13:41 |
| hwoarang | odyssey4me: evrardjp: there is something strange with ansible http://logs.openstack.org/52/535252/4/check/openstack-ansible-upgrade-ubuntu-xenial/fba54c1/job-output.txt.gz#_2018-01-22_12_08_01_406592 | 13:42 |
| evrardjp | of hosts for example* | 13:42 |
| jrosser | ok | 13:42 |
| hwoarang | look at the path on that line and the one after that | 13:42 |
| hwoarang | the 'previous_role' includes a playbook from the current role | 13:42 |
| hwoarang | :( | 13:42 |
| evrardjp | jrosser: so I think the best is, for something at the really beginning of a deploy, we would generate the content of the var deployment_environment_variables | 13:42 |
| *** cmart has joined #openstack-ansible | 13:43 | |
| evrardjp | it would still allow the deployers to edit this var, if it was in a user_*.yml file | 13:43 |
| evrardjp | I'd make it separate though. | 13:43 |
| evrardjp | hwoarang: I know. | 13:43 |
| evrardjp | https://github.com/ansible/ansible/pull/34790 | 13:43 |
| evrardjp | I have to discuss with people in ansible | 13:44 |
| jrosser | evrardjp: yes i think that is best | 13:44 |
| hwoarang | evrardjp: ok thanks! | 13:44 |
| evrardjp | hwoarang: what i did for previous roles, maybe I forgot for galera, is to install ansible 2.3 before installing previous role, then using current version in the current version of the role | 13:45 |
| evrardjp | hwoarang: let me show you that | 13:45 |
| jrosser | so perhaps a generated user_proxy_autogen.yml or something, perhaps at the start of setup_hosts.yml or similar? | 13:45 |
| hwoarang | evrardjp: right that's a good suggestion | 13:45 |
| evrardjp | hwoarang: https://github.com/openstack/openstack-ansible-os_keystone/commit/91b7524603650edbef984fd9fe83e352b90604d2 | 13:45 |
| evrardjp | we should probably extend the pattern to all the roles, but I haven't got the chance to do it yet. | 13:46 |
| *** deus_ex has joined #openstack-ansible | 13:46 | |
| evrardjp | probably worth adding the generation of the venv etc... in the -tests repo, and then use it from all the roles that need it. | 13:46 |
| hwoarang | awesome thanks | 13:47 |
| evrardjp | jrosser: yeah. | 13:47 |
| evrardjp | jrosser: a new playbook for offline or limited connectivity installs | 13:47 |
| *** pedja has quit IRC | 13:47 | |
| evrardjp | you'd have the full inventory view anyway when you run this | 13:48 |
| evrardjp | it sounds better to me. | 13:48 |
| *** deus_ex is now known as pedja | 13:49 | |
| evrardjp | offline-proxy_generation.yml could be the playbook name, generating something like user_proxy_autogen.yml indeed. | 13:49 |
| *** sxc731 has quit IRC | 13:49 | |
| jrosser | evrardjp: ok, we'll take that as a starting point and work something up.... thanks for the imput :) | 13:49 |
| jrosser | *input | 13:49 |
| evrardjp | it would skip the generation if the var is already set for example | 13:49 |
| gillesMo | jafeha: thank you, yes I have already one question. During my fight, it seemed that neutron-openvswitch-agent was searching for a OVS bridge (br-vlan of me) inside the container. if it needs it why openstack-ansible playbook does not create it ? | 13:50 |
| evrardjp | this way it's backwards compatible and avoid overrides or weird surprises | 13:50 |
| evrardjp | hwoarang: that's why the -vv is useful ;) | 13:51 |
| evrardjp | -v wouldn't show you what ansible loads :D | 13:52 |
| hwoarang | :D | 13:52 |
| jafeha | gillesMo, thats a missing piece that needs fixing. you have to create a provider bridge manually. | 13:52 |
| jafeha | gillesMo, and add it to the correct physical interface | 13:52 |
| evrardjp | hwoarang: if you need help on that, I can help you, it seems a prio item | 13:52 |
| evrardjp | it's just that I'd like to start getting our stuff included for ansible 2.5, to avoid shipping things around all the time | 13:53 |
| hwoarang | evrardjp: i think moving the venv script is a good suggestion | 13:53 |
| hwoarang | *moving to -tests | 13:53 |
| evrardjp | yeah we discussed that with odyssey4me and we agreed that if it was needed for another role, we should move to tests :D | 13:54 |
| evrardjp | and boom, you're it! | 13:54 |
| hwoarang | ok im on it | 13:54 |
| gillesMo | jafeha: Ah ! Ok. I wonder if I should not go to Linux bridge everywhere... But no DVR... | 13:55 |
| evrardjp | :D | 13:55 |
| * hwoarang kicks galera | 13:55 | |
| evrardjp | haha | 13:55 |
| *** esberglu has joined #openstack-ansible | 13:55 | |
| evrardjp | gillesMo: so? | 13:55 |
| evrardjp | DVR is not a big deal | 13:55 |
| jafeha | gillesMo, step 8 and 9 in the example config here. https://docs.openstack.org/neutron/pike/admin/deploy-ovs-provider.html | 13:56 |
| evrardjp | it looks great on paper, I admit it. | 13:56 |
| *** ykarel is now known as ykarel|afk | 13:56 | |
| jafeha | gillesMo, that should do the trick | 13:57 |
| gillesMo | evrardjp: Do you mean DVR is not worth the effort ? Right, I don't know how many compute nodes I will have, but it seems weird to direct all the traffic throught network nodes... | 14:00 |
| jafeha | evrardjp, afair the reason for developing dvr was neutrons limitation to run vrouters on the network nodes. and that would be our neutron-server container as far as i can tell. have there been any changes to this situation that i'm not aware of? | 14:00 |
| openstackgerrit | Markos Chandras (hwoarang) proposed openstack/openstack-ansible-tests master: Use a venv with previous ansible version for upgrades https://review.openstack.org/536372 | 14:01 |
| jafeha | evrardjp, exactly what gillesMo is saying from my perspective ;) | 14:02 |
| jafeha | thats why we've deployed it. | 14:02 |
| *** ykarel|afk has quit IRC | 14:03 | |
| *** hamza21 has joined #openstack-ansible | 14:03 | |
| jafeha | evrardjp, talking about ovs-dvr, who do i turn to for meriging it's documentation? spotz? | 14:04 |
| gillesMo | jafeha: Certainly, but I'm trying not to add extra steps to openstack-ansible, Our deployment should be simple. Add a compute node is documented in openstack-ansible. Also, the official doc is for bare metal installation, there is no LXC containers that add some layers... | 14:04 |
| jafeha | https://review.openstack.org/#/c/532798/ | 14:04 |
| odyssey4me | jafeha mind if you tweak the commit message there? you have a -1 because it makes no sense in the patch context | 14:06 |
| odyssey4me | the content looks fine | 14:06 |
| jafeha | gillesMo, you're right that has to be handled by osa. i've contributed a simple aio role for ovs where this bridge handling is covered, but haven't had any time looking at the other roles why this provider bridge isn't being created. | 14:06 |
| odyssey4me | something simple like 'This patch adds documentation content to help users setup OVS-DVR.' will do fine | 14:07 |
| jafeha | odyssey4me, missing my dev maschine right now, but might be able to fix this on launchpad | 14:08 |
| gillesMo | jafeha, is it something like the playbook at the end of that blog post ? https://trumant.github.io/openstack-ansible-openvswitch.html | 14:09 |
| odyssey4me | jafeha never mind then, I'll fix it up if you don't mind? | 14:09 |
| jafeha | gillesMo, exactly. that should be handled upstream | 14:10 |
| jafeha | or in osa itself | 14:10 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_neutron master: Adds Neutron OVS DVR Documentation https://review.openstack.org/532798 | 14:10 |
| jafeha | thank you! | 14:10 |
| evrardjp | gillesMo: jafeha both right, but I doubt you'd reach an issue if you have dedicated deploy nodes before long. | 14:12 |
| evrardjp | dedicated network nodes* | 14:13 |
| evrardjp | simple usage and debugging too :D | 14:13 |
| openstackgerrit | Markos Chandras (hwoarang) proposed openstack/openstack-ansible-galera_server master: Use a venv with previous ansible version for upgrades https://review.openstack.org/536374 | 14:13 |
| jafeha | evrardjp, with that constraint i see your point. | 14:14 |
| evrardjp | well it's not my cloud, you do whatever you want, but if you're afraid of scaling, maybe you should head towards a sdn solution to allow more flexibility, reporting, etc. | 14:14 |
| evrardjp | just my 0.02£ | 14:14 |
| evrardjp | don't get me wrong, I like ovs-dvr, it's fun. | 14:15 |
| evrardjp | I don't feel the need to put that into prod :D | 14:15 |
| odyssey4me | evrardjp ie your 0.00013 BTC ;) | 14:16 |
| evrardjp | HAHA | 14:16 |
| hwoarang | lol | 14:16 |
| mgariepy | lol | 14:16 |
| odyssey4me | oh wait, 0.00010 BTC ... oh wait 0.00017 BTC | 14:16 |
| evrardjp | odyssey4me: it's not wgat it used to be! | 14:16 |
| *** ivve has quit IRC | 14:16 | |
| evrardjp | :p | 14:16 |
| evrardjp | XBT is evil | 14:17 |
| jafeha | evrardjp, i get it now. my considertions were different: we're simply not having any hardware for deploying separate network nodes. dvr is a fine solution until we scale out one day. | 14:17 |
| evrardjp | I'd still go for lxb :p | 14:17 |
| odyssey4me | jafeha hope you got your security waxed, exposing every compute as a router with a public address is quite a thing to be doing | 14:17 |
| jafeha | odyssey4me, thats indeed an issue. | 14:20 |
| jafeha | odyssey4me, mbe i should add that to the docs. | 14:21 |
| evrardjp | odyssey4me: people will start to run away screaming now | 14:22 |
| evrardjp | congratulations! | 14:22 |
| evrardjp | :D | 14:22 |
| jafeha | even though i'm not convinced that makes much of a difference | 14:23 |
| gillesMo | evrardjp: like jafeha, my first deployment is with limited hardware and I try also some kind of hyperconverged infrastructure : Everything on 4 nodes. 3 controlers, 4 computes, also 4 Ceph OSD node on the same nodes... I know I search for problems ! | 14:23 |
| *** bhujay has joined #openstack-ansible | 14:27 | |
| evrardjp | gillesMo: yup. | 14:28 |
| *** ivve has joined #openstack-ansible | 14:28 | |
| evrardjp | You must love your kernel panics do you | 14:28 |
| evrardjp | :D | 14:29 |
| *** ykarel|afk has joined #openstack-ansible | 14:29 | |
| jafeha | talking about convergence and saving money: i know for fact that it is possible to run an entire osa backend (excluding storage and compute) on intel atoms :D | 14:30 |
| gillesMo | evrardjp: none for the moment, but I still not have any instances :-O until I manage to have working network bridges... | 14:30 |
| *** jwitko has joined #openstack-ansible | 14:37 | |
| *** cmart has quit IRC | 14:47 | |
| dcdamien | gillesMo: watch out, somebody can MELT your hyperconverged nodes DOWN :P | 14:52 |
| gillesMo | dcdamien: I'm clearly not convinced about hyperconverged infra for a cloud. If this Lab deployment fails, I will have more weight to go in another direction for production :-) But in this case, I need at least 9 servers | 14:56 |
| evrardjp | dcdamien: I see what you did there :D | 14:57 |
| cloudnull | mornings | 14:58 |
| *** cmart has joined #openstack-ansible | 15:00 | |
| *** sxc731 has joined #openstack-ansible | 15:02 | |
| evrardjp | morning cloudnull | 15:04 |
| *** RandomTech has joined #openstack-ansible | 15:05 | |
| RandomTech | Hello, Does anyome have an example of adding a seperate public network to the configs? | 15:06 |
| RandomTech | public/external* | 15:06 |
| RandomTech | or am i misinterprating the role of a seperate network that has to be the external one? | 15:08 |
| *** cjloader has joined #openstack-ansible | 15:12 | |
| *** cjloader has quit IRC | 15:13 | |
| *** bhujay has quit IRC | 15:13 | |
| evrardjp | RandomTech: generally we do not care about the external network. We do some wiring to make sure the interface arrives to the neutron nodes, but then it's up to the deployer | 15:14 |
| evrardjp | that wiring is in the networks part of the openstack_user_config | 15:15 |
| evrardjp | you can modify it the way you like depending on your use case (vlan or not...) | 15:15 |
| *** cjloader has joined #openstack-ansible | 15:16 | |
| *** cjloader has quit IRC | 15:16 | |
| *** cjloader has joined #openstack-ansible | 15:16 | |
| RandomTech | I understand that the networks are configured there im just not sure how to configure it in my setup as we have two seperate networks. An private one that we want to use for the storage/management/operations of the openstack, and a public one that we want to use for floating Ips/external container access | 15:17 |
| *** hamza21 has quit IRC | 15:21 | |
| gameon | RandomTech, this may help: https://docs.openstack.org/project-deploy-guide/openstack-ansible/latest/app-config-prod.html | 15:21 |
| gameon | You could put your MGMT net on a VLAN or similar if you wanted to restrict access to the hypervisors etc. | 15:21 |
| gillesMo | Ah ! Is there somewhere a playbook or a tag in openstack-ansible, to remove components ? I want to remove everything related to openvswitch (package and bridges) to start clean | 15:21 |
| evrardjp | gillesMo: no I don't think so... It's re-roll generally. We have a playbook to remove containers, but I think in some cases (like yours), it's safer to reroll your nodes | 15:25 |
| *** phalmos has joined #openstack-ansible | 15:25 | |
| dcdamien | RandomTech: You should provide connectivity to this networks on appropriate hosts yourself and include them in `openstack_user_config.yml`. Document mentioned by gameon explains it well. | 15:26 |
| gillesMo | evrardjp: Rhaaaa ! It takes ages ! But I understand. | 15:26 |
| *** phalmos_ has joined #openstack-ansible | 15:27 | |
| *** phalmos has quit IRC | 15:27 | |
| dcdamien | gillesMo: lxc-containers-destroy playbook can be your friend in this situation :) | 15:27 |
| gameon | gillesMo: What evrardjp said, I deployed with OVS but had to re-roll from scratch to use the default LXB | 15:27 |
| gameon | dcdamien: I found that the stuff left on the hosts caused issues and was quicker to rebuild | 15:28 |
| gillesMo | dcdamien: yes for the config inside containers (neutron-agent), but I also have bridge config on the compute nodes, baremetal. That is up to me. | 15:30 |
| gillesMo | gameon: OK, I think I will certainly have to do the same... I need to find at least how to rebuild without destroying my Ceph cluster (as it is on the same nodes...) | 15:32 |
| gameon | gillesMo: Are you using bluestore? | 15:32 |
| *** cjloader has quit IRC | 15:32 | |
| gameon | Whats your disk setup like? | 15:32 |
| *** cjloader has joined #openstack-ansible | 15:33 | |
| openstackgerrit | Markos Chandras (hwoarang) proposed openstack/openstack-ansible-galera_server master: tasks: Drop use_percona_upstream variable variable https://review.openstack.org/535252 | 15:33 |
| gameon | Because AFAIK, (don't quote me on this!!!) - when using Bluestore with collocated journals i.e on the same disk as your OSDs, you should be able to reinstall the OS etc and when you redeploy it will pickup your existing Ceph data... however I could be totally wrong | 15:33 |
| gillesMo | Not yet. I want to deploy Ocata, to see after that how to handle an upgrade to Pike. As I see, bluestore is not handled by openstack-ansible/ceph-ansible integration that I use. | 15:33 |
| gameon | Hmm - I deployed Pike recently with BS | 15:33 |
| gameon | (in a hyperconverged fashion) | 15:34 |
| gillesMo | gameon: I hope there's an upgrade path frome filestore to bluestore | 15:34 |
| gameon | You probably could get rid of OVS if you wanted to, for me it was quicker to rebuild | 15:34 |
| gameon | And I'm not sure, AFAIK there isn't but again, could be wrong, my Ceph knowledge isn't great | 15:35 |
| evrardjp | is there someone with a good knowledge of tempest here? | 15:35 |
| *** sxc731 has quit IRC | 15:36 | |
| *** ivveh has quit IRC | 15:37 | |
| *** eglute has quit IRC | 15:40 | |
| *** tacco has quit IRC | 15:40 | |
| *** portdirect has quit IRC | 15:40 | |
| *** mrhillsman has quit IRC | 15:40 | |
| *** kmARC has quit IRC | 15:40 | |
| *** electrocucaracha has quit IRC | 15:40 | |
| *** mancdaz has quit IRC | 15:40 | |
| *** Adri2000 has quit IRC | 15:40 | |
| *** homerp_ has quit IRC | 15:40 | |
| *** nwonknu has quit IRC | 15:40 | |
| *** ivveh has joined #openstack-ansible | 15:40 | |
| *** homerp has joined #openstack-ansible | 15:40 | |
| *** eglute_s has joined #openstack-ansible | 15:40 | |
| *** tacco_ has joined #openstack-ansible | 15:40 | |
| *** Adri2000 has joined #openstack-ansible | 15:40 | |
| *** Adri2000 has quit IRC | 15:40 | |
| *** Adri2000 has joined #openstack-ansible | 15:40 | |
| *** mancdaz_ has joined #openstack-ansible | 15:40 | |
| *** mancdaz_ is now known as mancdaz | 15:41 | |
| *** portdirect has joined #openstack-ansible | 15:41 | |
| gillesMo | Does someone know for sure if host_bind_override, in a linuxbridge context, has to point to a bridge or an interface on the host ? It's for my external public network, for which I had configured a Linuxbridge br-public on hosts, connected to the appropriate VLAN (bond0.46). Should I use br-public or bond0.46 for host_bind_override ? | 15:41 |
| *** kmARC has joined #openstack-ansible | 15:41 | |
| *** mrhillsman has joined #openstack-ansible | 15:41 | |
| *** electrocucaracha has joined #openstack-ansible | 15:41 | |
| cloudnull | gillesMo: host_bind_override is generally only needed when using a flat type network | 15:44 |
| cloudnull | **flat type neutron network | 15:44 |
| *** hw_wutianwei has joined #openstack-ansible | 15:45 | |
| *** nwonknu has joined #openstack-ansible | 15:45 | |
| gillesMo | cloudnull: Yes, my br-public is flat (even if underneath, it's a VLAN) | 15:45 |
| gameon | Does anyone know how I can identify which minor version of each component I'm running? Now I've done 16.0.5 to 16.0.6 | 15:45 |
| cloudnull | linux bridge will create bridges from an interface and it can't bridge a bridge | 15:45 |
| gameon | And I guess as a follow on, how I can clean up 16.0.5 | 15:45 |
| openstackgerrit | Merged openstack/openstack-ansible-specs master: Improve user experience by simplifying the guides https://review.openstack.org/531850 | 15:46 |
| cloudnull | gillesMo: in our example we'd hang a veth off the bridge and point the host_bind_override to one side of the veth | 15:46 |
| cloudnull | which gives neutron the ability to create more bridges from the interface. | 15:47 |
| hwoarang | cloudnull: Hello! could you re-apply your C/W on https://review.openstack.org/#/c/534764/ please? :) | 15:48 |
| cloudnull | hwoarang: yup | 15:50 |
| cloudnull | done | 15:50 |
| cloudnull | gameon: if you've upgraded the environment (re-bootstrapped ansible, run the playbooks) you should be good to go. | 15:51 |
| RandomTech | gameon: Im seeing that lxcbr0 provides external access but im not seeing where to configure that | 15:51 |
| hwoarang | cloudnull: awesome thanks a lot | 15:51 |
| gameon | cloudnull: I have run, all OK. But I see 16.0.5 in the venvs, I'd like to clean them up if possible (no big deal though). But I do want to confirm that 16.0.6 is running which I'm not sure on how to ascertain | 15:51 |
| gameon | RandomTech: lxcbr0 is used for containers to connect to the host AFAIK | 15:52 |
| RandomTech | okay then im still confused on how to tell it which network/interface to use for external network connections | 15:53 |
| gameon | RandomTech: Sorry, my mistake | 15:54 |
| gameon | My config has containers with eth0 (or 1) which connect to lxcbr0 (which runs on the host) | 15:55 |
| *** sxc731 has joined #openstack-ansible | 15:57 | |
| *** shardy has quit IRC | 16:00 | |
| *** bhujay has joined #openstack-ansible | 16:01 | |
| *** shardy has joined #openstack-ansible | 16:02 | |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-os_cinder master: Fix upgrade test https://review.openstack.org/535791 | 16:03 |
| *** ykarel|afk is now known as ykarel | 16:05 | |
| *** chyka has joined #openstack-ansible | 16:06 | |
| gillesMo | RandomTech: That's where I am too. You define the mapping in openstack_user_config.yml, provider_networks. I think the part defining the first br-vlan as flat is an external network in the example config here : https://docs.openstack.org/project-deploy-guide/openstack-ansible/ocata/app-config-test.html | 16:06 |
| RandomTech | i just noticed the lxcbr0 has an ip on a network i never told it | 16:07 |
| RandomTech | thankfully it didnt pick a used network but still | 16:07 |
| gameon | RandomTech: That is what the containers communicate with the host on - see: https://docs.openstack.org/openstack-ansible-lxc_hosts/ocata/ | 16:10 |
| gameon | I'd suggest going through every single component you want to deploy and configuring each one by one | 16:10 |
| *** chyka has quit IRC | 16:11 | |
| cloudnull | gameon: you can confirm the services are executing out of the 16.0.6 via process list | 16:17 |
| cloudnull | and then delete the old venv if so | 16:18 |
| cloudnull | I would check to see if anything is executing our of the old venv, 16.0.5 and if nothing is returned delete it | 16:18 |
| gameon | Thanks cloudnull, a simple ps | aux ? | 16:18 |
| cloudnull | yea. maybe something like `ansible -m shell -a 'ps auxf | grep "16.0.5" || rm -rf "/openstack/venvs/*16.0.5*"'` all | 16:20 |
| cloudnull | ^ totally untested | 16:20 |
| gameon | cloudnull, there I was writing a little bash thing. Awesome, thank you | 16:20 |
| cloudnull | might want to limit the scope of that command a bit but you get the idea. | 16:21 |
| cloudnull | :) | 16:21 |
| cloudnull | RandomTech: yea. the lxcbr0 interface should use a 10.x.y.z/24 network. but it's not plugged into any interfaces by default | 16:21 |
| gameon | cloudnull: I've got rid of the rm bit to manually check any discrepencies. Very useful, thank you | 16:22 |
| cloudnull | it routes traffic from containers out of the box via iptables shinanigans | 16:22 |
| cloudnull | gameon: ++ anytime! | 16:22 |
| cloudnull | gillesMo: https://github.com/openstack/openstack-ansible/blob/master/etc/network/interfaces.d/aio_interfaces.cfg#L44-L59 | 16:24 |
| cloudnull | in that example we're hanging the veth off the bridge | 16:24 |
| cloudnull | so host_bind_override would be set to "eth12" | 16:24 |
| cloudnull | which provides a virtual interface into the bridge | 16:24 |
| cloudnull | and allows neutron to create bridges for flat networks you setup in neutron | 16:25 |
| cloudnull | gillesMo: you said "my br-public is flat (even if underneath, it's a VLAN)" do you need the br-public bridge for other reasons ? | 16:28 |
| cloudnull | do you're hosts need access to the public networks you'll be creating? | 16:29 |
| cloudnull | if no, it may be better just to have the VLAN trunked to the host and setup neutron to use a VLAN type network | 16:30 |
| *** ivve has quit IRC | 16:31 | |
| *** rstarmer has joined #openstack-ansible | 16:32 | |
| gillesMo | cloudnull: for br-public, yes, I use it fot the lb external vip, to access horizon/API endpoints | 16:33 |
| cloudnull | ah ok. | 16:33 |
| cloudnull | so then hanging the veth off the bridge is probably the easiest way forward. | 16:34 |
| openstackgerrit | Merged openstack/openstack-ansible-tests master: Vagrantfile: Use rsync to copy working directory to VM https://review.openstack.org/534764 | 16:34 |
| gillesMo | cloudnull: I don't think we can bridge an interface in several bridges ? If I use the bond0.46 interface that is already in br-public, it won't work ? | 16:35 |
| cloudnull | you can't bridge a bridge, however you can create a bridge using an an interface that's plugged into a bridge. | 16:36 |
| cloudnull | which is what the veth is providing | 16:37 |
| *** rstarmer has quit IRC | 16:37 | |
| cloudnull | host_bind_override is used to set a specific interface to bind to instead of the assumed bridge. | 16:38 |
| evrardjp | cloudnull: well | 16:40 |
| cloudnull | evrardjp: and then :) | 16:41 |
| evrardjp | that definition is not clear to me... a veth is like a cable, so you plug the side to whatever you want, so you can plug two bridges together with a cable | 16:41 |
| evrardjp | cloudnull: haha I laughed. | 16:41 |
| evrardjp | and then? | 16:41 |
| cloudnull | yes | 16:41 |
| evrardjp | haha | 16:41 |
| cloudnull | the host is effectively a switch | 16:41 |
| evrardjp | not sure if it helps people | 16:41 |
| evrardjp | the upgrades are better, I think tomorrow we'll see if everything is fine. I think I should freeze the whole thing for m3. | 16:42 |
| evrardjp | when I have the confirmation it works | 16:42 |
| evrardjp | then we can iterate on the nspawn and others | 16:42 |
| cloudnull | yupie ! | 16:43 |
| cloudnull | anyone want to review https://review.openstack.org/#/c/527592/ | 16:43 |
| evrardjp | yeah, it took a while, but it's better that way for me: less stress, and we can, at the end, iterate faster. | 16:43 |
| evrardjp | cloudnull: don't be too hasteful! | 16:43 |
| evrardjp | tomorrow! | 16:43 |
| cloudnull | haste! | 16:43 |
| evrardjp | more speed less haste, that's the expression? | 16:44 |
| cloudnull | move fast, break things! | 16:44 |
| cloudnull | release often | 16:44 |
| cloudnull | :) | 16:44 |
| evrardjp | or is that hasters are gonna haste? | 16:44 |
| *** ykarel is now known as ykarel|afk | 16:44 | |
| evrardjp | I will -2 for today and remove all the -2 tomorrow. I hope we'll get some positive votes in the meantime | 16:45 |
| cloudnull | :\ | 16:46 |
| cloudnull | might as well -2 everthing then | 16:46 |
| cloudnull | if nothing is going in because of "reasons" then -2 all of the approved patches. | 16:47 |
| *** chyka has joined #openstack-ansible | 16:47 | |
| evrardjp | but it doesn't prevent people to vote positively! | 16:47 |
| evrardjp | that's how I see speed: We just just prepare all the patches, and then, when the -2 is lifted, MERGE PARTY! | 16:48 |
| cloudnull | why hold only my patches? | 16:48 |
| evrardjp | it's not only yours | 16:48 |
| *** ivve has joined #openstack-ansible | 16:48 | |
| cloudnull | it is | 16:48 |
| evrardjp | mine are -2 too | 16:48 |
| evrardjp | I am checking all that's big and not a bug fix | 16:48 |
| openstackgerrit | Merged openstack/openstack-ansible-specs master: Blueprint for Blazar integration with OSA https://review.openstack.org/528567 | 16:49 |
| evrardjp | sometimes some slip through my cracks, like hwoarang 's patch earlier. But it's fine, I can do mistakes | 16:49 |
| evrardjp | :D | 16:49 |
| hwoarang | hmm what patch... | 16:49 |
| evrardjp | the one to change the galera | 16:49 |
| evrardjp | I was concerned | 16:49 |
| evrardjp | I don't know if you recall | 16:49 |
| evrardjp | but anyway | 16:49 |
| evrardjp | it's done | 16:50 |
| cloudnull | that was a clean PR. | 16:50 |
| RandomTech | cloudnull: how should i configure what to use for my external network then? | 16:50 |
| cloudnull | I think we're being overly conservative... | 16:50 |
| gillesMo | Damned dynamic-inventory. I don't know where it finds it, but it always report a br-provider I had configurer once, but is no more in my openstack_user_config file... | 16:50 |
| evrardjp | I can remove the -2 to all those patches, but tomorrow, I am pretty sure I will be spending the day to fix stuff... | 16:50 |
| hwoarang | ah the upgrade one | 16:50 |
| cloudnull | 100+ line change, one line change. its all change which can break things. | 16:50 |
| evrardjp | we are close to m3 | 16:51 |
| RandomTech | does that mean 17 is coming out soon evrardjp ? | 16:51 |
| cloudnull | none of that matters for us, as a deployment project we release late. | 16:51 |
| evrardjp | RandomTech: b3 yes | 16:51 |
| evrardjp | not final | 16:51 |
| cloudnull | RandomTech: what do you need to configure? | 16:52 |
| evrardjp | cloudnull: we've always released on time for the milestones and rc. | 16:52 |
| cloudnull | sure. but if we needed to release later we could. | 16:52 |
| evrardjp | I agree it doesn't fully matter. But I still think it's being a good citizen | 16:52 |
| RandomTech | im needing to specify a interface/network for external use but i cant figure out how | 16:52 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-apt_package_pinning master: Updated from OpenStack Ansible Tests https://review.openstack.org/536408 | 16:53 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-ceph_client master: Updated from OpenStack Ansible Tests https://review.openstack.org/536409 | 16:53 |
| cloudnull | RandomTech: so the cloud is online and you need to create the neutron networks? | 16:53 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-galera_client master: Updated from OpenStack Ansible Tests https://review.openstack.org/536410 | 16:53 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-galera_server master: Updated from OpenStack Ansible Tests https://review.openstack.org/536411 | 16:53 |
| cloudnull | or something else at a lower level. | 16:53 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-haproxy_server master: Updated from OpenStack Ansible Tests https://review.openstack.org/536412 | 16:53 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-lxc_container_create master: Updated from OpenStack Ansible Tests https://review.openstack.org/536413 | 16:53 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-lxc_hosts master: Updated from OpenStack Ansible Tests https://review.openstack.org/536414 | 16:53 |
| RandomTech | would it help if i explain our setup breifly? | 16:53 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-memcached_server master: Updated from OpenStack Ansible Tests https://review.openstack.org/536415 | 16:53 |
| cloudnull | sure. | 16:53 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-openstack_hosts master: Updated from OpenStack Ansible Tests https://review.openstack.org/536416 | 16:54 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-openstack_openrc master: Updated from OpenStack Ansible Tests https://review.openstack.org/536417 | 16:54 |
| cloudnull | Might have to wait for the proposal bot to finish it's run :) | 16:54 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-os_almanach master: Updated from OpenStack Ansible Tests https://review.openstack.org/536418 | 16:54 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-os_aodh master: Updated from OpenStack Ansible Tests https://review.openstack.org/536419 | 16:54 |
| RandomTech | ill type it out and wait | 16:54 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-os_barbican master: Updated from OpenStack Ansible Tests https://review.openstack.org/536420 | 16:54 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-os_ceilometer master: Updated from OpenStack Ansible Tests https://review.openstack.org/518909 | 16:54 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-os_cinder master: Updated from OpenStack Ansible Tests https://review.openstack.org/536421 | 16:54 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-os_cloudkitty master: Updated from OpenStack Ansible Tests https://review.openstack.org/536422 | 16:54 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-os_designate master: Updated from OpenStack Ansible Tests https://review.openstack.org/536423 | 16:54 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-os_glance master: Updated from OpenStack Ansible Tests https://review.openstack.org/536424 | 16:55 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-os_gnocchi master: Updated from OpenStack Ansible Tests https://review.openstack.org/536425 | 16:55 |
| dcdamien | gillesMo: look into /etc/openstack_deploy/openstack_inventory.json | 16:55 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-os_heat master: Updated from OpenStack Ansible Tests https://review.openstack.org/536426 | 16:55 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-os_horizon master: Updated from OpenStack Ansible Tests https://review.openstack.org/536427 | 16:55 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-os_ironic master: Updated from OpenStack Ansible Tests https://review.openstack.org/536428 | 16:55 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-os_keystone master: Updated from OpenStack Ansible Tests https://review.openstack.org/536429 | 16:55 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-os_magnum master: Updated from OpenStack Ansible Tests https://review.openstack.org/536430 | 16:55 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-os_monasca master: Updated from OpenStack Ansible Tests https://review.openstack.org/536431 | 16:55 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-os_neutron master: Updated from OpenStack Ansible Tests https://review.openstack.org/536432 | 16:56 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-os_nova master: Updated from OpenStack Ansible Tests https://review.openstack.org/536433 | 16:56 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-os_octavia master: Updated from OpenStack Ansible Tests https://review.openstack.org/536434 | 16:56 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-os_rally master: Updated from OpenStack Ansible Tests https://review.openstack.org/536435 | 16:56 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-os_sahara master: Updated from OpenStack Ansible Tests https://review.openstack.org/536436 | 16:56 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-os_searchlight master: Updated from OpenStack Ansible Tests https://review.openstack.org/536437 | 16:56 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-os_swift master: Updated from OpenStack Ansible Tests https://review.openstack.org/536438 | 16:56 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-os_tacker master: Updated from OpenStack Ansible Tests https://review.openstack.org/536439 | 16:56 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-os_tempest master: Updated from OpenStack Ansible Tests https://review.openstack.org/536440 | 16:57 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-os_trove master: Updated from OpenStack Ansible Tests https://review.openstack.org/536441 | 16:57 |
| gillesMo | dcdamien: yes, it's there but I thought it will be re-created when we change something in openstack_user_config ? | 16:57 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-os_watcher master: Updated from OpenStack Ansible Tests https://review.openstack.org/536442 | 16:57 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-os_zaqar master: Updated from OpenStack Ansible Tests https://review.openstack.org/536443 | 16:57 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-pip_install master: Updated from OpenStack Ansible Tests https://review.openstack.org/536444 | 16:57 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-plugins master: Updated from OpenStack Ansible Tests https://review.openstack.org/536445 | 16:57 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-rabbitmq_server master: Updated from OpenStack Ansible Tests https://review.openstack.org/536446 | 16:57 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-repo_build master: Updated from OpenStack Ansible Tests https://review.openstack.org/536447 | 16:57 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-repo_server master: Updated from OpenStack Ansible Tests https://review.openstack.org/536448 | 16:57 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-rsyslog_client master: Updated from OpenStack Ansible Tests https://review.openstack.org/536449 | 16:57 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-rsyslog_server master: Updated from OpenStack Ansible Tests https://review.openstack.org/536450 | 16:58 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/ansible-hardening master: Updated from OpenStack Ansible Tests https://review.openstack.org/536451 | 16:58 |
| dcdamien | gillesMo: not sure but according to my knowledge things are appended to inventory file but not deleted except of inventory-manage script | 17:00 |
| *** smatzek has joined #openstack-ansible | 17:02 | |
| RandomTech | There are two networks, A isolated one just for the openstack and a public one we lease an ip range of from the university. Were wanting most of openstack to run on the isolated network, however, we need the containers to have external access through the university network. Internal subnets being(10.35.0.0/16 - storage 10.45.0.0 -container & 10.55.0.0 - tunnel) on one interface. The public network being on a seperate in | 17:02 |
| RandomTech | cloudnull: ^ | 17:02 |
| RandomTech | sorry that i used public/university network interchangably | 17:04 |
| dcdamien | gillesMo: as I know inventory file can't be recreated because - among others - it keeps containers names which are pre-generated | 17:04 |
| cloudnull | ^ gillesMo you might need to edit the openstack_inventory.json file to remove that old network | 17:05 |
| gillesMo | dcdamien, cloudnull: thanks, I see. I don't want to regenerate names and IPs ! So I can edit the inventory file without fear. | 17:07 |
| cloudnull | yes there's backups in the tar | 17:08 |
| *** bhujay has quit IRC | 17:08 | |
| cloudnull | but you can create a copy, edit, run a playbook and it if doesn't all die in a fire, it's good. | 17:08 |
| cloudnull | you can test inventory without a playbook by executing the `dynamic_inventory.py` script | 17:09 |
| *** pcaruana has quit IRC | 17:11 | |
| *** sauloaislan has joined #openstack-ansible | 17:15 | |
| sauloaislan | Morning! | 17:15 |
| gillesMo | Evening ! ;-) | 17:16 |
| openstackgerrit | Merged openstack/openstack-ansible-os_swift master: Fix upgrade test https://review.openstack.org/535788 | 17:16 |
| sauloaislan | Can someone explain to me why this? VIRTUAL_ENV = "/ tmp / openstack-venv-builder / venvs / ironic" | 17:16 |
| sauloaislan | Would that be correct? VIRTUAL_ENV = "/ openstack / venvs / ironic" | 17:17 |
| evrardjp | sauloaislan: what do you mean? | 17:20 |
| evrardjp | they are generated in /tmp/openstack-venv-builder in the repo build process | 17:20 |
| evrardjp | they are then updated in the role using virtualenv-tools (or now the shell script that replaces it) | 17:21 |
| evrardjp | sauloaislan: where do you see that? | 17:21 |
| *** smatzek has quit IRC | 17:22 | |
| *** gameon has quit IRC | 17:22 | |
| sauloaislan | evrardjp: I am trying to install pip packages and they are being installed on /tmp/ | 17:23 |
| *** smatzek has joined #openstack-ansible | 17:23 | |
| sauloaislan | I run export show all variables | 17:23 |
| *** woodard has quit IRC | 17:23 | |
| evrardjp | sauloaislan: which role? | 17:24 |
| evrardjp | which branch? | 17:24 |
| evrardjp | which venv? | 17:24 |
| evrardjp | ... | 17:24 |
| sauloaislan | evrardjp: Master | 17:24 |
| evrardjp | ok, which role? | 17:25 |
| sauloaislan | in ironic(host) and neutron_server venvs | 17:26 |
| evrardjp | neutron is pending | 17:26 |
| evrardjp | https://review.openstack.org/#/c/535813/ | 17:26 |
| evrardjp | ironic should be fine | 17:26 |
| *** smatzek_ has joined #openstack-ansible | 17:27 | |
| evrardjp | sauloaislan: check your last ironic role version? | 17:27 |
| evrardjp | the last patch in: https://github.com/openstack/openstack-ansible-os_ironic/commits/master should have fixed that | 17:27 |
| RandomTech | any idea what i may need to do cloudnull ? | 17:28 |
| *** smatzek has quit IRC | 17:28 | |
| *** smatzek_ has quit IRC | 17:33 | |
| *** smatzek has joined #openstack-ansible | 17:35 | |
| sauloaislan | evrardjp: as roles are outdated, thks !! | 17:35 |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-os_swift master: Reinitialize the virtualenv https://review.openstack.org/535814 | 17:36 |
| *** hamza21 has joined #openstack-ansible | 17:36 | |
| evrardjp | sauloaislan: haha no worries. Neutron is not fixed yet. It needs a few patches to go in first. | 17:36 |
| evrardjp | oh no, just that one patch | 17:37 |
| *** openstacking_123 has joined #openstack-ansible | 17:37 | |
| sauloaislan | evrardjp: great, if you need help testing, just talk. | 17:39 |
| evrardjp | I trust zuul | 17:42 |
| sauloaislan | evrardjp: haha ok | 17:45 |
| *** jbadiapa_ has joined #openstack-ansible | 17:46 | |
| d34dh0r53 | is /etc/openstack_deploy/group_vars deprecated? | 17:48 |
| *** sxc731 has quit IRC | 17:51 | |
| d34dh0r53 | nvm, think I found what I was looking for | 17:52 |
| *** RandomTech has quit IRC | 17:56 | |
| *** shardy has quit IRC | 17:59 | |
| *** mbuil has quit IRC | 17:59 | |
| *** gillesMo has quit IRC | 18:00 | |
| *** rmcall has joined #openstack-ansible | 18:05 | |
| *** jbadiapa_ has quit IRC | 18:07 | |
| *** woodard has joined #openstack-ansible | 18:18 | |
| *** woodard has quit IRC | 18:22 | |
| *** woodard has joined #openstack-ansible | 18:23 | |
| *** pcaruana has joined #openstack-ansible | 18:32 | |
| *** DimGR has quit IRC | 18:41 | |
| *** DimGR has joined #openstack-ansible | 18:41 | |
| *** ykarel|afk has quit IRC | 18:44 | |
| *** esberglu has quit IRC | 18:46 | |
| *** mrtenio has quit IRC | 18:49 | |
| *** esberglu has joined #openstack-ansible | 18:49 | |
| *** stuartgr has quit IRC | 18:52 | |
| *** sxc731 has joined #openstack-ansible | 18:53 | |
| openstackgerrit | Merged openstack/openstack-ansible stable/pike: Update all SHAs for 16.0.7 https://review.openstack.org/535873 | 18:58 |
| *** chyka has quit IRC | 19:00 | |
| *** poopcat has joined #openstack-ansible | 19:01 | |
| openstackgerrit | Antony Messerli proposed openstack/openstack-ansible-ops master: When linking a release, re-bootstrap Ansible https://review.openstack.org/536537 | 19:01 |
| *** openstacking_123 has quit IRC | 19:01 | |
| *** sauloaislan has quit IRC | 19:05 | |
| *** poopcat1 has joined #openstack-ansible | 19:06 | |
| *** poopcat has quit IRC | 19:08 | |
| *** electrofelix has quit IRC | 19:10 | |
| mgariepy | qq, what is the priority for the conf override ? I would think that host_vars overrides would take precedence over user_variables? | 19:14 |
| mgariepy | using newton that is. | 19:14 |
| *** armaan has quit IRC | 19:19 | |
| *** armaan has joined #openstack-ansible | 19:20 | |
| *** rstarmer has joined #openstack-ansible | 19:25 | |
| *** pbandark has quit IRC | 19:26 | |
| logan- | user_variables have precedence over everything | 19:36 |
| logan- | https://docs.ansible.com/ansible/latest/playbooks_variables.html#variable-precedence-where-should-i-put-a-variable scroll down just a bit, user_variables are loaded as "extra vars" | 19:37 |
| mgariepy | hmm, :( | 19:38 |
| openstackgerrit | Antony Messerli proposed openstack/openstack-ansible-ops master: On failure return output of all tasks to be reran https://review.openstack.org/536541 | 19:40 |
| *** pcaruana has quit IRC | 19:43 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_neutron master: Reinitialize the virtualenv https://review.openstack.org/535813 | 19:44 |
| *** dave-mcc_ is now known as dave-mcowan | 19:59 | |
| *** dave-mcowan is now known as dave-mccowan | 19:59 | |
| *** dave-mccowan is now known as dave-mcc_ | 20:00 | |
| *** hybridpollo has joined #openstack-ansible | 20:03 | |
| *** chyka has joined #openstack-ansible | 20:06 | |
| *** hybridpollo has quit IRC | 20:10 | |
| *** hybridpollo has joined #openstack-ansible | 20:12 | |
| *** rstarmer has quit IRC | 20:26 | |
| *** hw_wutianwei has quit IRC | 20:31 | |
| *** hw_wutianwei has joined #openstack-ansible | 20:31 | |
| *** smatzek has quit IRC | 20:38 | |
| *** idlemind has quit IRC | 20:54 | |
| *** ivve has quit IRC | 21:03 | |
| *** dave-mcc_ has quit IRC | 21:05 | |
| *** sxc731 has quit IRC | 21:07 | |
| *** hamza21 has quit IRC | 21:10 | |
| *** esberglu has quit IRC | 21:16 | |
| *** esberglu has joined #openstack-ansible | 21:16 | |
| *** esberglu_ has joined #openstack-ansible | 21:19 | |
| *** esberglu has quit IRC | 21:20 | |
| *** dave-mccowan has joined #openstack-ansible | 21:30 | |
| *** strobelight has quit IRC | 21:47 | |
| *** dave-mcc_ has joined #openstack-ansible | 21:52 | |
| *** dave-mccowan has quit IRC | 21:52 | |
| *** jwitko_ has joined #openstack-ansible | 22:02 | |
| cloudnull | RandomTech: sorry was AFK for a bit, still about | 22:03 |
| cloudnull | ? | 22:03 |
| *** openstacking_123 has joined #openstack-ansible | 22:05 | |
| *** jwitko has quit IRC | 22:06 | |
| *** jwitko_ has quit IRC | 22:08 | |
| *** openstackgerrit has quit IRC | 22:18 | |
| *** jwitko has joined #openstack-ansible | 22:31 | |
| *** smatzek has joined #openstack-ansible | 22:38 | |
| *** woodard has quit IRC | 22:42 | |
| *** smatzek has quit IRC | 22:43 | |
| *** openstacking_123 has quit IRC | 23:07 | |
| *** openstacking_123 has joined #openstack-ansible | 23:08 | |
| *** pmannidi has joined #openstack-ansible | 23:09 | |
| *** openstacking_123 has quit IRC | 23:09 | |
| *** gokhan_ has quit IRC | 23:14 | |
| *** mrhillsman has quit IRC | 23:26 | |
| *** mrhillsman has joined #openstack-ansible | 23:26 | |
| *** electrocucaracha has quit IRC | 23:26 | |
| *** kmARC has quit IRC | 23:26 | |
| *** kmARC has joined #openstack-ansible | 23:27 | |
| *** electrocucaracha has joined #openstack-ansible | 23:27 | |
| *** mancdaz has quit IRC | 23:27 | |
| *** Adri2000 has quit IRC | 23:28 | |
| *** dhellmann has quit IRC | 23:28 | |
| *** dhellmann has joined #openstack-ansible | 23:29 | |
| *** mancdaz has joined #openstack-ansible | 23:29 | |
| *** Adri2000 has joined #openstack-ansible | 23:29 | |
| *** Adri2000 has quit IRC | 23:29 | |
| *** Adri2000 has joined #openstack-ansible | 23:29 | |
| *** openstackgerrit has joined #openstack-ansible | 23:45 | |
| openstackgerrit | Antony Messerli proposed openstack/openstack-ansible-ops master: Remove BOOTSTRAP_ANSIBLE_FOLDER variable https://review.openstack.org/536537 | 23:45 |
| *** phalmos_ has quit IRC | 23:46 | |
| *** openstack has quit IRC | 23:55 | |
| *** openstack has joined #openstack-ansible | 23:58 | |
| *** ChanServ sets mode: +o openstack | 23:58 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!