| *** tosky has quit IRC | 00:01 | |
| *** ianychoi has joined #openstack-ansible | 00:14 | |
| *** macz_ has quit IRC | 00:15 | |
| *** ChiTo has quit IRC | 01:13 | |
| *** tinwood has quit IRC | 02:08 | |
| *** tinwood has joined #openstack-ansible | 02:11 | |
| *** macz_ has joined #openstack-ansible | 02:16 | |
| *** macz_ has quit IRC | 02:21 | |
| *** spatel has joined #openstack-ansible | 02:31 | |
| *** miloa has joined #openstack-ansible | 04:06 | |
| *** gyee has quit IRC | 05:08 | |
| *** evrardjp has quit IRC | 05:20 | |
| *** evrardjp has joined #openstack-ansible | 05:24 | |
| *** maharg101 has joined #openstack-ansible | 05:27 | |
| *** maharg101 has quit IRC | 05:33 | |
| *** evrardjp has quit IRC | 05:33 | |
| *** evrardjp has joined #openstack-ansible | 05:35 | |
| *** spatel has quit IRC | 05:58 | |
| *** pto has joined #openstack-ansible | 06:44 | |
| *** pto has quit IRC | 06:53 | |
| *** pto has joined #openstack-ansible | 06:54 | |
| *** maharg101 has joined #openstack-ansible | 07:29 | |
| *** maharg101 has quit IRC | 07:33 | |
| *** openstackgerrit has quit IRC | 07:47 | |
| noonedeadpunk | ThiagoCMC: yep, that totally should be ips of other participants. Also this this pretty shitty role, as I actually need to find time to make pacemaker-remote working | 08:07 |
|---|---|---|
| *** andrewbonney has joined #openstack-ansible | 08:13 | |
| *** pto_ has joined #openstack-ansible | 08:16 | |
| *** pto_ has quit IRC | 08:16 | |
| *** pto_ has joined #openstack-ansible | 08:17 | |
| *** pto has quit IRC | 08:19 | |
| *** pto_ has quit IRC | 08:22 | |
| *** pto has joined #openstack-ansible | 08:23 | |
| *** maharg101 has joined #openstack-ansible | 08:23 | |
| jrosser | i guess we need https://review.opendev.org/c/openstack/openstack-ansible-os_nova/+/770457 to merge ASAP then backport to victoria | 08:23 |
| jrosser | and possibly any branch with buster jobs..... | 08:24 |
| *** akahat|rover is now known as akahat|lunch | 08:25 | |
| *** rpittau|afk is now known as rpittau | 08:25 | |
| noonedeadpunk | yep, and update V bump as well | 08:31 |
| noonedeadpunk | ThiagoCMC: oh, I see why this has happened... Will fix it in a while | 08:32 |
| *** tosky has joined #openstack-ansible | 08:39 | |
| masterpe | noonedeadpunk: about https://github.com/noonedeadpunk/ansible-pacemaker-corosync/pull/1#discussion_r556338541 I need to dig into this why I choose this to force it as eth1, eventual it is on a container always eth1. | 08:43 |
| noonedeadpunk | I believe that won;'t be the case for some of my setups I guess, where I will be passing extra interface, since for me pacemaker does not run on mgmt interface | 08:45 |
| noonedeadpunk | and thanks for pushing this! and sorry I missed it ;( | 08:46 |
| noonedeadpunk | you should ping me haha | 08:46 |
| masterpe | Also after getting pacemaker-remote into a acceptance environment and after the PR I think I made a config error in pacemaker-remote: "error: Error in connection setup (3108-1833-14): Remote I/O error" | 08:46 |
| noonedeadpunk | I should add CI tests to the repo for sure.... | 08:47 |
| masterpe | noonedeadpunk: I think I try'd to ping you but I think at that time you where not available or sorts. | 08:49 |
| *** openstackgerrit has joined #openstack-ansible | 08:50 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_keystone master: Move keystone pip package from a constraint to a requirement https://review.opendev.org/c/openstack/openstack-ansible-os_keystone/+/770271 | 08:50 |
| * noonedeadpunk ashamed | 08:51 | |
| masterpe | noonedeadpunk: is it wise that this ansible-pacemaker-corosync is on github on your private repo and not within opendev? | 08:52 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_cinder master: Move cinder pip package from a constraint to a requirement https://review.opendev.org/c/openstack/openstack-ansible-os_cinder/+/770272 | 08:52 |
| noonedeadpunk | masterpe: we can move it to opendev later. Its just that I had it and it was working for me | 08:52 |
| noonedeadpunk | we have keepalived repo the same way actually | 08:52 |
| noonedeadpunk | (and not limited to it) | 08:53 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_nova master: Move nova pip package from a constraint to a requirement https://review.opendev.org/c/openstack/openstack-ansible-os_nova/+/770279 | 08:53 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_placement master: Move placement pip package from a constraint to a requirement https://review.opendev.org/c/openstack/openstack-ansible-os_placement/+/770280 | 08:54 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_horizon master: Move horizon pip packages from constraints to requirements https://review.opendev.org/c/openstack/openstack-ansible-os_horizon/+/770283 | 08:54 |
| *** pto_ has joined #openstack-ansible | 08:55 | |
| *** pto_ has quit IRC | 08:55 | |
| *** pto_ has joined #openstack-ansible | 08:56 | |
| admin0 | morning | 08:56 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_tempest master: Move tempest pip package from a constraint to a requirement https://review.opendev.org/c/openstack/openstack-ansible-os_tempest/+/770281 | 08:57 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_neutron master: Move neutron pip packages from constraints to requirements https://review.opendev.org/c/openstack/openstack-ansible-os_neutron/+/770276 | 08:58 |
| *** pto has quit IRC | 08:59 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible master: Update pip/setuptools/wheel to latest version https://review.opendev.org/c/openstack/openstack-ansible/+/770284 | 08:59 |
| *** stduolc has quit IRC | 09:00 | |
| *** stduolc has joined #openstack-ansible | 09:00 | |
| masterpe | admin0: morning | 09:01 |
| noonedeadpunk | masterpe: I psuhed https://github.com/noonedeadpunk/ansible-pacemaker-corosync/pull/2 do you have anywhere to check this out? | 09:03 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_glance master: Move glance pip package from a constraint to a requirement https://review.opendev.org/c/openstack/openstack-ansible-os_glance/+/770546 | 09:04 |
| admin0 | when i do curl curl 172.29.239.13:8776 ( cinder api ip of container) , it works ... but when I do curl internal-vip:8776, it says 503 and my cinder-api is failing .. how do I troubleshoot/fix this ? | 09:04 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible master: Update pip/setuptools/wheel to latest version https://review.opendev.org/c/openstack/openstack-ansible/+/770284 | 09:05 |
| jrosser | you look at haproxy and debug that | 09:05 |
| *** rfolco has joined #openstack-ansible | 09:05 | |
| openstackgerrit | Dmitriy Rabotyagov proposed openstack/openstack-ansible-openstack_hosts master: Fix debian release https://review.opendev.org/c/openstack/openstack-ansible-openstack_hosts/+/770455 | 09:05 |
| admin0 | haproxy has the right entry . | 09:05 |
| *** stduolc1 has joined #openstack-ansible | 09:06 | |
| admin0 | how do i enable haproxy stats ? | 09:06 |
| admin0 | found it | 09:07 |
| admin0 | will enable stats and check again | 09:07 |
| *** stduolc has quit IRC | 09:08 | |
| jrosser | if it says 503 then that is service unavailable, remember the codes are from the perspective of haproxy | 09:10 |
| jrosser | so it means "as haproxy i cannot contact the backend for this service@ | 09:10 |
| jrosser | so that should give plenty if hint about where to start | 09:10 |
| masterpe | noonedeadpunk: I will test | 09:16 |
| *** dasp_ has quit IRC | 09:16 | |
| *** dasp has joined #openstack-ansible | 09:17 | |
| noonedeadpunk | I will try to test with aio as well at least... | 09:17 |
| admin0 | jrosser, i think i found a bug .. here is the screenshot -- https://snipboard.io/geVuJ8.jpg -- when haproxy is "re-started", it allows the curl to work ( bottom right of the screen) .. when I run the cinder playbook again, i see that it does not open the MAINT mode .. so it just tries to call the API while the service is still in MAINT mode .. -- this is happening on all osa + ceph above 21.0.1 | 09:30 |
| *** pto_ has quit IRC | 09:30 | |
| *** pto has joined #openstack-ansible | 09:30 | |
| *** pto has quit IRC | 09:31 | |
| admin0 | sorry .. 21.1.0 is the last success of this | 09:31 |
| *** pto has joined #openstack-ansible | 09:32 | |
| admin0 | where it works . | 09:32 |
| jrosser | well not much changes on stable branches so if you think something has been broken it should be easy to find | 09:34 |
| admin0 | i am more of an operator then a developer.. but does the screenshot make sense | 09:35 |
| admin0 | i see it failing it and giving up while the service is still in maintainance mode | 09:35 |
| admin0 | if i rerun haproxy playbook, the api is up and i can do the curl | 09:35 |
| jrosser | i cant really help with the screenshot | 09:40 |
| jrosser | as always, if you can come up with a reproducible test case in an AIO which fails, then thats the most helpful thing | 09:41 |
| admin0 | i think i need to edit this file https://opendev.org/openstack/openstack-ansible-os_cinder/src/branch/master/tasks/cinder_backends.yml .. and then search for a code that puts the API out of MAINT | 09:42 |
| jrosser | well the roles don't know anything about the loadbalancer, becasue they are only concerned with deploying a service | 09:43 |
| jrosser | the playbooks stitch everything together, and this is where the loadbalancer state is managed https://github.com/openstack/openstack-ansible/blob/master/playbooks/os-cinder-install.yml#L158-L164 | 09:44 |
| admin0 | copy/pasting the block did not helped :) | 09:50 |
| admin0 | i will try to run it in an AIO | 09:50 |
| admin0 | but if it works in an AIO but not in my greenfield, not sure how it will help.. i see it not taking out the maint mode before testing the api endpoint | 09:51 |
| admin0 | will setup an aio and report back | 09:52 |
| openstackgerrit | Merged openstack/openstack-ansible-os_nova master: Add extra packages from Debian updates https://review.opendev.org/c/openstack/openstack-ansible-os_nova/+/770457 | 09:57 |
| openstackgerrit | Andrew Bonney proposed openstack/openstack-ansible-os_nova stable/victoria: Add extra packages from Debian updates https://review.opendev.org/c/openstack/openstack-ansible-os_nova/+/770513 | 10:01 |
| *** akahat|lunch is now known as akahat|rover | 10:04 | |
| *** gshippey has joined #openstack-ansible | 11:05 | |
| *** sshnaidm|afk is now known as sshnaidm|ruck | 11:19 | |
| openstackgerrit | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_glance master: Add glance cache variables into glance-api https://review.opendev.org/c/openstack/openstack-ansible-os_glance/+/770606 | 11:53 |
| openstackgerrit | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_glance master: Add glance cache variables into glance-api https://review.opendev.org/c/openstack/openstack-ansible-os_glance/+/770606 | 11:59 |
| *** mgariepy has quit IRC | 12:04 | |
| *** fanfi has quit IRC | 12:13 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_zun master: Use global service variables https://review.opendev.org/c/openstack/openstack-ansible-os_zun/+/769965 | 12:15 |
| noonedeadpunk | can we also merge https://review.opendev.org/c/openstack/openstack-ansible-openstack_hosts/+/770455 ? | 12:25 |
| *** spatel has joined #openstack-ansible | 12:47 | |
| *** yann-kaelig has joined #openstack-ansible | 12:49 | |
| *** spatel has quit IRC | 12:52 | |
| *** rh-jelabarre has joined #openstack-ansible | 12:53 | |
| andrewbonney | noonedeadpunk: any thoughts on https://review.opendev.org/c/openstack/openstack-ansible-os_nova/+/770513 ? I can just re-check, just seems a little coincidental it's a tempest issue on buster | 13:04 |
| noonedeadpunk | yeah, I think we can just recheck | 13:05 |
| andrewbonney | Ok | 13:05 |
| *** mgariepy has joined #openstack-ansible | 13:21 | |
| *** yann-kaelig has left #openstack-ansible | 13:33 | |
| dmsimard | noonedeadpunk, jrosser: o/ question, have you considered using --diff in ansible-playbook runs ? everything that I see recorded by ara doesn't have it enabled | 13:49 |
| dmsimard | I have a patch up to implement diff rendering in result details and wanted to test it with OSA but if --diff isn't used, ansible doesn't pass it | 13:49 |
| noonedeadpunk | well... I guess it makes zuul output totally unreadable? | 13:51 |
| noonedeadpunk | as we have so much stuff.... | 13:51 |
| dmsimard | fair enough | 13:51 |
| noonedeadpunk | I think that also might slow down jobs | 13:51 |
| noonedeadpunk | (not sure though) | 13:52 |
| dmsimard | I would expect the performance impact to be near negligible but I can understand the perspective of bloated output | 13:53 |
| dmsimard | though I would also consider the diff output to be useful for understanding what actually changed | 13:53 |
| dmsimard | ¯\_(ツ)_/¯ | 13:53 |
| noonedeadpunk | Actually I find diff pretty usefull while debugging, but if were were talking about single role, and not dozen of them.... | 13:54 |
| dmsimard | I'll put up a patch just to test what it looks like | 13:55 |
| noonedeadpunk | yep, go on | 13:55 |
| noonedeadpunk | I think you can just add --diff here https://opendev.org/openstack/openstack-ansible/src/branch/master/scripts/gate-check-commit.sh#L176 | 13:56 |
| noonedeadpunk | and futher ofc as well) | 13:56 |
| dmsimard | openstack-ansible is an alias to ansible-playbook ? | 13:57 |
| dmsimard | alright ok I understand | 13:58 |
| noonedeadpunk | it's more like a wrapper, but yes | 13:59 |
| noonedeadpunk | pretty simple though https://opendev.org/openstack/openstack-ansible/src/branch/master/scripts/openstack-ansible.sh | 13:59 |
| openstackgerrit | David Moreau Simard proposed openstack/openstack-ansible master: DNM: Test ara 1.5.4 with --diff https://review.opendev.org/c/openstack/openstack-ansible/+/696634 | 13:59 |
| dmsimard | ^ should be ok I guess | 14:00 |
| openstackgerrit | David Moreau Simard proposed openstack/openstack-ansible master: DNM: Test ara 1.5.4 with --diff https://review.opendev.org/c/openstack/openstack-ansible/+/696634 | 14:02 |
| *** spatel has joined #openstack-ansible | 14:03 | |
| spatel | Any good utility to measure disk performance of openstack for local-disk (no ceph storage) | 14:05 |
| dmsimard | spatel: fio ? | 14:05 |
| noonedeadpunk | +1 :) | 14:06 |
| spatel | dmsimard: great! let me read about it. | 14:06 |
| *** d34dh0r53 has quit IRC | 14:10 | |
| *** d34dh0r53 has joined #openstack-ansible | 14:13 | |
| *** fresta has quit IRC | 14:18 | |
| *** ThiagoCMC has quit IRC | 14:18 | |
| *** fresta has joined #openstack-ansible | 14:18 | |
| openstackgerrit | Merged openstack/openstack-ansible-openstack_hosts master: Fix debian release https://review.opendev.org/c/openstack/openstack-ansible-openstack_hosts/+/770455 | 14:21 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-openstack_hosts stable/victoria: Fix debian release https://review.opendev.org/c/openstack/openstack-ansible-openstack_hosts/+/770582 | 14:22 |
| openstackgerrit | Andrew Bonney proposed openstack/openstack-ansible-openstack_hosts stable/victoria: Fix debian release https://review.opendev.org/c/openstack/openstack-ansible-openstack_hosts/+/770582 | 14:23 |
| admin0 | how safe is it to select 22.0.0.0rc1 for public facing production ? | 15:08 |
| *** dwilde has joined #openstack-ansible | 15:08 | |
| *** macz_ has joined #openstack-ansible | 15:13 | |
| admin0 | 22.0.0.0.rc1 is the stable/victoria right ? | 15:14 |
| *** macz_ has quit IRC | 15:17 | |
| *** pto has quit IRC | 15:18 | |
| *** klamath_atx has quit IRC | 15:23 | |
| *** ThiagoCMC has joined #openstack-ansible | 15:25 | |
| spatel | admin0: i am running on production :) | 15:28 |
| spatel | I have odd question, I have 3 controller node and not using ceph storage. I have upload image in glance so it will endup on one of random node. can i rsync my /var/glance folder across 3 node to make that image available on all nodes? | 15:30 |
| spatel | anyone did that before? | 15:30 |
| ThiagoCMC | Before Ceph, I used to host the images in an NFS (which was a SPOF). | 15:32 |
| spatel | I have plan for ceph in future and this is bandit for now.. let me do rsync and test it out.. | 15:35 |
| admin0 | spatel, you need to use nfs | 15:35 |
| admin0 | or gluster :) | 15:35 |
| admin0 | you can rsync | 15:35 |
| admin0 | you can rsync .. and it will just work fine | 15:35 |
| spatel | again this is just bandit and trying not to add extra services at this point, we don't upload image everyday. we have standard image which we run for years | 15:36 |
| mgariepy | spatel, maybe you can use lsyncd for the sync part. | 15:38 |
| spatel | that is a good idea to make it sync incase someone add image etc... | 15:39 |
| mgariepy | is the simple rsync works yep. | 15:39 |
| spatel | rsync works! i will setup lsyncd to make them sync regularly | 15:48 |
| *** dwilde has quit IRC | 15:49 | |
| *** dwilde has joined #openstack-ansible | 15:51 | |
| MrClayPole | My openstack-ansible AIO stein deployment is currently failing on os-ceilometer-install.yml on Ubuntu 18.04.5 with the error http://paste.openstack.org/raw/801594/ | 15:55 |
| MrClayPole | The journal in the ceilometere container indicates it's failing with "CRITICAL ceilometer [-] Unhandled error: ImportError: /openstack/venvs/ceilometer-19.1.7/lib/python2.7/site-packages/ujson.so: undefined symbol: Buffer_AppendShortHexUnchecked" error. | 15:55 |
| MrClayPole | I was wondering if it was due to this bug in ultrajson https://github.com/ultrajson/ultrajson/issues/346 | 15:56 |
| *** dwilde has quit IRC | 15:59 | |
| admin0 | is it possible to check ( from inventory file ) on which hypervisors will be with ceph and which will use local storage . before i run the playbooks ? | 16:02 |
| *** macz_ has joined #openstack-ansible | 16:02 | |
| spatel | admin0: you can create virtual group in inventory call ceph | 16:03 |
| spatel | and put those node in ceph group | 16:03 |
| admin0 | no no .. i mean i have already listed 2 hypervisors with host_vars: nova_libvirt_images_rbd_pool:vms and rest without | 16:04 |
| admin0 | before i deploy ,i want to check if ansible picked it up | 16:04 |
| admin0 | as opposed to checking it post-deploy | 16:04 |
| spatel | i don't think inventory knows about compute nodes variables | 16:05 |
| spatel | wish we have tag field in inventory to pass tag in openstack-ansible command to filter out some host. | 16:06 |
| jrosser | spatel: isnt that just this? https://docs.ansible.com/ansible/latest/user_guide/intro_patterns.html | 16:09 |
| spatel | hmm that should work, i vaguely remember i had issue in regex last time when i was trying to run playbook and N number of compute.. | 16:11 |
| spatel | possible my regex was issue. | 16:11 |
| spatel | i have 1 to 300 compute nodes and had issue to craft regex to run to pick intermediate numbers (like 20,41,45,73,100 etc..) in that case i create file.txt and add all nodes there to just run playbook on them | 16:13 |
| spatel | I like groups in inventory that is very easy to handle | 16:14 |
| *** mgariepy has quit IRC | 16:20 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_horizon master: Move horizon pip packages from constraints to requirements https://review.opendev.org/c/openstack/openstack-ansible-os_horizon/+/770283 | 16:25 |
| ThiagoCMC | How hard/easy is to create my own host groups with OSA? Just made up some 'mysrv-infra_hosts:' and put it under '/etc/openstack_deploy/conf.d'? What about its group_vars? | 16:30 |
| noonedeadpunk | you should also create env.d I think | 16:31 |
| noonedeadpunk | this might be helpful https://docs.openstack.org/openstack-ansible/latest/reference/inventory/understanding-inventory.html | 16:33 |
| dmsimard | noonedeadpunk: example of rendered diff: https://api.trunk.demo.recordsansible.org/results/435291.html although there is an issue I need to troubleshoot like config_template task has an error 500 https://api.trunk.demo.recordsansible.org/results/435741.html | 16:35 |
| noonedeadpunk | this might need config_template adjustment ... | 16:36 |
| dmsimard | I don't know if it's ara's fault or config_template yet | 16:36 |
| dmsimard | busy on something else, I'll check later | 16:36 |
| ThiagoCMC | noonedeadpunk, thank you! | 16:36 |
| dmsimard | the diff format might differ from the copy and template modules | 16:37 |
| ThiagoCMC | My plan is to create a OSA-based container for Ceph Dashboard [grafana-server] Ceph's group. Anyone doing something like this? | 16:37 |
| CeeMac | noonedeadpunk: you've dabbled in the telemetry projects right? | 16:39 |
| noonedeadpunk | at least some of them, yes:) | 16:39 |
| CeeMac | any change you've come across that issue ^ MrClayPole posted earlier? | 16:40 |
| CeeMac | and hi also :D | 16:40 |
| spatel | ThiagoCMC: check this out - http://paste.openstack.org/show/770649/ | 16:40 |
| noonedeadpunk | oh, yes, I was facing it | 16:40 |
| ThiagoCMC | spatel, neat! | 16:40 |
| noonedeadpunk | it needs to bump ujson to verson 2.0 or higher | 16:41 |
| CeeMac | please tell me there is a quick and easy fix for it :) | 16:41 |
| *** tosky has quit IRC | 16:41 | |
| CeeMac | via constraints or there is a better way? | 16:42 |
| noonedeadpunk | I put in /opt/openstack-ansible/global-requirement-pins.txt `ujson==2.0.3` | 16:42 |
| *** tosky has joined #openstack-ansible | 16:42 | |
| CeeMac | perfect, thanks! We'll give that a go | 16:42 |
| MrClayPole | noonedeadpunk: thanks | 16:42 |
| CeeMac | noonedeadpunk: do we need to re-run repo plays or anything, or that'll just pick up from the ceilometer play | 16:43 |
| noonedeadpunk | iirc should pick up, but maybe re-run with `-e venv_rebuild=true` | 16:46 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_tempest master: Move tempest pip package from a constraint to a requirement https://review.opendev.org/c/openstack/openstack-ansible-os_tempest/+/770281 | 16:50 |
| *** gyee has joined #openstack-ansible | 17:00 | |
| *** mgariepy has joined #openstack-ansible | 17:20 | |
| openstackgerrit | Merged openstack/openstack-ansible-openstack_hosts stable/ussuri: Fix debian release https://review.opendev.org/c/openstack/openstack-ansible-openstack_hosts/+/770195 | 17:29 |
| *** sshnaidm|ruck is now known as sshnaidm|afk | 17:30 | |
| *** maharg101 has quit IRC | 17:31 | |
| openstackgerrit | Jonathan Rosser proposed openstack/ansible-role-python_venv_build master: Add source code and bugs url link to README https://review.opendev.org/c/openstack/ansible-role-python_venv_build/+/598580 | 17:37 |
| *** miloa has quit IRC | 17:40 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_aodh master: Fix documentation formatting https://review.opendev.org/c/openstack/openstack-ansible-os_aodh/+/587019 | 17:41 |
| jrosser | noonedeadpunk: i wonder if we should abandon open patches on stable/rocky? | 17:53 |
| noonedeadpunk | Yeah, I guess that makes sense. Not sure how much CI is broken there, but anyway we have no time on them... | 17:54 |
| noonedeadpunk | Maybe even stein should go the same way actually | 17:54 |
| jrosser | yes i am trying to clean up the review dashboard | 17:55 |
| jrosser | i'll get rid of the rocky stuff to start with | 17:55 |
| mgariepy | hmm. old patch in merge conflict from 2018 could probably be abandoned also .. | 17:58 |
| jrosser | i did a quick pass bottom-up on the current patches that have passes, anything else that you think looks bogus we can also abandon | 17:59 |
| jrosser | some stuff looked reasonable but needed a manual rebase | 17:59 |
| CeeMac | sad times, i really need to get my environments up to train minimum | 18:03 |
| *** rpittau is now known as rpittau|afk | 18:05 | |
| mgariepy | CeeMac, which release are you on ? | 18:05 |
| jrosser | CeeMac: you've only really got 4 months before Train is extended-maintainance too...... | 18:05 |
| mgariepy | once on T you can do U ;p | 18:06 |
| ThiagoCMC | wow... Have you guys ever used this ClusterControl to visualize OSA's Galera Cluster? I just installed it here inside of an OSA-based custom container in an AIO deploy to test it... It looks awesome! I connected CC to OSA's 1 node Galera (from my custom container) :-D | 18:06 |
| jrosser | ThiagoCMC: you should make some stuff for the openstack-ansible-ops repo :) | 18:07 |
| noonedeadpunk | btw regarding it.. we should somehow unblock CI there... | 18:07 |
| noonedeadpunk | I tried to spend time on it but :( | 18:08 |
| jrosser | yes we should | 18:08 |
| jrosser | ebbex: were you looking at the ELK stuff? | 18:08 |
| noonedeadpunk | didn't have any result actually | 18:08 |
| ThiagoCMC | jrosser, I believe that I'm going do it someday! lol | 18:08 |
| jrosser | we were unable to reproduce the error in the ops repo CI locally | 18:08 |
| jrosser | there is something strange going on | 18:08 |
| noonedeadpunk | oh, rly? because I was.... | 18:08 |
| jrosser | oh, interesting | 18:09 |
| noonedeadpunk | maybe it's smth different now... | 18:09 |
| jrosser | well i think there is a patch from gshippey which is a starting point | 18:09 |
| jrosser | virtualenv > 20 stuff i think or something | 18:09 |
| noonedeadpunk | yeah, and I adjusted it even more, but now it's stuck on populting data for me | 18:09 |
| CeeMac | mgariepy: rocky still. | 18:10 |
| mgariepy | on 16.04 or 18.04 ? | 18:10 |
| CeeMac | jrosser: really? Where has the time gone! | 18:10 |
| CeeMac | 18.04 | 18:10 |
| jrosser | i'm assuming that this is accurate https://releases.openstack.org/ | 18:11 |
| *** cloudnull has quit IRC | 18:12 | |
| noonedeadpunk | ah, yes, I just used virtualenv one and placed there several adjustments... | 18:12 |
| *** cloudnull has joined #openstack-ansible | 18:12 | |
| jrosser | i think it was broken even before the virtualenv thing came up | 18:12 |
| jrosser | so its kind of broken^2 | 18:12 |
| CeeMac | Hopefully trilio supports U then we can go upto U then. Quite a run from R to U though :| | 18:12 |
| *** andrewbonney has quit IRC | 18:12 | |
| noonedeadpunk | I think it got borken with some elastic upgrade | 18:13 |
| CeeMac | U is transition release for focal? | 18:13 |
| noonedeadpunk | as we don't stick to the fixed version but use kind of latest one | 18:13 |
| noonedeadpunk | V still has bionic on board | 18:13 |
| CeeMac | It does? Handy | 18:14 |
| jrosser | we have to make a decision for W on that i guess | 18:14 |
| CeeMac | I figured with 20.04 being LTS now that would have happened already. | 18:15 |
| CeeMac | Was it U first supported focal? | 18:15 |
| jrosser | U & V support both | 18:15 |
| CeeMac | Kk | 18:15 |
| noonedeadpunk | yep, U is first where focal appeared | 18:15 |
| openstackgerrit | Merged openstack/openstack-ansible-os_nova stable/victoria: Add extra packages from Debian updates https://review.opendev.org/c/openstack/openstack-ansible-os_nova/+/770513 | 18:15 |
| mgariepy | i run u on focal for one deployment, | 18:15 |
| CeeMac | Good to know | 18:15 |
| jrosser | nice big upgrade window | 18:15 |
| jrosser | rocky was a real pain for that | 18:16 |
| CeeMac | Did they ditch netplan yet? | 18:16 |
| jrosser | apt-install ifupdown <- worksforme | 18:16 |
| mgariepy | need to upgarde 16 to 18 on R on some other. but i want to do it live and there are quite a few annoyance with R. lol | 18:16 |
| CeeMac | Sad but reassuring at the same time :) | 18:16 |
| mgariepy | netplan once you have fixed 2 or 3 things that don't works is ok.. | 18:17 |
| CeeMac | Did you get a chance to play with ifupdown2? I seem to recall you mentioning it | 18:17 |
| jrosser | it looked really interesting and seemed to attempt to actually have a concept of current->desired config and try to resolve that | 18:18 |
| CeeMac | Is "OK" sufficient though? :D | 18:18 |
| jrosser | but tbh all my ifupdown stuff is working right now and there are more interesting things to work on | 18:19 |
| CeeMac | True fact | 18:19 |
| jrosser | better spend time on the new installer for focal becasue thats so much nicer than debian-installer | 18:19 |
| jrosser | yaml config and mostly sane | 18:20 |
| mgariepy | but it uses netplan ;p | 18:20 |
| mgariepy | haha | 18:20 |
| jrosser | until my ansible comes along and 'fixes that' :) | 18:20 |
| mgariepy | haha :) | 18:21 |
| CeeMac | Yeah I need to work on that whole deployment and config piece still | 18:23 |
| mgariepy | i played a bit with ifupdown2 on cumulus switches.. | 18:24 |
| CeeMac | How'd you find it? | 18:27 |
| MrClayPole | noonedeadpunk: It failed with the same error running os-ceilometer-install.yml even with `-e venv_rebuild=true' with `ujson==2.0.3` in /opt/openstack-ansible/global-requirement-pins.txt. I removed the ceilometer container and then ran setup-everything.yml and it's now working. Thanks for you help. | 18:29 |
| CeeMac | Nice. :) | 18:35 |
| dmsimard | noonedeadpunk: looked a bit at config_template -- the difference between it and copy/template is that copy/template provide a "before_header" and "after_header" key to display the file paths involved | 18:57 |
| dmsimard | as well as before/after keys | 18:57 |
| dmsimard | this is shown in the unpatched version, taking the same links as earlier today https://api.demo.recordsansible.org/results/435291.html vs https://api.demo.recordsansible.org/results/435741.html | 18:58 |
| ThiagoCMC | Guys, is there any easy way to enable Ceph Dashboard (Ceph Ansible '[grafana-server]' host groups) in an OSA AIO + Ceph 'SCENARIO=aio_lxc_ceph'? I'm wondering if ceph-ansible could deploy the Dashboard (which are Docker containers, if I'm not mistaken) inside of 'ceph-moh_container-XXX'... | 19:00 |
| spatel | Do you folks use "nohz_full" kernel option to host compute node? | 19:02 |
| admin0 | so even with host_vars: nova_libvirt_images_rbd_pool: vms .. it did not worked | 19:06 |
| admin0 | spatel, what was your way | 19:06 |
| admin0 | can give a quick example | 19:06 |
| spatel | you want to run playbook only on host which has "nova_libvirt_images_rbd_pool: vms" right? | 19:07 |
| admin0 | yes .. i had it wth host_vars: , but it did not worked | 19:07 |
| spatel | can you should me what file ? | 19:08 |
| spatel | host_vars is directory | 19:08 |
| admin0 | spatel, i had it like this: https://gist.github.com/a1git/a401cdf68b760f2495be0ae0fc0acaa3 | 19:10 |
| admin0 | this format worked on 21.0 .. i am trying 22 | 19:10 |
| spatel | it should work in 22 also (i am not using ceph with 22) | 19:11 |
| spatel | you need to put that variable in specific section right? | 19:12 |
| spatel | this is what i am doing | 19:12 |
| spatel | admin0: http://paste.openstack.org/show/801604/ | 19:13 |
| spatel | i create file for each compute node and add specific setting there instead of single file. | 19:14 |
| admin0 | ok | 19:14 |
| admin0 | good idea | 19:14 |
| spatel | in your case you can do h1.yml h2.yml etc.. | 19:16 |
| spatel | check out this example here - https://docs.openstack.org/openstack-ansible/latest/reference/configuration/using-overrides.html | 19:16 |
| admin0 | got it | 19:16 |
| admin0 | made those | 19:16 |
| spatel | http://paste.openstack.org/show/801605/ | 19:16 |
| admin0 | rerunning playbook | 19:16 |
| spatel | i like keeping each compute file separate for simplify :) | 19:17 |
| spatel | i mad bad typos in past when using single file and with 300 compute nodes its not fun to find typo :) | 19:18 |
| *** maharg101 has joined #openstack-ansible | 19:28 | |
| admin0 | spatel, https://gist.github.com/a1git/93a0092034207fdcfbae36abe97f0b01 -- this is the end result . but it still spawn locally and does not use ceph | 19:32 |
| *** maharg101 has quit IRC | 19:32 | |
| spatel | look like it didn't added nova_libvirt_images_rbd_pool option in nova.conf file | 19:34 |
| spatel | This is what i have http://paste.openstack.org/show/801606/ | 19:36 |
| admin0 | i am going to drop 22 and go back to 21 i think | 19:36 |
| admin0 | or .. any other options i can try ? | 19:37 |
| spatel | admin0: try this in that file. - http://paste.openstack.org/show/801607/ | 19:37 |
| spatel | put layers to add that option | 19:38 |
| jrosser | remeber also anything you put in user_*.yml will take precendence over host/group_vars | 19:38 |
| spatel | not for host_var i believe | 19:39 |
| jrosser | well this https://docs.ansible.com/ansible/latest/user_guide/playbooks_variables.html#variable-precedence-where-should-i-put-a-variable | 19:40 |
| jrosser | -e beats everything | 19:40 |
| spatel | you are right! | 19:40 |
| jrosser | this can often give "i set the variable but it doesnt work" kind of problem | 19:40 |
| jrosser | because the same thing in user_*.yml will win | 19:40 |
| admin0 | you both confuse me on what to try next :) | 19:41 |
| admin0 | i think i will drop 22 for a bit and go back on 21 | 19:41 |
| admin0 | and try this on lab | 19:41 |
| admin0 | this one is going to be prod | 19:41 |
| spatel | jrosser: worth putting that notes in this page about precedence layer - https://docs.openstack.org/openstack-ansible/latest/reference/configuration/using-overrides.html | 19:42 |
| ebbex | jrosser, noonedeadpunk : sorry to chime in a bit late on the elk stuff. But yeah, I'm looking into it. Right now the CI stalls on an undocumented rest api call to set a default index, I think i've found a way to do the same with documented calls. Might try to replace it, but that doesn't really explain the timeout. Looking into whether there are some issues between kibana and elastic. | 19:44 |
| openstackgerrit | Merged openstack/openstack-ansible-os_sahara master: Use global service variables https://review.opendev.org/c/openstack/openstack-ansible-os_sahara/+/769958 | 20:19 |
| *** nurdie has joined #openstack-ansible | 20:38 | |
| nurdie | Anyone know why my l3_agent.ini on controller nodes would be completely missing after Stein -> Train upgrade? | 20:39 |
| ThiagoCMC | So, guys, I managed to create my custom OSA-based container and it has IP from br-mgmt (its eth1) and a eth0 IP from lxcbr0... How hard is to create a very simple HAProxy entry for it? Let's say that in my custom container I have running netcat on port 80 & 443 to simulate it... | 20:41 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_tempest master: Move tempest pip package from a constraint to a requirement https://review.opendev.org/c/openstack/openstack-ansible-os_tempest/+/770281 | 20:41 |
| jrosser | ThiagoCMC: take a look at how haproxy_extra_services is used on here https://opendev.org/openstack/openstack-ansible-ops/src/branch/master/elk_metrics_6x/README.rst | 20:43 |
| ThiagoCMC | jrosser, awesome! | 20:43 |
| nurdie | I have 3 controllers. neutron-l3-agent is somehow running just fine on two of them but is restarting over and over on my first node | 20:47 |
| nurdie | All three nodes have no /etc/neutron/l3_agent.ini | 20:47 |
| openstackgerrit | Merged openstack/openstack-ansible-openstack_hosts stable/victoria: Fix debian release https://review.opendev.org/c/openstack/openstack-ansible-openstack_hosts/+/770582 | 20:54 |
| *** theintern_ has joined #openstack-ansible | 21:02 | |
| nurdie | Looks like the 20.2.0 upgrade didn't update systemd to use the new venvs on controller nodes o.0 | 21:02 |
| nurdie | * for neutron | 21:09 |
| jrosser | nurdie: please file a bug with as much info as possible if you think something is wrong | 21:21 |
| nurdie | I'm testing 20.2.1 on one of my controllers. | 21:24 |
| nurdie | jrosser: I will, thanks! | 21:24 |
| dmsimard | noonedeadpunk: btw fixed the error 500 by giving up if it can't parse it ¯\_(ツ)_/¯ https://api.trunk.demo.recordsansible.org/results/435741.html | 21:55 |
| dmsimard | I can file a bug if it's something you'd be interested in eventually fixing, I guess it's a convention elsewhere: https://codesearch.recordsansible.org/?q=before_header | 21:57 |
| dmsimard | I dug a bit around https://github.com/openstack/ansible-config_template/blob/master/action/config_template.py and it looks like there's a purpose for the current diff key format it's using -- I don't have an opinion, just pointing it's different :) | 22:00 |
| openstackgerrit | Dmitriy Rabotyagov proposed openstack/openstack-ansible stable/victoria: Bump Victoria for the release https://review.opendev.org/c/openstack/openstack-ansible/+/770073 | 22:05 |
| *** klamath_atx has joined #openstack-ansible | 22:11 | |
| *** spatel has quit IRC | 22:56 | |
| *** rh-jelabarre has quit IRC | 23:17 | |
| *** maharg101 has joined #openstack-ansible | 23:29 | |
| *** maharg101 has quit IRC | 23:34 | |
| *** theintern_ has quit IRC | 23:34 | |
| *** nurdie has quit IRC | 23:37 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
