| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-tests stable/wallaby: Update tox.ini to work with 4.0 https://review.opendev.org/c/openstack/openstack-ansible-tests/+/869846 | 08:55 |
|---|---|---|
| opendevreview | Dmitriy Rabotyagov proposed openstack/ansible-role-systemd_networkd stable/wallaby: Handle omitted variables which appear as empty strings https://review.opendev.org/c/openstack/ansible-role-systemd_networkd/+/869842 | 08:56 |
| opendevreview | Merged openstack/openstack-ansible-os_neutron stable/zed: Update OVN northd group documentation https://review.opendev.org/c/openstack/openstack-ansible-os_neutron/+/869834 | 09:10 |
| opendevreview | Merged openstack/ansible-role-systemd_networkd stable/zed: Fix static routes to use Destination rather than Source key https://review.opendev.org/c/openstack/ansible-role-systemd_networkd/+/869835 | 10:18 |
| opendevreview | Merged openstack/ansible-role-systemd_networkd stable/zed: Handle omitted variables which appear as empty strings https://review.opendev.org/c/openstack/ansible-role-systemd_networkd/+/869839 | 10:18 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible master: Allow empty records for container_skel https://review.opendev.org/c/openstack/openstack-ansible/+/869762 | 11:36 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible master: Add is_nest property for container_skel https://review.opendev.org/c/openstack/openstack-ansible/+/869762 | 11:38 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible stable/zed: Revert "Use /healthcheck URI for backends" https://review.opendev.org/c/openstack/openstack-ansible/+/869967 | 11:55 |
| noonedeadpunk | andrewbonney: for fixing galera thing I can propose only run galera playbook before haproxy... Or maybe we should just say to run haproxy as a last step during this upgrade | 11:56 |
| noonedeadpunk | oh, well, maybe I know better option | 12:02 |
| noonedeadpunk | "better" | 12:03 |
| noonedeadpunk | let me check in some aio... | 12:04 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible master: Skip haproxy with setup-infrastructure for upgrades https://review.opendev.org/c/openstack/openstack-ansible/+/869974 | 12:53 |
| noonedeadpunk | ^ maybe this is good enough? | 12:53 |
| andrewbonney | Yes, that was going to be my approach for our production deployments :) | 13:10 |
| noonedeadpunk | well other solution would be to somehow rework our approach with backend configuration, and damiandabrowski is going to look into that (though I'm a bit skeptical due to complexity of what we're doing) | 13:13 |
| opendevreview | Merged openstack/openstack-ansible-tests stable/wallaby: Update tox.ini to work with 4.0 https://review.opendev.org/c/openstack/openstack-ansible-tests/+/869846 | 13:39 |
| damiandabrowski | actually I already wrote this change and now I'm deploying AIO with this, in a next few days I should be able to propose a change | 13:49 |
| damiandabrowski | but it's not really complex | 13:50 |
| damiandabrowski | the main goal is to avoid configuring haproxy services when running haproxy-install.yml, but instead configure them in pre_tasks in each role separately | 13:52 |
| damiandabrowski | it's also fixing our problem with variables scope(now haproxy does not really know for ex. if glance is listening on https or http because it does not have access to glance group vars) | 13:53 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/ansible-role-systemd_mount stable/zed: Fix mount's systemd unit dependency logic https://review.opendev.org/c/openstack/ansible-role-systemd_mount/+/869937 | 14:29 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/ansible-role-systemd_mount stable/yoga: Fix mount's systemd unit dependency logic https://review.opendev.org/c/openstack/ansible-role-systemd_mount/+/869938 | 14:29 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/ansible-role-systemd_mount stable/xena: Fix mount's systemd unit dependency logic https://review.opendev.org/c/openstack/ansible-role-systemd_mount/+/869939 | 14:29 |
| noonedeadpunk | I should have written down list of concerns with changing the approach, as I've already forgotten them :D | 14:41 |
| damiandabrowski | i have some notes, for ex. you were concerned about haproxy-endpoint-manage.yml when endpoint does not exist(yet) | 14:48 |
| damiandabrowski | but haproxy module does not return error i. this case | 14:49 |
| damiandabrowski | jrosser mentioned some possible issue with horizon and letsencrypt but i didn't test horizon test yet | 14:50 |
| noonedeadpunk | aha, ok, that's awesome you noted that :D | 14:50 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/ansible-role-pki master: Use loop labels to suppress output https://review.opendev.org/c/openstack/ansible-role-pki/+/870015 | 15:08 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible master: Add is_nest property for container_skel https://review.opendev.org/c/openstack/openstack-ansible/+/869762 | 15:21 |
| noonedeadpunk | jrosser: I wonder if instead adding variables to mq/service/etc for no-log we jsut leverage loop labels? | 15:29 |
| noonedeadpunk | As I think you can get output of error while keeping password secret. Unless it's run with verbosity | 15:30 |
| noonedeadpunk | I guess I will play with it.... | 15:30 |
| opendevreview | Merged openstack/ansible-role-systemd_networkd stable/yoga: Handle omitted variables which appear as empty strings https://review.opendev.org/c/openstack/ansible-role-systemd_networkd/+/869840 | 16:15 |
| opendevreview | Merged openstack/ansible-role-systemd_networkd stable/yoga: Fix static routes to use Destination rather than Source key https://review.opendev.org/c/openstack/ansible-role-systemd_networkd/+/869836 | 16:22 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_nova master: Fix scheduler track_instance_changes option https://review.opendev.org/c/openstack/openstack-ansible-os_nova/+/870023 | 16:35 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_nova master: Fix scheduler track_instance_changes option https://review.opendev.org/c/openstack/openstack-ansible-os_nova/+/870023 | 17:11 |
| opendevreview | Merged openstack/ansible-role-systemd_networkd stable/xena: Handle omitted variables which appear as empty strings https://review.opendev.org/c/openstack/ansible-role-systemd_networkd/+/869841 | 18:11 |
| spatel | How do i change IP subnet range of AIO deployment? | 18:16 |
| noonedeadpunk | Well, you can't | 18:17 |
| spatel | hmm | 18:17 |
| opendevreview | Merged openstack/openstack-ansible master: Remove "warn" parameter from command module https://review.opendev.org/c/openstack/openstack-ansible/+/869657 | 18:24 |
| prometheanfire | rocky9 switch from libvirtd to virtqemud from what I can tell | 19:26 |
| opendevreview | Matthew Thode proposed openstack/openstack-ansible-ceph_client master: run on rocky-9 https://review.opendev.org/c/openstack/openstack-ansible-ceph_client/+/870027 | 19:40 |
| opendevreview | Matthew Thode proposed openstack/openstack-ansible-os_nova master: run on rocky-9 https://review.opendev.org/c/openstack/openstack-ansible-os_nova/+/870028 | 19:42 |
| opendevreview | Matthew Thode proposed openstack/openstack-ansible-os_nova master: run on rocky-9 https://review.opendev.org/c/openstack/openstack-ansible-os_nova/+/870028 | 19:43 |
| prometheanfire | might need virtnodedevd too | 19:46 |
| jrosser | prometheanfire: why do only re-runs fail? | 20:02 |
| jrosser | like why is our rocky9 CI not broken | 20:02 |
| prometheanfire | I'm honestly not certian, what I can tell is the socket/service files for virtsecretd were enabled, and the have a Conflict= line with libvirtd in it | 20:14 |
| prometheanfire | I'm still trying to figure out how those services/sockets were enabled in the first place, since cent/rocky doesn't do that like debian | 20:14 |
| prometheanfire | I wasn't around for the first part of the install, which is when that would have been installed/enabled | 20:15 |
| jrosser | you could ask NeilHanlon about what we should expect to find in Rocky….. | 20:16 |
| NeilHanlon | don't we do upgrade tests on rocky now? | 20:17 |
| jrosser | I think that prometheanfire is suggesting a previous Rocky 9 install did things one way, and rerunning the nova playbook gives different libvirt services | 20:18 |
| jrosser | which would surprise me if there’s been a breaking change there | 20:18 |
| NeilHanlon | ah | 20:18 |
| prometheanfire | my problem is that virtqemud and it's brothers are enabled at the same time libvirtd is enabled (they conflict and starting one will shut the others down) | 20:19 |
| jrosser | I wonder where the packages are coming from | 20:20 |
| prometheanfire | I don't know what enabled virtqemud, that's my biggest question really | 20:20 |
| jrosser | because you may/may not have some RDO or other repos | 20:20 |
| jrosser | I forget how this works on RHEL-slimes | 20:20 |
| jrosser | hah autocorrect :) | 20:20 |
| NeilHanlon | i am guessing it's coming from RDO or something, yeah. because the libvirtd daemon split shouldn't affect RHEL until RHEL 10 at least | 20:20 |
| jrosser | RHEL-alikes, apologies to NeilHanlon there :) | 20:20 |
| prometheanfire | upstream libvirt is breaking out it's monolithic stuff to virt* services/sockets | 20:21 |
| NeilHanlon | RHEL 9 is based on fedora 34, and the libvirtd split happened during fedora 35 | 20:21 |
| jrosser | prometheanfire: I would start by checking what repos you have enabled and if they look correct | 20:22 |
| jrosser | we could easily have an error there picking up master RDO or something | 20:22 |
| NeilHanlon | annnd i'd be wrong https://git.centos.org/rpms/libvirt/blob/c9/f/SPECS/libvirt.spec#_188 | 20:22 |
| prometheanfire | rdo-deps is enabled | 20:22 |
| prometheanfire | rdo for centos9-zed that is | 20:23 |
| NeilHanlon | where do libvirtd and virtqemud say they come from with `dnf list installed`? | 20:23 |
| prometheanfire | NeilHanlon: that mean that redhat-9 vars for the libvirt daemons should switch https://libvirt.org/daemons.html#switching-to-modular-daemons ? | 20:24 |
| prometheanfire | virtqemud is not a package, libvirt-daemon is installed from appstream | 20:25 |
| prometheanfire | virtqemud is installed by libvirt-daemon-driver-qemu | 20:26 |
| prometheanfire | 8.5.0-7.el9_1 for rocky9 | 20:27 |
| * prometheanfire has to copy from another system to this chat | 20:27 | |
| opendevreview | Merged openstack/openstack-ansible-lxc_hosts stable/yoga: Ensure tar is installed on LXC host https://review.opendev.org/c/openstack/openstack-ansible-lxc_hosts/+/868177 | 20:33 |
| opendevreview | Merged openstack/openstack-ansible-haproxy_server stable/zed: Fix dict object key error when haproxy interfaces not defined https://review.opendev.org/c/openstack/openstack-ansible-haproxy_server/+/869843 | 21:02 |
| prometheanfire | NeilHanlon: does that mean that rocky/cent9 needs to have nova enable the virt* services instead of libvirtd? | 21:44 |
| NeilHanlon | I am confirming what (if anything) may have changed. but quite possibly yes | 22:38 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!