| *** yann-kaelig has quit IRC | 00:02 | |
| *** yann-kaelig has joined #openstack-ansible | 00:03 | |
| *** tosky has quit IRC | 00:06 | |
| *** rfolco has quit IRC | 00:12 | |
| *** lemko4 has joined #openstack-ansible | 00:20 | |
| *** lemko has quit IRC | 00:21 | |
| *** lemko4 is now known as lemko | 00:21 | |
| *** yann-kaelig has quit IRC | 00:40 | |
| *** macz_ has quit IRC | 00:56 | |
| *** cshen has joined #openstack-ansible | 01:52 | |
| *** cshen has quit IRC | 01:57 | |
| *** KurtB has joined #openstack-ansible | 03:43 | |
| *** cshen has joined #openstack-ansible | 03:52 | |
| *** cshen has quit IRC | 03:57 | |
| *** KurtB has quit IRC | 04:02 | |
| *** sri_ has quit IRC | 04:07 | |
| *** sri_ has joined #openstack-ansible | 04:11 | |
| *** johnsom has quit IRC | 04:46 | |
| *** johnsom has joined #openstack-ansible | 04:49 | |
| *** fyx has quit IRC | 05:11 | |
| *** fyx has joined #openstack-ansible | 05:14 | |
| *** johnsom has quit IRC | 05:18 | |
| *** johnsom has joined #openstack-ansible | 05:19 | |
| *** evrardjp has quit IRC | 05:33 | |
| *** evrardjp has joined #openstack-ansible | 05:33 | |
| *** cshen has joined #openstack-ansible | 05:53 | |
| *** cshen has quit IRC | 05:57 | |
| *** johnsom has quit IRC | 06:27 | |
| *** johnsom has joined #openstack-ansible | 06:27 | |
| *** SiavashSardari has joined #openstack-ansible | 06:54 | |
| *** cshen has joined #openstack-ansible | 07:00 | |
| *** cshen has quit IRC | 07:04 | |
| *** miloa has joined #openstack-ansible | 07:21 | |
| *** jbadiapa has joined #openstack-ansible | 07:27 | |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible master: Bump SHAs for master https://review.opendev.org/762762 | 07:39 |
|---|---|---|
| openstackgerrit | Siavash Sardari proposed openstack/openstack-ansible-openstack_openrc master: Adding support of system scoped openrc and clouds.yaml https://review.opendev.org/762090 | 07:41 |
| SiavashSardari | is there a problem with gerrit? gitweb links return 404. | 07:43 |
| noonedeadpunk | indeed it does... I think it was working a while ago... | 07:45 |
| openstackgerrit | Siavash Sardari proposed openstack/openstack-ansible-openstack_openrc master: Adding support of system scoped openrc and clouds.yaml https://review.opendev.org/762090 | 07:53 |
| *** macz_ has joined #openstack-ansible | 07:57 | |
| *** macz_ has quit IRC | 08:02 | |
| *** jamesgibo has joined #openstack-ansible | 08:04 | |
| *** rpittau|afk is now known as rpittau | 08:05 | |
| *** andrewbonney has joined #openstack-ansible | 08:17 | |
| *** cshen has joined #openstack-ansible | 08:27 | |
| *** ericzolf has joined #openstack-ansible | 08:29 | |
| *** partlycloudy has quit IRC | 08:35 | |
| *** partlycloudy has joined #openstack-ansible | 08:36 | |
| *** tosky has joined #openstack-ansible | 08:56 | |
| *** nurdie has quit IRC | 09:01 | |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-os_manila master: Updated from OpenStack Ansible Tests https://review.opendev.org/754172 | 09:01 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-os_manila master: Add LXC CI job https://review.opendev.org/705108 | 09:08 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-os_manila master: Add Manila key generation and distribution https://review.opendev.org/705019 | 09:13 |
| *** SiavashSardari has quit IRC | 09:15 | |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible master: [WIP] Adjustments for LXC CI https://review.opendev.org/705107 | 09:21 |
| *** SiavashSardari has joined #openstack-ansible | 09:22 | |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-os_manila master: Add LXC CI job https://review.opendev.org/705108 | 09:22 |
| SiavashSardari | noonedeadpunk this bug should be assigned to you and I think status should change to fix released. https://bugs.launchpad.net/openstack-ansible/+bug/1904029 | 09:33 |
| openstack | Launchpad bug 1904029 in openstack-ansible "Cinder API V2 disabling drives to os_cinder role failure" [Medium,In progress] | 09:33 |
| noonedeadpunk | SiavashSardari: I don't think fix is released for osa.. We need collection to be tagged and updated in galaxy so we could bump collection version here https://opendev.org/openstack/openstack-ansible/src/branch/stable/ussuri/ansible-collection-requirements.yml#L3 | 09:35 |
| noonedeadpunk | But yeah, assigned it to myself, thanks! | 09:36 |
| *** odyssey4me has quit IRC | 09:50 | |
| jamesdenton | Hey noonedeadpunk - QQ for you. I am trying to fix the Calico job, and it turns out that the package on PyPi is not current and not Py3 compatible. Is it possible to install from git instead? https://github.com/openstack/openstack-ansible-os_neutron/blob/master/vars/main.yml#L52 | 09:58 |
| noonedeadpunk | good question | 09:59 |
| noonedeadpunk | and git repo is still https://github.com/projectcalico/networking-calico ? | 10:00 |
| jamesdenton | yes | 10:00 |
| jamesdenton | which also needs to be updated here: https://github.com/openstack/openstack-ansible-os_neutron/blob/master/defaults/main.yml#L60 | 10:01 |
| noonedeadpunk | yeah and it seems that pypi is still from opendev | 10:04 |
| *** jamesdenton has quit IRC | 10:10 | |
| *** jamesdenton has joined #openstack-ansible | 10:10 | |
| jamesdenton | noonedeadpunk i could probably ask for a new release, just wasn't sure if i was overlooking something easy. will be back in a few hrs | 10:13 |
| noonedeadpunk | I think they just don't own this pypi package | 10:13 |
| noonedeadpunk | as it's owned by openstack | 10:13 |
| noonedeadpunk | so it's about transfering ownage of pypi I think | 10:14 |
| jamesdenton | ahh i see, ok | 10:14 |
| noonedeadpunk | I will try to find out the way in the meanwhile | 10:14 |
| jamesdenton | sure, thanks for looking | 10:15 |
| noonedeadpunk | well, I'm thinking if it's already installed from source.... | 10:19 |
| *** jpward has quit IRC | 10:20 | |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible master: Bump calico version https://review.opendev.org/762985 | 10:27 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-os_neutron master: Return calico to voting https://review.opendev.org/702657 | 10:29 |
| openstackgerrit | Merged openstack/openstack-ansible-ops master: Fix repo link in docs https://review.opendev.org/762801 | 10:52 |
| openstackgerrit | Merged openstack/openstack-ansible-os_horizon master: Replace deprecated UPPER_CONSTRAINTS_FILE variable https://review.opendev.org/762055 | 10:58 |
| openstackgerrit | Georgina Shippey proposed openstack/openstack-ansible-galera_server master: Fix to mariadb backup script https://review.opendev.org/762990 | 10:59 |
| *** MickyMan77 has quit IRC | 11:00 | |
| openstackgerrit | Merged openstack/openstack-ansible stable/ussuri: Disable manila 2 vms tempest scenario https://review.opendev.org/761806 | 11:31 |
| openstackgerrit | Merged openstack/openstack-ansible master: Move nspawn to unmaintained state https://review.opendev.org/760954 | 11:31 |
| *** macz_ has joined #openstack-ansible | 11:33 | |
| *** macz_ has quit IRC | 11:38 | |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible master: Fix serialized playbook runs https://review.opendev.org/752040 | 11:45 |
| *** Gimmix has joined #openstack-ansible | 11:47 | |
| Gimmix | hello everyone . I have a question relating to uploading packages on murano using horizon. The error I get : When I use the upload button on horizon I get : [wsgi:error] [pid 18891:tid 140077317805824] /openstack/venvs/horizon-21.1.0/lib/python3.8/site-packages/muranodashboard/packages/views.py:72: UnhashableKeyWarning: The key of | 11:49 |
| Gimmix | openstack_dashboard.api.glance glanceclient is not hashable and cannot be memoized: ((<weakref at 0x7f6635229900; to 'WSGIRequest' at 0x7f6635408b50>,), (('version', '2'),))I can't find anything relating to this on IRC anyone has hit this issue ?I'm using ussuri | 11:49 |
| Gimmix | I'm running ussuri version of openstack-ansible | 11:49 |
| Gimmix | thank you in advance | 11:50 |
| *** spatel has joined #openstack-ansible | 11:56 | |
| *** cshen_ has joined #openstack-ansible | 11:58 | |
| jrosser | Gimmix: this looks like a similar thing in horizon https://bugs.launchpad.net/horizon/+bug/1839295 | 11:59 |
| openstack | Launchpad bug 1839295 in OpenStack Dashboard (Horizon) "UnhashableKeyWarning: The key of openstack_dashboard .api.nova novaclient is not hashable and cannot be memoized" [Low,Confirmed] | 11:59 |
| *** spatel has quit IRC | 12:01 | |
| *** cshen has quit IRC | 12:01 | |
| *** ianychoi_ has quit IRC | 12:10 | |
| *** yann-kaelig has joined #openstack-ansible | 12:11 | |
| Gimmix | jrosser thank you for your pointer. I had seen this previously . This errors refers to nova and suggests that removing the It seems to suggest removing @memoized resolves this | 12:13 |
| Gimmix | I tried this on : /openstack/venvs/horizon-21.1.0/lib/python3.8/dist-packages/openstack_dashboard/api/glance.py | 12:14 |
| Gimmix | there are to references to @memoized which I removed. The error persists just not mentioning memoized | 12:14 |
| Gimmix | I think something is broken on the horizon murano deployment. The playbook activates /openstack/venvs/horizon-21.1.0/lib/python3.8/site-packages/openstack_dashboard/local/enabled/_50_murano.py | 12:16 |
| jrosser | well this really would be a murano bug, OSA can only deploy what is in the murano-dashboard code | 12:16 |
| Gimmix | which from what it read is obsolete. Removing this file the playbook finishes | 12:16 |
| Gimmix | yeah that's true | 12:16 |
| jrosser | unless we pick a wrong version of the dashboard by accident perhaps | 12:17 |
| jrosser | thats always worth checking | 12:17 |
| *** jpward has joined #openstack-ansible | 12:20 | |
| jrosser | Gimmix: actually i think it might be an issue with OSA not controlling the version of murano-dashboard correctly | 12:21 |
| Gimmix | this is from default role | 12:22 |
| Gimmix | murano_dashboard_git_repo: https://opendev.org/openstack/murano-dashboard#murano_dashboard_git_install_branch: "{{ horizon_git_track_branch }}" | 12:22 |
| Gimmix | I have tried changing it and I'm not having much luck which verifies what you just said | 12:23 |
| *** rfolco has joined #openstack-ansible | 12:26 | |
| jrosser | there are probably two things really | 12:26 |
| jrosser | we don't pin/track the version very well which i can fix | 12:27 |
| jrosser | but there quite likley still some underlying issue in the murano-dashboard code | 12:27 |
| *** rfolco has quit IRC | 12:28 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible master: Pin SHA of murano-dashboard so it is controlled by OSA releases https://review.opendev.org/763002 | 12:28 |
| openstackgerrit | Merged openstack/openstack-ansible-os_aodh master: Remove duplicated aodh_db_setup_host https://review.opendev.org/761922 | 12:38 |
| admin0 | osa+magnum day2 struggle start :) | 12:38 |
| jamesdenton | noonedeadpunk re: calico, we also need to add etcd3gw to python package list | 12:43 |
| jamesdenton | and i forgot completely about openstack_services.yml | 12:44 |
| jamesdenton | which begs the question... is the pip package list only used for distro install? | 12:44 |
| *** SiavashSardari has quit IRC | 12:47 | |
| noonedeadpunk | jamesdenton: no, it's added to neutron_venv_packages. I think what takes come controll here is constraints as we're providing it here https://github.com/openstack/openstack-ansible-os_neutron/blob/master/defaults/main.yml#L82 | 12:49 |
| noonedeadpunk | *some control | 12:49 |
| *** SiavashSardari has joined #openstack-ansible | 12:49 | |
| jamesdenton | there might be some duplication here, since the package is also listed in neutron_optional_calico_pip_packages. So maybe there is a preference? or one overwrites the other? | 12:49 |
| admin0 | guilhermesp, container_infra_prefix=<USE A REGIRTRY WITHOUT PULL RATE LIMIT> --- any ideas what to use here or what registries are available out there that hold the required k8 packages ? | 12:52 |
| Gimmix | jrosser thank you . I have manually changed this and indeed it built a new package for the murano-dashboard . The problem does persist however . I think something is broken in horizon / murano | 12:55 |
| Gimmix | I will keep trying ! | 12:55 |
| ThiagoCMC | Morning guys! Quick question.. I noticed that OSA ignores the `ceph_origin: distro` only at the Compute Nodes, the Ceph Nodes (mons, osds, etcs) they came from `distro`, but at the Compute Nodes, it came from a third party repo! Which I don't wanna do. Any idea why? | 12:58 |
| noonedeadpunk | ThiagoCMC: https://opendev.org/openstack/openstack-ansible-ceph_client/src/branch/master/defaults/main.yml#L32-L41 | 13:00 |
| *** nurdie has joined #openstack-ansible | 13:02 | |
| *** nurdie has quit IRC | 13:08 | |
| *** alvinstarr has joined #openstack-ansible | 13:16 | |
| *** Gimmix has quit IRC | 13:22 | |
| guilhermesp | admin0: you can use vexxhost's one | 13:30 |
| guilhermesp | container_infra_prefix=registry.public.mtl1.vexxhost.net/magnum/ | 13:30 |
| admin0 | they won't mind ? | 13:31 |
| admin0 | i won't mind hosting a registry if it helps .. have a few servers with lots of bandwidth sitting | 13:32 |
| admin0 | do you know how to make a registry mirror ? | 13:32 |
| guilhermesp | we are using harbor admin0 https://goharbor.io/ | 13:32 |
| guilhermesp | well you can use for now, no problems but if you wanna take control to replicate docker images by yourself, go for your harbor :) | 13:33 |
| *** jamesgibo has quit IRC | 13:41 | |
| ThiagoCMC | noonedeadpunk, thanks! | 13:45 |
| ThiagoCMC | noonedeadpunk, so, just curious, this openstack-ansible-ceph_client role isn't part of ceph-ansible, right? | 13:45 |
| openstackgerrit | James Denton proposed openstack/openstack-ansible-os_neutron master: Return calico to voting https://review.opendev.org/702657 | 13:46 |
| admin0 | ThiagoCMC, what i have seen is more and more operators are now asking for ceph to be deployed independently of osa and osa just uses what is provided .. kolla already removed integrated ceph build from the latest( as per irc) and my guess is osa might also move in that direction someday | 13:48 |
| *** jamesgibo has joined #openstack-ansible | 13:49 | |
| admin0 | guilhermesp, in your command, towards the end is " --coe kubernetes -f value -c uuid" -- is that also a part of the command ? | 13:52 |
| SiavashSardari | in my point of view the integration of ceph and kola is quite different with OSA and ceph. because OSA uses ceph-ansible and also supports to have an independent ceph installation. | 13:52 |
| *** Gimmix has joined #openstack-ansible | 13:53 | |
| *** spatel has joined #openstack-ansible | 13:58 | |
| ThiagoCMC | admin0, this is exactly what I'm trying to do at home! I already deployed Ceph isolated and now, I'm trying to "import" the Ceph configs from OSA subdirs, into my Ceph standalone and do the same but, without OSA. | 14:01 |
| admin0 | if you provide the mons and deploy has keys to ssh to them, it works out of the box | 14:02 |
| *** nurdie has joined #openstack-ansible | 14:02 | |
| ThiagoCMC | Sure! I know... But my "ceph-ansible" machine is difference from the "osa-ansible" machine... :-P | 14:03 |
| openstackgerrit | Merged openstack/openstack-ansible-os_manila master: Updated from OpenStack Ansible Tests https://review.opendev.org/754172 | 14:03 |
| ThiagoCMC | I want them fully isolated | 14:03 |
| admin0 | create the pools and users beforehand .. that way, a single ceph can handle multiple separate osa clusters | 14:03 |
| admin0 | they are fully isolated .. ceph-deploy box is diff from osa-deploy box | 14:03 |
| ThiagoCMC | cool | 14:03 |
| admin0 | only thing is osa-deploy box should be able to ssh to the mons | 14:03 |
| admin0 | to download the auth stuff | 14:03 |
| ThiagoCMC | Hmm Got it | 14:04 |
| *** cshen_ has quit IRC | 14:06 | |
| *** pto has quit IRC | 14:12 | |
| guilhermesp | admin0: no that's just to print me the uuid of the cluster template after creation | 14:16 |
| guilhermesp | actually just -f value -c uuid | 14:17 |
| guilhermesp | you should keep --coe kubernetes | 14:17 |
| *** macz_ has joined #openstack-ansible | 14:20 | |
| *** d34dh0r53 has quit IRC | 14:22 | |
| spatel | jrosser: or noonedeadpunk i checked out latest master branch and trying to build 6 node api/infra controller nodes and getting stuck here - http://paste.openstack.org/show/800097/ | 14:23 |
| *** macz_ has quit IRC | 14:24 | |
| spatel | why journalctl trying to send logs to physical host "Data could not be sent to remote host \"ostack-phx-api-1-1\ " | 14:25 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible master: Make the different methods of integrating with a ceph cluster clearer https://review.opendev.org/763019 | 14:26 |
| admin0 | ThiagoCMC, https://www.openstackfaq.com/openstack-and-ceph/ | 14:26 |
| jrosser | admin0: ThiagoCMC i clarified the documentation (hopefully!) as there are now more options for ceph integration than there used to be https://review.opendev.org/763019 | 14:27 |
| *** d34dh0r53 has joined #openstack-ansible | 14:30 | |
| noonedeadpunk | spatel: it's not journalctl sends, but it's how osa connection plugin works overall - to reach container we enter host and running commands inside them using lxc tooling | 14:33 |
| noonedeadpunk | This tasks is just creating directory with file module iirc | 14:33 |
| spatel | hmm that is something went wrong with my setup-host.yml | 14:33 |
| noonedeadpunk | but the question why it failed connecting to the physical host considering it reached it prevously without any issues is a good one | 14:34 |
| jrosser | really the trouble is here "Could not resolve hostname ostack-phx-log-1: Name or service not known" | 14:34 |
| spatel | look at this http://paste.openstack.org/show/800098/ | 14:35 |
| spatel | what is -undefined- component ? | 14:35 |
| jamesdenton | jrosser mgariepy if you have a moment, can you give your blessing here? https://review.opendev.org/#/c/762661/ | 14:35 |
| spatel | I have found some of my container doesn't have IP address also | 14:35 |
| jrosser | spatel: previously you found a sitation where the inventory_manage tool would segfault | 14:36 |
| jrosser | so i 'fixed' it by making it print '-undefined-' for things that were undefined rather than a stacktrace | 14:36 |
| spatel | jrosser: so its normal ? | 14:37 |
| noonedeadpunk | sounds like worth dropping containers and inventory... | 14:37 |
| spatel | noonedeadpunk: i am going to try that out next, to drop everything | 14:37 |
| jrosser | hold on | 14:37 |
| spatel | k | 14:38 |
| jrosser | this is quite possibly the first task to try to target ostack-phx-log-1_rsyslog_container-d49b5280 | 14:38 |
| jrosser | can you ssh to ostack-phx-log-1 from the deploy host, by hostname | 14:38 |
| noonedeadpunk | but ostack-phx-api-1-1 failed the same way kind of? | 14:38 |
| spatel | ping: ostack-phx-log-1: Name or service not known | 14:38 |
| jrosser | right | 14:38 |
| spatel | i don't have DNS | 14:38 |
| spatel | I did build many time OSA without any DNS, pure IP | 14:39 |
| jrosser | but you see my point | 14:39 |
| spatel | it use /etc/hosts for all its container and host | 14:39 |
| jrosser | the error you are getting is"Could not resolve hostname ostack-phx-log-1: Name or service not known" | 14:39 |
| jrosser | that was in your paste, and for whatever reason is the root cause | 14:40 |
| ThiagoCMC | admin0, jrosser - thanks!!! | 14:40 |
| noonedeadpunk | and what do you have in /etc/hosts? | 14:40 |
| noonedeadpunk | jrosser: hm, I though ansible should use `ansible_host` for connection.... | 14:40 |
| jrosser | well, i think this is mixed up with the connection plugin | 14:40 |
| jrosser | so i'm thinking it's going to take the IP specificed in openstack_user_config.yml for the host? | 14:41 |
| noonedeadpunk | I think it should... | 14:41 |
| spatel | noonedeadpunk: in container /etc/hosts is empty but on infra node i have 100 lines of /etc/hosts entries for each container and host | 14:41 |
| spatel | look like my containers are half baked | 14:41 |
| jrosser | spatel: and in openstack_user_config you gave the IPs for the hosts? | 14:41 |
| spatel | Yes all IP | 14:42 |
| jrosser | you can run any trivial ansible command to test this | 14:42 |
| jrosser | cd /opt/openstack-ansible | 14:42 |
| spatel | http://paste.openstack.org/show/800099/ | 14:43 |
| jrosser | hrm | 14:43 |
| noonedeadpunk | And I'm pretty sure we were ok without hosts file on the deploy host.... | 14:45 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-plugins master: Use privilege escalation to attach to containers https://review.opendev.org/615403 | 14:45 |
| jrosser | spatel: do something like this http://paste.openstack.org/show/800100/ | 14:46 |
| jrosser | with -vv you can see how it resolves the ip for the hostname in the connection plugin | 14:46 |
| noonedeadpunk | wow, I think you need to set env properly as well for this | 14:47 |
| noonedeadpunk | as by default you wont have inventory | 14:48 |
| spatel | jrosser: getting no output but WARNING | 14:49 |
| spatel | ansible -vv ostack-api-phx-1-1 -m debug -a var=inventory_hostname | 14:49 |
| spatel | [WARNING]: Could not match supplied host pattern, ignoring: ostack-api-phx-1-1 | 14:49 |
| noonedeadpunk | provide `-i /opt/openstack-ansible/inventory/dynamic_inventory.py` | 14:50 |
| spatel | http://paste.openstack.org/show/800101/ | 14:50 |
| *** ericzolf has quit IRC | 14:50 | |
| noonedeadpunk | but yeah, you need to set env I think first - source /usr/local/bin/openstack-ansible.rc | 14:52 |
| spatel | noonedeadpunk: i didn't get that command? | 14:52 |
| noonedeadpunk | oh, well then you don't even need to provide inventory:) | 14:52 |
| spatel | noonedeadpunk: you are right as soon as i load (source /usr/local/bin/openstack-ansible.rc) | 14:52 |
| noonedeadpunk | `source /usr/local/bin/openstack-ansible.rc` and then run what jrosser asked | 14:52 |
| -spatel- ostack-phx-infra-1-1 | SUCCESS => { | 14:53 | |
| -spatel- "inventory_hostname": "ostack-phx-infra-1-1" | 14:53 | |
| -spatel- } | 14:53 | |
| openstackgerrit | James Denton proposed openstack/openstack-ansible master: Create OSA-specific tempest resources https://review.opendev.org/733894 | 14:53 |
| noonedeadpunk | I think interesting part is right above that | 14:53 |
| spatel | here is the full output - http://paste.openstack.org/show/800102/ | 14:54 |
| jrosser | noonedeadpunk: i get the inventory automatically so long as i'm in /opt/openstack-ansible | 14:54 |
| noonedeadpunk | wtf | 14:54 |
| openstackgerrit | James Denton proposed openstack/openstack-ansible-ops master: Make ELK Journalbeat seek configuration a variable https://review.opendev.org/744180 | 14:54 |
| noonedeadpunk | jrosser: sorry for stupid question, but how did you did that?:) | 14:55 |
| jrosser | hmm? :) | 14:55 |
| noonedeadpunk | it's default ? O_o | 14:55 |
| noonedeadpunk | lol | 14:55 |
| jrosser | i *think* so? | 14:55 |
| noonedeadpunk | haha | 14:55 |
| jrosser | it depends on the dir i think | 14:55 |
| noonedeadpunk | ok, til | 14:55 |
| jrosser | https://github.com/openstack/openstack-ansible/blob/master/scripts/openstack-ansible.sh#L38 | 14:56 |
| spatel | yes i have noticed many time when i want to run ansible adhoc command then i have to do cd /opt/openstack-ansible/ | 14:57 |
| noonedeadpunk | jrosser: lmfao I just realized /usr/local/bin/ansible is a symlink to /usr/local/bin/openstack-ansible | 14:58 |
| noonedeadpunk | I was just sourcing rc file.... | 14:59 |
| spatel | so what is the verdict of that output | 15:00 |
| spatel | do you want me to drop and re-build again or keep digging in current failure ? | 15:00 |
| noonedeadpunk | verdict is that connection plugin seems ok | 15:01 |
| jrosser | try for the log host now | 15:01 |
| spatel | let me drop log host and rebuild it again | 15:01 |
| jrosser | well if it can't resolve from the deploy host i'd be looking first at the openstack_user_config end first | 15:02 |
| spatel | but you are ok with -unidentified- component ? I don't have that in my lab that make me worry | 15:02 |
| spatel | Let me sanitize my config first to find any typo or issue. Thank for clue | 15:03 |
| *** SiavashSardari has quit IRC | 15:05 | |
| spatel | jrosser: this is normal warning right - [WARNING]: Unable to parse /etc/openstack_deploy/inventory.ini as an inventory source | 15:12 |
| jrosser | thats fine | 15:13 |
| jrosser | you can specify a static inventory there if you wish to | 15:13 |
| spatel | one thing is bothering me, why i am not seeing -undefined- component in my lab box but i can see that in production where i am deploying | 15:15 |
| spatel | something is not right | 15:15 |
| *** miloa has quit IRC | 15:15 | |
| admin0 | from cloud-init log on the kube-master, i was able to get port 9511: Connection timed out line .. moving further | 15:23 |
| kleini | ThiagoCMC: https://docs.openstack.org/openstack-ansible-ceph_client/latest/config-from-file.html <- here is description how to configure OSA without any access to Ceph cluster | 15:24 |
| ThiagoCMC | kleini, awesome!!! :-D | 15:25 |
| spatel | jamesdenton: i have commented on your patch - https://review.opendev.org/#/c/762661/5/vars/redhat.yml | 15:35 |
| noonedeadpunk | spatel: I'm wondering how does test passing then... | 15:36 |
| noonedeadpunk | btw have you noticed that it's not from rdo repo? | 15:37 |
| noonedeadpunk | maybe packages are built differently? | 15:37 |
| spatel | noonedeadpunk: check this out - http://paste.openstack.org/show/800105/ | 15:39 |
| spatel | I used same version of openvswitch which came from nfv repo | 15:40 |
| noonedeadpunk | fair enough | 15:40 |
| spatel | normal OVS deployment doesn't care about socket but if you start deploying dpdk then it need to find socket directory to attach interface | 15:41 |
| noonedeadpunk | you'd rather put -1 then:) | 15:41 |
| spatel | let me do -1 | 15:42 |
| spatel | noonedeadpunk: my patch has that included so not sure how to deal here - https://review.opendev.org/#/c/762729/2/vars/redhat.yml | 15:43 |
| noonedeadpunk | well, we can rebase jamesdenton patch on top of your then | 15:44 |
| spatel | make sense | 15:44 |
| noonedeadpunk | but you need to drop '{{ }}' in https://review.opendev.org/#/c/762729/2/tasks/providers/setup_ovs_dpdk.yml@92 to pass tests | 15:44 |
| noonedeadpunk | inside ternary | 15:44 |
| spatel | oh can i use ternary('vhost_socket_directory_group' ? | 15:45 |
| noonedeadpunk | you can use `ternary(vhost_socket_directory_group, '')` | 15:45 |
| spatel | ok let me do that and amend it | 15:47 |
| noonedeadpunk | you're already in jinja, so variables does not need to be quoted and tagged with jinja agian | 15:47 |
| spatel | +1 | 15:47 |
| openstackgerrit | Satish Patel proposed openstack/openstack-ansible-os_neutron master: Add centos-8 support for ovs-dpdk https://review.opendev.org/762729 | 15:50 |
| jamesdenton | spatel looks like neutron_ovs_socket_path is only referenced by the OVN metadata agent template. That was implemented for Centos7, so not sure if something changed there or not. | 15:52 |
| jamesdenton | but it doesn't have any impact on OVS-proper | 15:52 |
| spatel | jamesdenton: yes it doesn't have any impact but misleading path | 15:53 |
| jamesdenton | that could be true, but would rather see it addressed on a CentOS 8 OVN-related patch | 15:55 |
| spatel | noonedeadpunk: after dropping all container it passed journalctl step and now seeing DNS issue on my containers - http://paste.openstack.org/show/800106/ | 15:55 |
| jamesdenton | which isn't currently tested AFAIT | 15:55 |
| spatel | jamesdenton: soon i am going to start playing with OVN to give it a shot.. hope we can make it stable for OSA | 15:56 |
| admin0 | definitely stuck at: https://cloud:9511/v1/certificates/uuid | 15:56 |
| spatel | jamesdenton: we can live with that for now, no big deal with path | 15:56 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-plugins master: Fix linters https://review.opendev.org/763041 | 15:57 |
| spatel | noonedeadpunk: i found my issue, damn NetworkManager (i hate this daemon most) | 15:58 |
| noonedeadpunk | it may really appear on centos even after you've dropped it by dependency to smth unrelarted... | 15:59 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-plugins master: Fix linters https://review.opendev.org/763041 | 15:59 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-plugins master: Fix linters https://review.opendev.org/763041 | 16:00 |
| noonedeadpunk | #startmeeting openstack_ansible_meeting | 16:00 |
| openstack | Meeting started Tue Nov 17 16:00:40 2020 UTC and is due to finish in 60 minutes. The chair is noonedeadpunk. Information about MeetBot at http://wiki.debian.org/MeetBot. | 16:00 |
| openstack | Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. | 16:00 |
| *** openstack changes topic to " (Meeting topic: openstack_ansible_meeting)" | 16:00 | |
| openstack | The meeting name has been set to 'openstack_ansible_meeting' | 16:00 |
| noonedeadpunk | #topic bug triage | 16:01 |
| *** openstack changes topic to "bug triage (Meeting topic: openstack_ansible_meeting)" | 16:01 | |
| noonedeadpunk | https://bugs.launchpad.net/openstack-ansible/+bug/1904327 | 16:01 |
| openstack | Launchpad bug 1904327 in openstack-ansible "msgpack version causing linuxbridgeagent to fail on ubuntu 18" [Undecided,New] | 16:01 |
| openstackgerrit | James Denton proposed openstack/openstack-ansible-os_neutron master: Test OVS/OVN deployments on CentOS 8 https://review.opendev.org/762661 | 16:01 |
| noonedeadpunk | jamesdenton: have you ever seen smth like this? | 16:01 |
| jamesdenton | can't say i have | 16:02 |
| jrosser | python2 on train? is that right? | 16:03 |
| noonedeadpunk | I think stdout from stein | 16:03 |
| jamesdenton | i see related bug here: https://bugs.launchpad.net/oslo.privsep/+bug/1844822 | 16:03 |
| openstack | Launchpad bug 1844822 in oslo.privsep "linuxbridge agent crash after R ->S upgrade" [Undecided,New] | 16:03 |
| noonedeadpunk | it really is | 16:04 |
| jamesdenton | same OP | 16:04 |
| *** pcaruana has joined #openstack-ansible | 16:04 | |
| *** pcaruana has quit IRC | 16:04 | |
| jrosser | i guess the best we can do is prod oslo to see if they can patch it? | 16:06 |
| noonedeadpunk | ok, so now it's max_buffer_size | 16:07 |
| noonedeadpunk | https://msgpack-python.readthedocs.io/en/latest/api.html#msgpack.Unpacker | 16:08 |
| jrosser | that matches up with this https://bugs.launchpad.net/oslo.privsep/+bug/1844822/comments/5 | 16:09 |
| openstack | Launchpad bug 1844822 in oslo.privsep "linuxbridge agent crash after R ->S upgrade" [Undecided,New] | 16:09 |
| noonedeadpunk | but default is 1024*1024*100 | 16:09 |
| noonedeadpunk | Ok, well, I can bump ofc version of msgpack but it's in u-c so the only chance to do that is to use global-requirement-pins | 16:11 |
| noonedeadpunk | however I'm not sure if it's our bug or not | 16:12 |
| noonedeadpunk | I'd rather left it to oslo and try to figure out the best way here to proceed | 16:12 |
| openstackgerrit | James Denton proposed openstack/openstack-ansible-os_neutron master: Test OVS/OVN deployments on CentOS 8 https://review.opendev.org/762661 | 16:13 |
| noonedeadpunk | oh, wait, it's msgpack 1.0 docs | 16:13 |
| *** gshippey has joined #openstack-ansible | 16:13 | |
| noonedeadpunk | but it does not support py2.... | 16:14 |
| noonedeadpunk | ok, so interesting one was https://bugs.launchpad.net/openstack-ansible/+bug/1903846 | 16:17 |
| openstack | Launchpad bug 1903846 in openstack-ansible " libvirt fails to start because --listen parameter is not permitted with systemd activation socket " [Undecided,New] | 16:17 |
| noonedeadpunk | I haven't check it in detail yet | 16:17 |
| noonedeadpunk | I'm wondering what libvirt do we get on systmes... | 16:19 |
| noonedeadpunk | as it's failing since 5.7.0 | 16:20 |
| noonedeadpunk | bionic on train looks good | 16:21 |
| noonedeadpunk | so it's from https://github.com/libvirt/libvirt/commit/3a6a725b8f575890ee6c151ad1f46ea0ceea1f3b | 16:21 |
| jrosser | does this say which OS versions the bug report is for | 16:23 |
| noonedeadpunk | nope, don't see it | 16:24 |
| noonedeadpunk | But focal will have issues for sure | 16:24 |
| noonedeadpunk | libvirt docs are pretty confusing for me..... | 16:25 |
| jrosser | so do we need those socket listener services started/stopped on nova_libvirtd_listen_tcp/tls ansible vars | 16:25 |
| jrosser | systemctl start libvirtd-tls.socket | 16:25 |
| jrosser | that sort of thing | 16:25 |
| noonedeadpunk | and we should not use libvirtd.conf for it anymore | 16:26 |
| jrosser | https://opendev.org/openstack/puppet-nova/src/branch/master/releasenotes/notes/libvirt-systemd-socket-units-f7b0fc3a4f3c5219.yaml | 16:26 |
| noonedeadpunk | but um... how to define address to listen on... | 16:26 |
| noonedeadpunk | well, ok, it's valid and has high priority | 16:27 |
| jrosser | yeah so its config_template and [socket] a.b.c.d:port sort of format | 16:27 |
| noonedeadpunk | yeah, seems like that | 16:28 |
| jrosser | or systemd_service with overrides | 16:28 |
| jrosser | https://www.freedesktop.org/software/systemd/man/systemd.socket.html | 16:28 |
| noonedeadpunk | it was so easier in libvirtd.conf :) | 16:29 |
| noonedeadpunk | also what I'm wondering if we can just move to systemd for previous branches - if this new way won't break pre 5.7 libvirt | 16:30 |
| noonedeadpunk | I think that's most interesting unporcessed ones. | 16:32 |
| jrosser | we should also be close to a victoria RC? | 16:34 |
| noonedeadpunk | #topic ffice hours | 16:35 |
| *** openstack changes topic to "ffice hours (Meeting topic: openstack_ansible_meeting)" | 16:35 | |
| noonedeadpunk | well, regarding rc - we need branching for that and I'm pretty stuck with fixing upgrade jobs for magnum | 16:35 |
| noonedeadpunk | but I've placed beta | 16:35 |
| jrosser | is there something to look at for the magnum problem? | 16:36 |
| noonedeadpunk | https://review.opendev.org/#/c/762821/ | 16:36 |
| noonedeadpunk | So I've got https://review.opendev.org/#/c/758461/ | 16:36 |
| noonedeadpunk | but https://review.opendev.org/#/c/755569/5 still fails | 16:36 |
| noonedeadpunk | and it's not only about magnum - octavia fails the same way | 16:37 |
| noonedeadpunk | they both don't have ability to sent endpoint in tempest plugin | 16:37 |
| noonedeadpunk | magnm fails on tempest https://storage.gra.cloud.ovh.net/v1/AUTH_dcaab5e32b234d56b626f72581e3644c/zuul_opendev_logs_bf0/755569/5/check/openstack-ansible-upgrade-aio_metal-ubuntu-focal/bf0ca28/logs/openstack/aio1-utility/stestr_results.html | 16:38 |
| noonedeadpunk | I'm about to jsut abandon 758461 and set upgrade jobs non-oting for magnum and octavia.... | 16:39 |
| noonedeadpunk | will probably try to spend some little time on fixing.... | 16:39 |
| noonedeadpunk | but really concidering just setting then to nv | 16:39 |
| jrosser | interesting that octavia fails, becasue i have some memory of fixing the tempest plugin | 16:40 |
| jrosser | that was perhaps for no ssl verification rather than endpoint | 16:41 |
| johnsom | We have this: https://github.com/openstack/octavia-tempest-plugin/blob/master/octavia_tempest_plugin/config.py#L48 | 16:42 |
| noonedeadpunk | oh, will try this out now | 16:43 |
| noonedeadpunk | hm, we already have it https://opendev.org/openstack/openstack-ansible/src/branch/master/tests/roles/bootstrap-host/templates/user_variables_octavia.yml.j2#L24 | 16:45 |
| noonedeadpunk | but not during test https://zuul.opendev.org/t/openstack/build/d3e6a58048f944aea338979c65626418/log/logs/etc/host/tempest/tempest.conf.txt#164 | 16:45 |
| noonedeadpunk | interesting | 16:45 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-os_octavia master: Fix linter errors https://review.opendev.org/755576 | 16:46 |
| jrosser | why are those missing? | 16:49 |
| noonedeadpunk | maybe I depended on patch that wasn't rebased on top of the fix | 16:50 |
| *** cshen has joined #openstack-ansible | 16:50 | |
| noonedeadpunk | it was added not so far ago or for octavia we're re-defining | 16:51 |
| noonedeadpunk | https://opendev.org/openstack/openstack-ansible/commit/eae210982469d3fdb190e61acb6d19df5f008374 | 16:51 |
| noonedeadpunk | probably it was more than that | 16:51 |
| noonedeadpunk | when amphora was trying to reach octavia api or smth like this... | 16:52 |
| *** cshen has quit IRC | 16:54 | |
| noonedeadpunk | but from my point this is the major things that stops us at the moment | 16:56 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible master: Added Netplan examples for a Openstack Infrastructure node. https://review.opendev.org/760090 | 16:58 |
| noonedeadpunk | and also we need to get to https://review.opendev.org/#/c/752040/3 one day... | 16:58 |
| noonedeadpunk | as it's really a bug we need to look at | 16:58 |
| noonedeadpunk | #endmeeting | 17:00 |
| *** openstack changes topic to "Launchpad: https://launchpad.net/openstack-ansible || Weekly Meetings: https://wiki.openstack.org/wiki/Meetings/openstack-ansible || Review Dashboard: http://bit.ly/osa-review-board-v2" | 17:00 | |
| openstack | Meeting ended Tue Nov 17 17:00:28 2020 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | 17:00 |
| openstack | Minutes: http://eavesdrop.openstack.org/meetings/openstack_ansible_meeting/2020/openstack_ansible_meeting.2020-11-17-16.00.html | 17:00 |
| openstack | Minutes (text): http://eavesdrop.openstack.org/meetings/openstack_ansible_meeting/2020/openstack_ansible_meeting.2020-11-17-16.00.txt | 17:00 |
| openstack | Log: http://eavesdrop.openstack.org/meetings/openstack_ansible_meeting/2020/openstack_ansible_meeting.2020-11-17-16.00.log.html | 17:00 |
| jrosser | noonedeadpunk: does this help https://github.com/openstack/magnum-tempest-plugin/blob/master/magnum_tempest_plugin/config.py#L44-L45 | 17:01 |
| noonedeadpunk | and get public IP during bootstrap? | 17:01 |
| jrosser | not standard endpoint terminology there had to dig deep in the code, but maybe that will let us override? | 17:01 |
| noonedeadpunk | worth trying | 17:02 |
| noonedeadpunk | oh, well, not public but private | 17:02 |
| noonedeadpunk | yeah, good idea! | 17:02 |
| *** jamesgibo has quit IRC | 17:03 | |
| jrosser | we must already have a var somewhere for the internal endpoint when setting up the service catalog | 17:03 |
| jrosser | though that will be defined inside the os_magnum role defaults though | 17:03 |
| noonedeadpunk | in aio we've hardcoded mgmt ip address anyway | 17:04 |
| jrosser | right, so just need to make it into a url i guess | 17:05 |
| jrosser | that var is used here https://github.com/openstack/magnum-tempest-plugin/blob/master/magnum_tempest_plugin/common/manager.py#L50-L54 | 17:05 |
| jrosser | and i guess that it is this line https://github.com/openstack/magnum-tempest-plugin/blob/master/magnum_tempest_plugin/common/manager.py#L54 | 17:05 |
| jrosser | which is missing endpoint=internalURL/publicURL kind of setting | 17:06 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible stable/ussuri: Fix octavia tempest tests https://review.opendev.org/763048 | 17:16 |
| *** jamesgibo has joined #openstack-ansible | 17:17 | |
| noonedeadpunk | oh, we do bootstrap on ussuri during upgrade jobs | 17:17 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible master: Add magnum tempest URL https://review.opendev.org/763049 | 17:19 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-os_magnum master: Fix linter errors https://review.opendev.org/755569 | 17:20 |
| *** jamesgibo has quit IRC | 17:21 | |
| noonedeadpunk | also would be awesome to get some reviews https://review.opendev.org/#/q/is:open+owner:self+label:Verified :) | 17:22 |
| noonedeadpunk | eventually I'd love to merge adjutant support https://review.opendev.org/#/c/756313/ | 17:22 |
| jrosser | ^ almost that link..... but not quite | 17:22 |
| jrosser | 'self' -> me | 17:22 |
| noonedeadpunk | ahhhh | 17:22 |
| noonedeadpunk | https://review.opendev.org/#/q/is:open+owner:noonedeadpunk+label:Verified | 17:23 |
| jrosser | oh also i forgot to mention, andrewbonney has some opportunity to look at the zun role | 17:24 |
| noonedeadpunk | that is really cool | 17:25 |
| noonedeadpunk | we also have damiandabrowski who tries to look on elk but face that same thing we see in CI. | 17:25 |
| noonedeadpunk | I hope they will find out the way to get elk fixed | 17:26 |
| *** rpittau is now known as rpittau|afk | 17:27 | |
| spatel | noonedeadpunk: so finally we released victoria rc release right? | 17:37 |
| noonedeadpunk | beta | 17:37 |
| noonedeadpunk | not rc | 17:37 |
| noonedeadpunk | there's a big difference under it... | 17:38 |
| spatel | oh ok | 17:38 |
| noonedeadpunk | which means that we don't have victoria branches yet | 17:38 |
| noonedeadpunk | jamesdenton: calico seems jsut working:) https://review.opendev.org/#/c/702657/ | 17:39 |
| spatel | when do you think we will have that one. (not asking for specific date but just trying to tell my team that we are moving toward victoria and skipping ussuri) | 17:39 |
| noonedeadpunk | just needed to bump it version, as until they dropped pbr it was not instalkling at all | 17:39 |
| noonedeadpunk | I want to fix magnum and octavia and branch after that | 17:40 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-os_neutron master: Return calico to voting https://review.opendev.org/702657 | 17:40 |
| *** yann-kaelig has quit IRC | 17:58 | |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible master: Fix networking_nsx_git_track_branch https://review.opendev.org/763063 | 18:00 |
| spatel | noonedeadpunk: after dropping and re-creating container works. | 18:02 |
| noonedeadpunk | spatel: this year defenitely, not sure about if we will be able to land in november, but I'd love to place it asap tbh.... | 18:03 |
| *** andrewbonney has quit IRC | 18:05 | |
| spatel | noonedeadpunk: hmm, if its saying longer time then i can go back to stable/ussuri and later come back to victoria when its stable release | 18:05 |
| spatel | *taking | 18:05 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-os_ironic master: Define condition for the first play host one time https://review.opendev.org/756340 | 18:19 |
| *** cloudnull is now known as kecarter | 18:36 | |
| *** kecarter is now known as cloudnull | 18:36 | |
| *** klamath_atx has quit IRC | 18:36 | |
| ThiagoCMC | Guys, is the tag 22.0.0.0b1 about Victoria? | 18:38 |
| *** d34dh0r53 has quit IRC | 18:43 | |
| *** d34dh0r53 has joined #openstack-ansible | 18:46 | |
| spatel | ThiagoCMC: Yes i believe 22.x.x.0b1 is first beta | 19:03 |
| ThiagoCMC | Wheee! | 19:04 |
| ThiagoCMC | BTW, guys, I just had an idea... The introduction of a new Bridge, let's call it "br-ceph" for now. Today, Ceph Public is mixed with "br-mgmt" but, this is kinda bad when you want to isolate things. So, I'll be deploying a fresh OSA now, with this new bridge and make things separated (I have a 2.5Gbps switch where I want deploy Ceph OSDs and Clients, without VLANs, using MACVLAN actually, for "br-ceph | 19:08 |
| ThiagoCMC | / public" and regular "br-storage / cluster"). | 19:08 |
| *** gyee has joined #openstack-ansible | 19:18 | |
| *** cshen has joined #openstack-ansible | 19:28 | |
| *** jawad_axd has joined #openstack-ansible | 19:36 | |
| admin0 | ThiagoCMC, br-mgmt for ceph is only used by glance as far as i know | 19:37 |
| admin0 | ThiagoCMC .. help me make magnum work.. kubernetes on top of osa | 19:38 |
| spatel | admin0: glance should use br-storage | 19:39 |
| spatel | glance container is attached to br-storage interface | 19:40 |
| spatel | ceph use br-mgmt for mon nodes communication with OSD (control-plane) | 19:41 |
| admin0 | day2 osa magnum + kubernetes failed :) will continue again tomorrow | 19:41 |
| ThiagoCMC | admin0, the br-mgmt is actually used by all Ceph Clients of OSA! According to the examples... Including the Compute Nodes/Libvirt <-> OSDs traffic! That's why I want to isolate this. It will help to also, deploy Ceph independently from OpenStack and linke then together later on | 19:43 |
| ThiagoCMC | admin0, I'll have some Kubernetes lessons this Friday! lol Maybe later! haha! | 19:43 |
| spatel | Oh wait, i just did tcpdump on my br-mgmt and its not talking or doing anything with ceph | 19:43 |
| ThiagoCMC | check your rbd within some instance's libvirt, they might be on top of br-mgmt subnet | 19:44 |
| spatel | my all compute nodes using br-storage for RBD communication | 19:45 |
| ThiagoCMC | What about this: https://github.com/openstack/openstack-ansible/blob/master/etc/openstack_deploy/user_variables.yml.prod-ceph.example#L19 ? | 19:45 |
| spatel | i have br-storage interface on mon/osd/compute nodes | 19:45 |
| ThiagoCMC | I see... Well, the provided example mix Ceph's "public_network" with OSA br-mgmt and this is ... Bad (from my point of view). | 19:46 |
| spatel | This is what i have in my production config | 19:46 |
| -spatel- # Ceph network | 19:46 | |
| -spatel- monitor_address_block: "10.50.0.0/21" | 19:46 | |
| -spatel- public_network: "10.50.0.0/21" | 19:46 | |
| -spatel- cluster_network: "10.51.0.0/21" | 19:46 | |
| spatel | 10.50.0.0 is my br-storage and cluster_network network 10.51.0.0 is my dedicated network for OSD replication | 19:47 |
| spatel | br-mgmt has nothing to do with ceph except just deploy or run playbook | 19:48 |
| ThiagoCMC | Is your "cidr_networks.container" different? | 19:48 |
| ThiagoCMC | Got it | 19:48 |
| ThiagoCMC | That's what I want! | 19:48 |
| ThiagoCMC | cool | 19:48 |
| -spatel- cidr_networks: | 19:48 | |
| -spatel- container: 172.28.0.0/21 | 19:48 | |
| -spatel- tunnel: 172.29.0.0/21 | 19:48 | |
| -spatel- storage: 10.50.0.0/21 | 19:48 | |
| ThiagoCMC | Yeah | 19:48 |
| ThiagoCMC | good one | 19:48 |
| ThiagoCMC | :-P | 19:49 |
| spatel | admin0: what is the issue are you having with kube ? | 19:50 |
| spatel | I had lots of fun with that but they it work out nice | 19:50 |
| djhankb | admin0: could you get VMs to boot at all? | 19:51 |
| spatel | This is my kube go to command, always works for me - http://paste.openstack.org/show/800112/ | 19:52 |
| *** cshen has quit IRC | 20:02 | |
| admin0 | they boot, but in the end they say they are unable to connect to https://cloud:9511: Connection timed out | 20:03 |
| jrosser | do you have a proper cert? | 20:07 |
| jrosser | worth replicating that call with curl from the vm to see what’s happening | 20:08 |
| jrosser | spatel: do you deploy ceph with osa or separate? | 20:13 |
| spatel | with OSA | 20:13 |
| spatel | my mon nodes in lxc containers | 20:13 |
| jrosser | looks like you have a good config there with public_network and not follow the example | 20:15 |
| jrosser | the example doesn’t really look right | 20:15 |
| *** cshen has joined #openstack-ansible | 20:16 | |
| spatel | jrosser: I had lots of issue during first time deployment and finally figured out how to deal with that. totally agreed we should change that example | 20:17 |
| admin0 | i use storage network as both public and cluster network | 20:27 |
| spatel | that is also fine if its not hammering performance. | 20:28 |
| spatel | I kept it in separate interface so performance | 20:29 |
| spatel | in OSA document is there nice list of tags which we can use for quick reference? | 21:05 |
| spatel | i meant ansible tags to run playbook snippets | 21:06 |
| djhankb | I keep getting something similar to this "AuthorizationFailure: resources.kube_masters.resources[0].resources.kube-master-bfv: Authorization failed." when it gets to this part of the build: https://opendev.org/openstack/magnum/src/branch/master/magnum/drivers/k8s_fedora_coreos_v1/templates/kubemaster.yaml#L907 | 21:07 |
| *** cshen has quit IRC | 21:48 | |
| *** ThiagoCMC has quit IRC | 22:05 | |
| *** spatel has quit IRC | 22:31 | |
| *** ThiagoCMC has joined #openstack-ansible | 22:58 | |
| *** cshen has joined #openstack-ansible | 23:00 | |
| *** cshen has quit IRC | 23:04 | |
| *** tosky has quit IRC | 23:58 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!