| *** ysandeep|out is now known as ysandeep | 05:22 | |
| opendevreview | Merged openstack/openstack-ansible master: [doc] Add note regarding 127.0.1.1 removal https://review.opendev.org/c/openstack/openstack-ansible/+/830175 | 07:53 |
|---|---|---|
| noonedeadpunk | mornings | 08:10 |
| noonedeadpunk | dmsimard: it's installing correct version, yes, but in ansible-galaxy collection list it's show as *. So eventually you never know what you're using in fact and no way to check either, as you can't go and do `git status` | 08:11 |
| noonedeadpunk | jrosser: it's installing properly since 2.12 iirc, so we could use github everywhere technically | 08:12 |
| noonedeadpunk | it's still nasty though | 08:12 |
| jrosser | right yes that’s what I was going to say as well | 08:13 |
| jrosser | there’s pretty much no way to tell what you actually got | 08:13 |
| jrosser | which I why I’d really like an option for it to leave the git metadata in place | 08:14 |
| jrosser | noonedeadpunk: follow on from the zuul queues stuff - I think we are going to have to try it | 08:15 |
| noonedeadpunk | I think main there is "have to"?:) | 08:16 |
| noonedeadpunk | I had to leave yestarday as raised an alarm in the office - seems they don't like ppl leaving after hours ( ͡ᵔ ͜ʖ ͡ᵔ ) | 08:17 |
| jrosser | long discussion with clarkb yesterday, basically the zuul people say we are doing it wrong rather than the new behaviour being a bug | 08:17 |
| jrosser | and it’s kind of up to us to show that we have difficulty with a shared queue | 08:18 |
| noonedeadpunk | then I haven't missed a thing | 08:18 |
| jrosser | I still feel we are in for a very bad time with things like proposal not patches where they are truly unrelated and maybe 30% will fail | 08:19 |
| jrosser | *bot | 08:19 |
| noonedeadpunk | I think we will have to push for having 12Gb flavors.... | 08:25 |
| noonedeadpunk | as currently there's a 16gb option, but only in very limited providers | 08:26 |
| *** raukadah is now known as chandankumar | 08:29 | |
| jrosser | is that for centos OOM? | 08:31 |
| opendevreview | Jonathan Rosser proposed openstack/openstack-ansible master: Bump versions of ansible collections https://review.opendev.org/c/openstack/openstack-ansible/+/836379 | 08:37 |
| opendevreview | Jonathan Rosser proposed openstack/openstack-ansible master: Bump openstack collection to master for system scope https://review.opendev.org/c/openstack/openstack-ansible/+/836382 | 08:37 |
| *** ysandeep is now known as ysandeep|lunch | 09:08 | |
| *** ysandeep|lunch is now known as ysandeep | 10:03 | |
| *** dviroel|out is now known as dviroel | 11:21 | |
| *** ysandeep is now known as ysandeep|afk | 11:37 | |
| dmsimard | noonedeadpunk, jrosser: ah yeah, I see what you mean -- I found out I installed the right version by checking the changelog file but that's not very convenient I must admit | 12:46 |
| noonedeadpunk | I bet it might be even absent for some collections ;) | 12:48 |
| noonedeadpunk | then you would need to go to git, check for diffs to see if they present | 12:48 |
| noonedeadpunk | s/git/github/ | 12:49 |
| *** ysandeep|afk is now known as ysandeep | 12:56 | |
| opendevreview | Merged openstack/openstack-ansible-os_aodh master: Use openstack uri proto https://review.opendev.org/c/openstack/openstack-ansible-os_aodh/+/834845 | 13:15 |
| dmsimard | noonedeadpunk: yeah it's not something I would rely on for sure | 13:18 |
| noonedeadpunk | as other possible solution I'd say that maybe it's better to use provided by user version rather then just *.... | 13:20 |
| noonedeadpunk | or well, fallback to provided version | 13:20 |
| noonedeadpunk | even if it's SHA - it's still better I guess | 13:20 |
| noonedeadpunk | and then it's possible to verify if version in requiements has changed and update of collection is needed | 13:21 |
| NeilHanlon | pin it all to `master` /s | 13:22 |
| dmsimard | I will put visibility on the lack of version as an issue in the individual collections -- it's not perfect but it would address the issue moving forward (i.e, for newer versions the field would not be null..) | 13:22 |
| opendevreview | Merged openstack/openstack-ansible master: Update ansible-core to 2.12.4 https://review.opendev.org/c/openstack/openstack-ansible/+/836378 | 13:29 |
| dmsimard | I've "escalated" https://github.com/ansible-collections/ansible.netcommon/issues/239 && https://github.com/ansible-collections/openvswitch.openvswitch/issues/94 and told them ansible.utils is also problematic, let's see how it goes | 13:37 |
| opendevreview | Damian Dąbrowski proposed openstack/openstack-ansible-os_tempest master: Improve domain support https://review.opendev.org/c/openstack/openstack-ansible-os_tempest/+/836833 | 13:45 |
| damiandabrowski[m] | hey! Will You be able to have a look at my tempest changes and leave initial review? https://review.opendev.org/q/topic:tempest-damian-2021-12 | 13:51 |
| damiandabrowski[m] | There is a huge relation chain but I don't think we can avoid it as all of these changes focus on the same files :/ However, they are pretty straightforward so I hope all of them will be merged. | 13:51 |
| damiandabrowski[m] | So my idea is to: gather initial reviews from You -> (make corrections) -> rebase all of them -> merge all. What do You think? | 13:52 |
| magodde | Hello there, when performing the distribution upgrade between Ussuri and Victoria according to that guide https://docs.openstack.org/openstack-ansible/victoria/admin/upgrades/distribution-upgrades.html. Do we have to checkout to Victoria first or do we have to reinstall Ussuri on all host before running a mass upgrade ? Thanks | 14:03 |
| opendevreview | Merged openstack/openstack-ansible stable/xena: Connect openstack_pki_regen_ca variable to pki role https://review.opendev.org/c/openstack/openstack-ansible/+/834017 | 14:04 |
| *** dviroel is now known as dviroel|ptg | 14:13 | |
| damiandabrowski[m] | magodde: I | 14:46 |
| damiandabrowski[m] | I'm not 100% sure but i think that as long as both distros are supported by victoria&ussuri You can do it both ways | 14:47 |
| magodde | So, if we checkout to Victoria first, we'll have hosts with OS Victoria and hosts with OS Ussuri since we can't reinstall all hosts at the same time | 14:48 |
| magodde | isn't this an issue ? | 14:48 |
| damiandabrowski[m] | yeah and it should be fine if You upgrade control plane first | 14:49 |
| damiandabrowski[m] | for ex. You have to upgrade neutron-server before upgrading neutron-agents | 14:49 |
| magodde | all right so let me summarize just to be sure. | 14:50 |
| damiandabrowski[m] | and neutron-server on Victoria should work fine with both Ussuri&Victoria agents | 14:50 |
| magodde | we have 3 controllers (one controller is the primary one) and 4 computes. All hosts are running Ubuntu 18.04 and Ussuri | 14:50 |
| magodde | We checkout on stable/Victoria on our deployment host | 14:51 |
| magodde | then we upgrade each controller host to 20.04 and we run the Victoria playbook. The last controller has to be the primary one | 14:52 |
| magodde | Then we do the same with our compute nodes. | 14:52 |
| damiandabrowski[m] | ah, so You are going to mix distribution and openstack upgrade at the some time, honestly I don't have any experience with that | 14:53 |
| magodde | Yes, the point is that we have to perform the upgrade to 20.04. | 14:54 |
| damiandabrowski[m] | it may be safer to firstly perform openstack upgrade on all nodes(ubuntu 18.04) and then start ubuntu upgrade | 14:54 |
| magodde | But can we update to Victoria with Ubuntu 18.04 ? | 14:55 |
| damiandabrowski[m] | ah, You're using distro install? (as now I can see that distro install is not supported for victoria&ubuntu 18.04) | 14:55 |
| damiandabrowski[m] | but if You're using source install, then yes You can update to Victoria with Ubuntu 18.04 | 14:55 |
| damiandabrowski[m] | https://docs.openstack.org/openstack-ansible/latest/admin/upgrades/compatibility-matrix.html | 14:55 |
| magodde | We are using source (install_method: source) | 14:56 |
| damiandabrowski[m] | so yeah, You can upgrade openstack to Victoria with Ubuntu 18.04 ;) all of my regions were running victoria+18.04 at some point | 14:57 |
| magodde | All right then | 14:58 |
| magodde | Thx for all this :) | 14:58 |
| damiandabrowski[m] | You're welcome ;) | 14:59 |
| spatel | noonedeadpunk there | 15:00 |
| noonedeadpunk | semi-around | 15:00 |
| spatel | are you running wallaby anywhere in prod? | 15:02 |
| spatel | since i upgrade to wallaby i am noticing this error in compute nodes frequently - https://paste.opendev.org/show/bft9znewTxyXHkvIcQO0/ | 15:02 |
| spatel | solution is to restart nova-compute agent | 15:03 |
| spatel | look like some kind of connection re-try issue here | 15:03 |
| damiandabrowski[m] | spatel: not sure if it's related, but worth to have a look: https://bugs.launchpad.net/oslo.messaging/+bug/1934937 | 15:05 |
| noonedeadpunk | spatel: nope, only V and X | 15:07 |
| spatel | damiandabrowski[m] looks like my issue.. i will give it a try and see.. but this is causing massive issue in my production :( | 15:08 |
| noonedeadpunk | I bet that amqp version has been bmped in W as well | 15:08 |
| noonedeadpunk | And andrewbonney placed several patches to cover that | 15:09 |
| andrewbonney | Yeah, that paste looks a little different to anything I've seen before | 15:09 |
| spatel | andrewbonney do you think this will solve my issue - [oslo_messaging_rabbit]/heartbeat_in_pthread=false | 15:13 |
| andrewbonney | If it isn't already deployed then it's worth doing anyway. I don't think there's been an OSA Wallaby release since that was patched. I can't guarantee it'll fix your specific issue though | 15:14 |
| spatel | hmm | 15:17 |
| *** ysandeep is now known as ysandeep|out | 15:21 | |
| spatel | I will try and see if i can re-produce this issue | 15:21 |
| noonedeadpunk | it wasn't relased yet, no | 15:27 |
| spatel | hmm | 15:30 |
| spatel | so this option not released in wallaby? heartbeat_in_pthread=false | 15:30 |
| *** dviroel|ptg is now known as dviroel|ptg|lunch | 15:42 | |
| *** dviroel|ptg|lunch is now known as dviroel|ptg | 16:03 | |
| noonedeadpunk | you can use overrides easily to get it in | 16:12 |
| noonedeadpunk | but I kind of doubt that issue is related | 16:12 |
| noonedeadpunk | well, at least with it you would see pthread errors in log | 16:12 |
| noonedeadpunk | but I'm not sure how errors would look like with old amqp | 16:13 |
| noonedeadpunk | spatel: you should check that amqp==5.0.8 in your venv | 16:14 |
| noonedeadpunk | that was patch covering it https://review.opendev.org/c/openstack/requirements/+/823350 | 16:14 |
| spatel | hmm let me checvk | 16:15 |
| spatel | I am running amqp=5.0.5 | 16:16 |
| spatel | let me see if i can bump in dev and re-produce it | 16:18 |
| *** dviroel|ptg is now known as dviroel | 17:22 | |
| *** dviroel is now known as dviroel|mtg | 17:47 | |
| opendevreview | James Denton proposed openstack/openstack-ansible-os_ironic master: Remove [keystone] configuration block https://review.opendev.org/c/openstack/openstack-ansible-os_ironic/+/831544 | 18:36 |
| *** dviroel|mtg is now known as dviroel | 18:43 | |
| *** dviroel is now known as dviroel|afk | 20:34 | |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!