| *** threestrands has joined #openstack-ansible | 00:09 | |
| *** cshen has joined #openstack-ansible | 00:35 | |
| *** cshen has quit IRC | 00:41 | |
| *** NewJorg has quit IRC | 01:08 | |
| *** spatel has joined #openstack-ansible | 01:09 | |
| *** sshnaidm|afk is now known as sshnaidm|off | 01:34 | |
| *** redrobot has quit IRC | 02:22 | |
| *** cshen has joined #openstack-ansible | 02:25 | |
| *** spatel has quit IRC | 02:27 | |
| *** cshen has quit IRC | 02:30 | |
| *** threestrands has quit IRC | 03:49 | |
| *** cshen has joined #openstack-ansible | 04:26 | |
| *** evrardjp has quit IRC | 04:33 | |
| *** evrardjp has joined #openstack-ansible | 04:33 | |
| *** mrda has quit IRC | 04:58 | |
| *** mrda has joined #openstack-ansible | 05:00 | |
| *** udesale has joined #openstack-ansible | 05:18 | |
| janno | so what would be best practice regarding ceph and upgrade from stein to train? would upgrading ceph manually before the upgrade be a solution? | 05:35 |
|---|---|---|
| *** NewJorg has joined #openstack-ansible | 05:39 | |
| jrosser | noonedeadpunk: what do you think of this? https://opendev.org/openstack/openstack-ansible/src/branch/master/scripts/get-ansible-role-requirements.yml#L77-L93 | 05:40 |
| jrosser | that creates overrides so we find the openstack service repos with file:// in CI so they don't need cloning from opendev.org | 05:41 |
| jrosser | but i am trying to patch this, because the networking-calico repo is now at github https://opendev.org/openstack/openstack-ansible/src/branch/master/playbooks/defaults/repo_packages/openstack_services.yml#L172-L176 | 05:42 |
| jrosser | and it breaks like this ERROR: Command errored out with exit status 128: git clone -q file:///openstack/src/github.com/projectcalico/networking-calico | 05:43 |
| jrosser | janno: Erik has been working on ceph upgrades https://review.opendev.org/#/c/710778/ | 05:44 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible master: Update repo location for networking-calico https://review.opendev.org/731109 | 05:55 |
| *** fghaas has joined #openstack-ansible | 06:18 | |
| CeeMac | morning | 06:26 |
| noonedeadpunk | jrosser: but calico is not the only one on github | 06:29 |
| noonedeadpunk | we should already have smth to cover that | 06:29 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible master: Update repo location for networking-calico https://review.opendev.org/731109 | 06:31 |
| jrosser | noonedeadpunk: it's the only one in openstack_services.yml though from github | 06:31 |
| noonedeadpunk | indeed.... maybe add another select and try to get ones with opendev.org only? Or replace git_repo with opendev.org.... | 06:34 |
| noonedeadpunk | that way we will be able to return gnocchi as well | 06:35 |
| noonedeadpunk | but maybe your approach is also good | 06:36 |
| *** yolanda has joined #openstack-ansible | 06:43 | |
| jrosser | noonedeadpunk: yeah, so it's a choice of putting non-zuul things in their own file, or like you say add another select() | 06:50 |
| jrosser | i quite like adding the select anyway as it makes it more robust | 06:51 |
| noonedeadpunk | tbh change filter is just easier as we may have more stuff in github with time | 06:51 |
| noonedeadpunk | yeah, so we wont have to be bothered with it | 06:52 |
| noonedeadpunk | or here we may just do both things :) | 06:52 |
| noonedeadpunk | or, we can really split this file by services | 06:53 |
| noonedeadpunk | that way we won't include so much variables for each play | 06:53 |
| *** jawad_axd has joined #openstack-ansible | 07:03 | |
| janno | gnocchi comes from github: gnocchi_git_repo: https://github.com/gnocchixyz/gnocchi | 07:06 |
| noonedeadpunk | yeah and its also split int oseparate file | 07:28 |
| *** mgariepy has quit IRC | 07:32 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible master: Only use zuul cloned git repos for code hosted at opendev.org https://review.opendev.org/731451 | 07:37 |
| jrosser | maybe like that | 07:37 |
| *** tosky has joined #openstack-ansible | 07:44 | |
| openstackgerrit | Erik Berg proposed openstack/ansible-role-python_venv_build master: install venv_build_distro_package_list on venv_wheel_build_enable https://review.opendev.org/731453 | 07:47 |
| ebbex | jrosser: I tried to build from stable/train (no tag) last night and keystone failed to build. I'm not sure what's changed since 20.1.0, cause the builds were happening _in_ the keystone containers rather than repo..? It was missing a bunch of packages. Then looking for a solution I found this; https://review.opendev.org/731453 | 07:52 |
| ebbex | Am I missing something? It's been like that since the dawn of time, that couldn't ever have possibly worked or? | 07:53 |
| jrosser | ebbex: the builds should be delegated to the repo server | 07:58 |
| jrosser | and those packages would be installed here https://github.com/openstack/ansible-role-python_venv_build/blob/master/tasks/python_venv_wheel_build.yml#L29-L41 | 08:00 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-lxc_hosts master: Stop installing python2.7 on debian https://review.opendev.org/731457 | 08:05 |
| ebbex | I wonder why mine are not being delegated, have to look at that first i suppose. Is the logic correct in that task I sent a patch for? If venv_wheel_build_enable then only install distro packages? | 08:07 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-lxc_hosts master: Stop installing python2.7 on debian https://review.opendev.org/731457 | 08:07 |
| ebbex | because they should be enabled and installed on the repo servers? | 08:07 |
| jrosser | i suspect that is the flip side of the case that i linked to | 08:07 |
| jrosser | when there is no repo server the wheel build happens locally | 08:07 |
| jrosser | so you need the devel packages locally | 08:08 |
| ebbex | and if builds happen locally that should mean that venv_wheel_build_enable is false? | 08:09 |
| ebbex | Ah yes; venv_wheel_build_enable: "{{ venv_build_host != inventory_hostname }}" | 08:10 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-galera_server master: Bump MariaDB version to 10.4 https://review.opendev.org/717830 | 08:12 |
| jrosser | ebbex: if something has gone strange with your deploy that makes one half of that logic do the wrong thing then you could end up with packages on the repo server but trying to do a local build | 08:13 |
| jrosser | that would be odd, but possible | 08:13 |
| *** arkan has joined #openstack-ansible | 08:19 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_rally master: Do not install development packages on the target https://review.opendev.org/731459 | 08:21 |
| ebbex | Yeah, I can't really wrap my head around it yet. There's no changes to my openstack_deploy config. I'll check out 20.1.0 and bootstrap again and see what happens then. | 08:33 |
| *** ravsingh has joined #openstack-ansible | 08:47 | |
| noonedeadpunk | galera 10.4 is going to pass (finally) | 08:51 |
| noonedeadpunk | jrosser: should I rebase it on top of your focal patch or rebase focal one on top of it? | 08:51 |
| *** mgariepy has joined #openstack-ansible | 08:52 | |
| *** rpittau|afk is now known as rpittau | 08:53 | |
| *** arkan has quit IRC | 08:55 | |
| noonedeadpunk | btw we can't remove no_log from tests - ansible lint is unhappy with it then... So probably should skip lint then or dunno | 08:55 |
| *** ondrejburian has quit IRC | 08:57 | |
| *** arkan has joined #openstack-ansible | 08:57 | |
| *** jawad_ax_ has joined #openstack-ansible | 08:58 | |
| jrosser | noonedeadpunk: it's going to be several hours (at the very best) before we might have a focal CI image without python2 | 09:00 |
| jrosser | and thats blocking the focal patches, so it might be best not to rebase on top of those right now | 09:00 |
| jrosser | if there are focal tests we care about, of course | 09:01 |
| *** jawad_axd has quit IRC | 09:01 | |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-galera_server master: Bump MariaDB version to 10.4 https://review.opendev.org/717830 | 09:09 |
| *** jbadiapa has joined #openstack-ansible | 09:10 | |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-galera_server master: Bump MariaDB version to 10.4 https://review.opendev.org/717830 | 09:11 |
| arkan | hi guys, do you know something about this error: http://paste.openstack.org/show/794087/ ? (I'm using 'train' version, and production openstack_user_config.yml) | 09:12 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible master: Remove /usr/bin/python on debian/ubuntu for CI jobs https://review.opendev.org/731476 | 09:13 |
| arkan | also I see that ceph_osd is on metal (running scripts/inventory-manage.py -l), how can I install it into lxc container (if it is possible) ? | 09:15 |
| noonedeadpunk | I'm not sure ceph-osd in container is good idea as it should have access to you hard drives | 09:19 |
| arkan | noonedeadpunk: I don't know why is throwing this error? does it need a dedicated disks? I have aprox. 430GB free space | 09:21 |
| *** itandops has joined #openstack-ansible | 09:22 | |
| noonedeadpunk | so osds are designed to be a separate disks. However you can probably create sparse file and use loop device for testing purposes | 09:22 |
| arkan | and also there is no mention for such details in the example https://docs.openstack.org/openstack-ansible/train/user/ceph/full-deploy.html | 09:25 |
| arkan | it's using ceph-osd_hosts, ceph-mon_hosts and the user variables | 09:25 |
| jrosser | arkan: just to be completely clear about what is happening here, openstack-ansible is acting as a very thin wrapper around https://github.com/ceph/ceph-ansible | 09:26 |
| jrosser | you have to understand what ceph architecture you are trying to build, and set the appropriate ceph-ansible variables to make that happen | 09:26 |
| arkan | jrosser: ok, so I can check their config vars from ceph-ansible | 09:27 |
| arkan | openstack-ansible is using ceph-ansible | 09:27 |
| CeeMac | jrosser: i see you're doing some work on focal, is U release still reference on bionic or it will become the transition release for focal too? | 09:27 |
| jrosser | yes, and you have to supply those in the relevant host/group/user_variables from the OSA side | 09:27 |
| jrosser | arkan: it is optional, you have have OSA call the ceph-ansible roles to deploy the ceph cluster | 09:28 |
| jrosser | arkan: or you can deploy ceph yourself standalone (using ceph-ansible or whatever else you like) then just point OSA at the ceph monitor IP addresses | 09:28 |
| arkan | jrosser: great, I will play with it and see | 09:28 |
| jrosser | CeeMac: in a super ideal world U would become the transition release | 09:29 |
| CeeMac | but in a pragmatic world, if one was looking to deploy U stick with bionic? | 09:29 |
| jrosser | U will support bionic either way | 09:30 |
| CeeMac | ok | 09:30 |
| jrosser | and you could give it some time and with a following wind do an OS upgrade to focal | 09:30 |
| CeeMac | i'm a bit out of touch with OSA releases, was there a U release cut yet? | 09:31 |
| CeeMac | just trying to prep for a coming project | 09:31 |
| jrosser | not yet... | 09:31 |
| CeeMac | did U get officially release anyhow? I'm so out of touch, and have no idea what day it is most times since lockdown started lol | 09:33 |
| jrosser | U is released for the main components yes, but projects like OSA trail there releases behind | 09:33 |
| jrosser | otherwise it's chicken/egg | 09:33 |
| CeeMac | yes | 09:35 |
| CeeMac | makes sense | 09:35 |
| CeeMac | just trying to play a little mental catchup | 09:35 |
| CeeMac | sorry for the obvious/lazy questions :) | 09:35 |
| *** ravsingh has quit IRC | 09:39 | |
| *** khomesh24 has joined #openstack-ansible | 09:42 | |
| *** gshippey has joined #openstack-ansible | 09:44 | |
| *** ravsingh has joined #openstack-ansible | 09:52 | |
| *** dpaclt has joined #openstack-ansible | 09:57 | |
| *** tosky_ has joined #openstack-ansible | 09:57 | |
| *** tosky is now known as Guest48633 | 09:57 | |
| *** tosky_ is now known as tosky | 09:57 | |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-galera_server master: Bump MariaDB version to 10.4 https://review.opendev.org/717830 | 09:59 |
| *** khomesh24 is now known as kt|away | 10:11 | |
| *** kt|away has quit IRC | 10:15 | |
| *** rpittau is now known as rpittau|bbl | 10:18 | |
| *** itandops has quit IRC | 10:24 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible master: Remove /usr/bin/python on debian/ubuntu for CI jobs https://review.opendev.org/731476 | 10:25 |
| *** itandops has joined #openstack-ansible | 10:31 | |
| *** udesale_ has joined #openstack-ansible | 10:34 | |
| *** udesale has quit IRC | 10:36 | |
| *** sshnaidm|off has quit IRC | 10:55 | |
| noonedeadpunk | arg.... what have I changed that 10.4 fails again.... | 11:00 |
| *** cshen has quit IRC | 11:01 | |
| *** itandops has quit IRC | 11:05 | |
| *** dpaclt has quit IRC | 11:08 | |
| *** itandops has joined #openstack-ansible | 11:14 | |
| *** cshen has joined #openstack-ansible | 11:27 | |
| *** cshen has quit IRC | 11:32 | |
| *** cshen has joined #openstack-ansible | 11:36 | |
| *** mgariepy has quit IRC | 11:38 | |
| *** mgariepy has joined #openstack-ansible | 11:38 | |
| *** NewJorg has quit IRC | 11:39 | |
| *** NewJorg has joined #openstack-ansible | 11:39 | |
| *** mathlin has quit IRC | 11:40 | |
| noonedeadpunk | and only ssl ones | 11:41 |
| openstackgerrit | Marc Gariépy (mgariepy) proposed openstack/openstack-ansible-galera_server master: WIP Centos-8 support https://review.opendev.org/724721 | 11:43 |
| openstackgerrit | Marc Gariépy (mgariepy) proposed openstack/openstack-ansible-lxc_hosts master: [WIP] centos-8 support https://review.opendev.org/728108 | 11:52 |
| *** jawad_ax_ has quit IRC | 11:52 | |
| *** udesale_ has quit IRC | 11:58 | |
| *** udesale_ has joined #openstack-ansible | 11:59 | |
| *** dpaclt has joined #openstack-ansible | 12:02 | |
| *** rh-jelabarre has joined #openstack-ansible | 12:03 | |
| *** rh-jelabarre has quit IRC | 12:05 | |
| *** rpittau|bbl is now known as rpittau | 12:05 | |
| *** rh-jelabarre has joined #openstack-ansible | 12:10 | |
| *** arkan has quit IRC | 12:17 | |
| *** ravsingh has quit IRC | 12:38 | |
| openstackgerrit | amolkahat proposed openstack/openstack-ansible-os_tempest master: Added tempest ironic resources setup. https://review.opendev.org/720705 | 13:02 |
| *** ansmith_ has quit IRC | 13:07 | |
| openstackgerrit | Duncan Martin Walker proposed openstack/openstack-ansible-ops master: Prevent duplicate initial_master_nodes https://review.opendev.org/731524 | 13:08 |
| *** ansmith has joined #openstack-ansible | 13:10 | |
| *** itandops has quit IRC | 13:23 | |
| *** itandops has joined #openstack-ansible | 13:24 | |
| *** dpaclt has quit IRC | 13:59 | |
| *** rpittau is now known as rpittau|brb | 14:18 | |
| *** rpittau|brb is now known as rpittau | 14:29 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-lxc_container_create master: Use a single vars file for all ubuntu releases https://review.opendev.org/731552 | 14:44 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-lxc_hosts master: Add functional tests for ubuntu focal https://review.opendev.org/731251 | 14:45 |
| *** Guest70400 has joined #openstack-ansible | 15:03 | |
| *** Guest70400 is now known as redrobot | 15:05 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-lxc_container_create master: Use a single vars file for all ubuntu releases https://review.opendev.org/731552 | 15:15 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible master: [WIP] Centos-8 support https://review.opendev.org/689629 | 15:22 |
| *** theintern has joined #openstack-ansible | 15:35 | |
| *** theintern has quit IRC | 15:39 | |
| openstackgerrit | Merged openstack/openstack-ansible-ops master: Add override for Auditbeat socket dataset https://review.opendev.org/730754 | 15:42 |
| *** udesale_ has quit IRC | 15:50 | |
| openstackgerrit | Duncan Martin Walker proposed openstack/openstack-ansible-ops master: Ensure that the beat setup role runs when elk_beat_setup=true https://review.opendev.org/729654 | 16:02 |
| *** rpittau is now known as rpittau|afk | 16:08 | |
| noonedeadpunk | jrosser: so 10.4 passed https://review.opendev.org/#/c/717830/ | 16:12 |
| jrosser | oh cool | 16:13 |
| noonedeadpunk | btw, had a talks with mariadb folks (inlcuding montywi) | 16:14 |
| noonedeadpunk | 10.4.13 confirmed has a bug that makes db crash | 16:14 |
| noonedeadpunk | 10.4.14 laready fixed (but not released) | 16:14 |
| jrosser | ok | 16:15 |
| jrosser | you want to merge this? | 16:15 |
| noonedeadpunk | and, they suggested us not removing root@localhost. Like it is impossible till 10.4.13 because of hardcode | 16:15 |
| noonedeadpunk | and instead of creating root@% - create like admin@% | 16:15 |
| noonedeadpunk | regarding merge - I'm not sure what will happen with upgrade jobs, but I think we can try merging it | 16:16 |
| *** cshen has quit IRC | 16:17 | |
| jrosser | the focal CI image is now without python2 | 16:17 |
| jrosser | so that gives, umm, different errors now :) | 16:17 |
| jrosser | this is really close https://review.opendev.org/#/c/731251/ | 16:18 |
| jrosser | it's failing to access /tmp in the container, but they are now created | 16:18 |
| noonedeadpunk | ok, cool. btw, regarding py - like run_tests.sh fails for bionic with missing py2 | 16:18 |
| noonedeadpunk | and weird failures for some other os | 16:19 |
| jrosser | is that because we have a hardcoded python2 somewhere in the bootstrap? | 16:20 |
| * jrosser never really uses run_tests, always do it manually | 16:20 | |
| noonedeadpunk | I didn't find anything specific, but fails really at the begining - like with first playbook run | 16:20 |
| * noonedeadpunk lazy to install bindep, tox, etc | 16:21 | |
| noonedeadpunk | so feels like hardocded ansible_python_interpreter | 16:23 |
| *** cshen has joined #openstack-ansible | 16:23 | |
| jrosser | i think we should try not to set that | 16:24 |
| jrosser | feels like trouble | 16:24 |
| noonedeadpunk | so fails on that https://opendev.org/openstack/openstack-ansible-tests/src/branch/master/test-prepare-keys.yml#L27 | 16:24 |
| jrosser | i wonder if we really need connection: local there | 16:26 |
| noonedeadpunk | there is already connection: local? | 16:29 |
| noonedeadpunk | ah, if we need | 16:29 |
| noonedeadpunk | so, do we merge 10.4 galera? | 16:37 |
| noonedeadpunk | I'm just about to do update freeze | 16:37 |
| noonedeadpunk | so probably let's merge whatever we want today not to backport | 16:38 |
| jrosser | yes we should merge it | 16:44 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-tests master: Do not set alternatives before package is present https://review.opendev.org/731585 | 17:03 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-tests master: Do not set alternatives before package is present https://review.opendev.org/731585 | 17:03 |
| jrosser | noonedeadpunk: do you think we should set the alternatives at all for python? | 17:08 |
| jrosser | for a python3 host ideally /usr/bin/python is not there | 17:08 |
| noonedeadpunk | maybe.... | 17:10 |
| noonedeadpunk | just this cause failure:) | 17:10 |
| jrosser | I’ve been taking /use/bin/python away in the focal containers, well undoing the symlinking of it anyway so n the cache prep | 17:12 |
| noonedeadpunk | we can really just remove alternatives | 17:13 |
| noonedeadpunk | I'm probably ok with that | 17:13 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-galera_server master: WIP Centos-8 support https://review.opendev.org/724721 | 17:20 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible master: [WIP] Centos-8 support https://review.opendev.org/689629 | 17:20 |
| mgariepy | wasn't the ansible_python_interpreter mostly needed for centos 7 selinux stuff at the start ? | 17:34 |
| noonedeadpunk | as appeared, no:( | 17:34 |
| mgariepy | centos8 is no fun. | 17:34 |
| mgariepy | btrfs is not supported anymore, no kernel modules avalaible anywhere for it. | 17:35 |
| noonedeadpunk | I _really_ hope, that once we come to services, this will be the way easier | 17:35 |
| noonedeadpunk | yeah | 17:35 |
| noonedeadpunk | I looking at it right now... | 17:35 |
| mgariepy | elrepo do have the btrfs tools stuff | 17:35 |
| noonedeadpunk | I think.... that we actually don't need btrfs | 17:36 |
| mgariepy | but you can't mount it haha | 17:36 |
| noonedeadpunk | yah, but can't mount | 17:36 |
| mgariepy | lol | 17:36 |
| noonedeadpunk | so https://www.man7.org/linux/man-pages/man1/machinectl.1.html | 17:36 |
| noonedeadpunk | says that we can use just raw images with mbr/gpt | 17:36 |
| noonedeadpunk | I'm going to try this out | 17:36 |
| noonedeadpunk | we won't have quotas, but tbh not sure if we need them? | 17:38 |
| mgariepy | well i guess it depends. but i would rather have it working in a ugly way than not.. | 17:39 |
| noonedeadpunk | can't agree more | 17:40 |
| mgariepy | i don't set quota on my controller but i do use a zpool | 17:40 |
| noonedeadpunk | but I think that won't stop you from using cgroups | 17:40 |
| noonedeadpunk | like at the moment I'm not sure what this btrfs quota affects | 17:40 |
| mgariepy | probably the rootfs for the containers quota i guess | 17:40 |
| mgariepy | not sure how it's supposed to be working tbh, i never used it really. | 17:41 |
| noonedeadpunk | btw I think it's only for machinectl or dunno even | 17:41 |
| noonedeadpunk | Like I was using cgroups memory and cpu quotas | 17:42 |
| *** NewJorg has quit IRC | 17:42 | |
| noonedeadpunk | but I feel it has nothing to do with that | 17:42 |
| noonedeadpunk | https://opendev.org/openstack/openstack-ansible-lxc_hosts/src/branch/master/tasks/lxc_volume.yml#L66 | 17:42 |
| mgariepy | yep cgroup for the system but the FS is other things | 17:42 |
| *** NewJorg has joined #openstack-ansible | 17:43 | |
| noonedeadpunk | btw, I'm not sure that container data is stored there... it feel more like container image storage | 17:43 |
| noonedeadpunk | but tbh not really an expert here... | 17:44 |
| mgariepy | maybe indeed. | 17:44 |
| noonedeadpunk | ok. next step.... mount -t ext4 /var/lib/machines.raw /var/lib/machines works | 17:48 |
| noonedeadpunk | but this systemd service says `mount: /var/lib/machines: wrong fs type, bad option, bad superblock on /dev/loop0, missing codepage or helper program, or other error.` | 17:49 |
| noonedeadpunk | https://opendev.org/openstack/openstack-ansible-lxc_hosts/src/branch/master/templates/var-lib-machines.mount | 17:49 |
| noonedeadpunk | oh | 17:49 |
| noonedeadpunk | sure | 17:49 |
| noonedeadpunk | hurray! | 17:50 |
| mgariepy | veverything works and the whole aoi works now ? :P | 17:51 |
| noonedeadpunk | lol | 17:56 |
| noonedeadpunk | I just mounted ext4 with that service :p | 17:57 |
| mgariepy | lol | 17:57 |
| jrosser | the machinectl stuff is only used during the cache prep | 18:11 |
| jrosser | I had to look at this for focal | 18:11 |
| openstackgerrit | amolkahat proposed openstack/openstack-ansible-os_tempest master: Added tempest ironic resources setup. https://review.opendev.org/720705 | 18:12 |
| jrosser | it dumps /car/cache/lxc/<foo>/rootfs.tgz | 18:12 |
| jrosser | /var grrr | 18:12 |
| jrosser | how that gets there kinda doesn’t matter | 18:12 |
| jrosser | it’s independent of the lxc backing store | 18:13 |
| *** yolanda has quit IRC | 18:22 | |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-lxc_hosts master: [WIP] centos-8 support https://review.opendev.org/728108 | 18:24 |
| noonedeadpunk | argh, it seems that we would really need to take more closer look at the way of lxc containers creation... | 18:26 |
| *** aspiers has quit IRC | 18:31 | |
| openstackgerrit | Andreas Jaeger proposed openstack/openstack-ansible master: Remove retired congress https://review.opendev.org/731594 | 18:33 |
| openstackgerrit | Andreas Jaeger proposed openstack/openstack-ansible master: Remove retired congress https://review.opendev.org/731594 | 18:36 |
| *** NewJorg has quit IRC | 18:37 | |
| *** AJaeger has joined #openstack-ansible | 18:39 | |
| jrosser | noonedeadpunk: given all the time in the world I would swap it all out for LXD :) | 18:40 |
| *** aspiers has joined #openstack-ansible | 18:40 | |
| noonedeadpunk | not sure I really understand why it's better? | 18:41 |
| openstackgerrit | Andreas Jaeger proposed openstack/openstack-ansible master: Remove retired congress https://review.opendev.org/731594 | 18:41 |
| jrosser | image management is all just there | 18:41 |
| noonedeadpunk | ah | 18:41 |
| jrosser | cloud-init | 18:41 |
| AJaeger | ansible cores, not sure what else is needed to fix your jobs with congress now retired. https://review.opendev.org/731594 is the start. Help is welcome. Without change https://review.opendev.org/731594 openstack-ansible-deploy-aio will not run at all! | 18:42 |
| noonedeadpunk | but cloud-init sounds as more complex stuff | 18:42 |
| noonedeadpunk | I guess we need to retire that too https://opendev.org/openstack/openstack-ansible-os_congress | 18:42 |
| AJaeger | noonedeadpunk: yeah, you might want to. | 18:43 |
| jrosser | is it gone totally, even stable branches? | 18:44 |
| AJaeger | jrosser: congress is gone from CI, you can install old versions from pypi | 18:44 |
| AJaeger | jrosser: there was some email a month ago on openstack-discuss | 18:45 |
| jrosser | urgh | 18:45 |
| noonedeadpunk | so we nw need to backport that to all of our stable branches | 18:45 |
| AJaeger | yes. | 18:45 |
| jrosser | not even sure we can install from pypi without some significant work | 18:46 |
| AJaeger | http://lists.openstack.org/pipermail/openstack-discuss/2020-April/014292.html | 18:46 |
| noonedeadpunk | what a timing.... | 18:46 |
| *** LuckyClover has joined #openstack-ansible | 18:47 | |
| jrosser | /o\ I just don’t even want to think about that :( | 18:50 |
| noonedeadpunk | ok, so I'm running freeze again | 18:52 |
| *** gshippey has quit IRC | 18:52 | |
| *** aspiers has quit IRC | 18:57 | |
| mgariepy | jrosser, LXD would be better indeed. | 19:04 |
| *** NewJorg has joined #openstack-ansible | 19:04 | |
| noonedeadpunk | but it's really soooo many work.... | 19:04 |
| mgariepy | lxc-to-lxd haha | 19:05 |
| prometheanfire | is ussuri ready to be tested? | 19:05 |
| noonedeadpunk | however, dunno how much should be spent on centos8 lxc | 19:05 |
| noonedeadpunk | nah | 19:05 |
| prometheanfire | out side of osa gate that is :P | 19:05 |
| *** cshen has quit IRC | 19:07 | |
| noonedeadpunk | we have so many features to be implemented for U lol | 19:08 |
| mgariepy | having multiple distros for the project is nice but it would be kinda a lot less work to have fewer. | 19:09 |
| prometheanfire | heh, ya | 19:10 |
| prometheanfire | train it is then | 19:10 |
| prometheanfire | noonedeadpunk: do you have a tracker? | 19:10 |
| noonedeadpunk | http://etherpad.opendev.org/p/openstack-ansible-ussuri kind of it | 19:10 |
| noonedeadpunk | oh, jrosser, btw, are collections ready for merge? | 19:11 |
| jrosser | bar tempest | 19:12 |
| jrosser | tripleo not ready yet | 19:12 |
| noonedeadpunk | we can leave it as is I think? | 19:12 |
| jrosser | need to bump the version we use of the collection | 19:12 |
| noonedeadpunk | we still can use modules? | 19:12 |
| jrosser | but apart from that good I think | 19:12 |
| openstackgerrit | Marc Gariépy (mgariepy) proposed openstack/openstack-ansible-lxc_hosts master: [WIP] centos-8 support https://review.opendev.org/728108 | 19:13 |
| mgariepy | hrm.. fixing 2 typoes ;p haha | 19:13 |
| noonedeadpunk | ah sure | 19:15 |
| noonedeadpunk | but it still doesn't work:( | 19:15 |
| noonedeadpunk | like machinectl just freeze during unpack | 19:15 |
| noonedeadpunk | guess it don't like ext4 | 19:16 |
| noonedeadpunk | and even plain dir | 19:16 |
| noonedeadpunk | despite it probablu should? https://github.com/systemd/systemd/issues/14130 | 19:16 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible master: Create Ussuri https://review.opendev.org/729258 | 19:18 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-tests master: Add ansible openstack collection https://review.opendev.org/718362 | 19:19 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-tests master: Synchronise move to openstack ansible collection to all roles https://review.opendev.org/718527 | 19:19 |
| noonedeadpunk | ooooops | 19:20 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-tests master: Add ansible openstack collection https://review.opendev.org/718362 | 19:21 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-tests master: Synchronise move to openstack ansible collection to all roles https://review.opendev.org/718527 | 19:22 |
| openstackgerrit | Matthew Thode proposed openstack/openstack-ansible-os_swift stable/train: split templates to work around configparser bug https://review.opendev.org/731601 | 19:22 |
| prometheanfire | I think that's the only thing I have to fix manually to get swift working | 19:23 |
| noonedeadpunk | jrosser: are we leaving OSA_COLLECTIONS_DIR as is as for now? https://review.opendev.org/#/c/718431/3/scripts/bootstrap-ansible.sh | 19:23 |
| noonedeadpunk | prometheanfire: did we backport the rest of them? | 19:24 |
| noonedeadpunk | (probably we are) | 19:24 |
| prometheanfire | noonedeadpunk: I think so | 19:25 |
| prometheanfire | I went though the commit log and it looks right | 19:25 |
| prometheanfire | I'm going to be deploying train in the next day or two | 19:25 |
| prometheanfire | so I'll find out | 19:25 |
| noonedeadpunk | hope to branch ussuri this week, but it's probably not smth that worth deploying on production at the moment | 19:26 |
| prometheanfire | well, the env I'm deploying is a test, prod may be in a week or two | 19:26 |
| mgariepy | dnf is annoying.,, I always type dns.. grr. | 19:29 |
| jrosser | noonedeadpunk: we can simplify that | 19:31 |
| jrosser | sorry forgot to revisit it | 19:32 |
| *** cshen has joined #openstack-ansible | 19:33 | |
| *** spatel has joined #openstack-ansible | 19:35 | |
| *** cshen has quit IRC | 19:38 | |
| prometheanfire | just need one more review for https://review.opendev.org/731601 then I can manually change the sha in role-requirements to test | 19:40 |
| mgariepy | noonedeadpunk, selinux ftw. | 19:43 |
| noonedeadpunk | prometheanfire: as you may see we don't have CI at the moment:( | 19:43 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible master: Create Ussuri https://review.opendev.org/729258 | 19:44 |
| prometheanfire | oh lol, just linters and docs | 19:44 |
| noonedeadpunk | mgariepy: selinux? | 19:44 |
| mgariepy | journalctl -xe | 19:45 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible stable/train: Remove retired congress https://review.opendev.org/731604 | 19:45 |
| mgariepy | checking it a bit. | 19:45 |
| mgariepy | but seems like selinux is preventing the machinctl import stuff. | 19:45 |
| noonedeadpunk | I saw it stuck but didn't realize why | 19:46 |
| noonedeadpunk | yeah, had selinux enforcing.... | 19:46 |
| noonedeadpunk | ok, then it's progress I think | 19:47 |
| mgariepy | almost encouraging. | 19:49 |
| mgariepy | let's port everything to arch . | 19:50 |
| *** fghaas has quit IRC | 19:58 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible master: Add ansible openstack collection https://review.opendev.org/718431 | 19:59 |
| jrosser | hopefully thats right | 19:59 |
| *** cshen has joined #openstack-ansible | 20:03 | |
| *** cshen has quit IRC | 20:07 | |
| mgariepy | machineclt seems to be holding the shell for $REASON | 20:23 |
| mgariepy | in the journal, i get Operation completed successfully | 20:23 |
| *** rgogunskiy has joined #openstack-ansible | 20:23 | |
| noonedeadpunk | dunno............... | 21:01 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible master: Create Ussuri https://review.opendev.org/729258 | 21:28 |
| *** jbadiapa has quit IRC | 21:33 | |
| *** rgogunskiy has quit IRC | 21:38 | |
| *** threestrands has joined #openstack-ansible | 21:45 | |
| *** stingrayza has joined #openstack-ansible | 21:47 | |
| *** also_stingrayza has quit IRC | 21:50 | |
| *** cshen has joined #openstack-ansible | 22:04 | |
| *** cshen has quit IRC | 22:08 | |
| *** waxfire has quit IRC | 22:13 | |
| *** waxfire has joined #openstack-ansible | 22:13 | |
| jrosser | noonedeadpunk: broken things https://zuul.opendev.org/t/openstack/build/4186391d44fe4f1480bf16e03e79a09e/log/logs/host/lxc-cache-prep-commands.log.txt#179 | 22:15 |
| jrosser | i'm not sure how we install ca-certificates when the repo is now https with an untrusted certificate | 22:15 |
| jrosser | we need to adjust the URL that zuul provides for the mirror to be http:// instead of https://, thats the smallest change | 22:38 |
| prometheanfire | looks like I missed https://review.opendev.org/718883 | 22:46 |
| *** itandops has quit IRC | 23:11 | |
| *** tosky has quit IRC | 23:20 | |
| openstackgerrit | Matthew Thode proposed openstack/openstack-ansible-lxc_hosts stable/train: ensure 20listchanges is not in debian container https://review.opendev.org/731625 | 23:30 |
| openstackgerrit | Matthew Thode proposed openstack/openstack-ansible-lxc_hosts stable/train: copy the actual keyring https://review.opendev.org/731626 | 23:32 |
| prometheanfire | hmm lxc_hosts is still failing for train it looks like | 23:34 |
| openstackgerrit | Matthew Thode proposed openstack/openstack-ansible-lxc_hosts master: move listchanges removal to where it's needed https://review.opendev.org/731628 | 23:41 |
| openstackgerrit | Matthew Thode proposed openstack/openstack-ansible-lxc_hosts stable/train: ensure 20listchanges is not in debian container https://review.opendev.org/731625 | 23:43 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!