| opendevreview | OpenStack Proposal Bot proposed openstack/openstack-ansible master: Imported Translations from Zanata https://review.opendev.org/c/openstack/openstack-ansible/+/948290 | 04:05 |
|---|---|---|
| opendevreview | Ivan Anfimov proposed openstack/openstack-ansible master: wip https://review.opendev.org/c/openstack/openstack-ansible/+/948111 | 07:24 |
| opendevreview | Ivan Anfimov proposed openstack/openstack-ansible-ops master: Remove tags from README https://review.opendev.org/c/openstack/openstack-ansible-ops/+/948393 | 07:29 |
| opendevreview | Ivan Anfimov proposed openstack/openstack-ansible-ops master: Remove tags from README https://review.opendev.org/c/openstack/openstack-ansible-ops/+/948393 | 07:29 |
| opendevreview | Ivan Anfimov proposed openstack/openstack-ansible master: Remove tags from README https://review.opendev.org/c/openstack/openstack-ansible/+/948394 | 07:30 |
| opendevreview | Ivan Anfimov proposed openstack/openstack-ansible master: Remove tags from README https://review.opendev.org/c/openstack/openstack-ansible/+/948394 | 07:31 |
| opendevreview | Ivan Anfimov proposed openstack/openstack-ansible-rabbitmq_server master: Remove tags from README https://review.opendev.org/c/openstack/openstack-ansible-rabbitmq_server/+/948410 | 09:31 |
| opendevreview | Ivan Anfimov proposed openstack/openstack-ansible-os_horizon master: Remove tags from README https://review.opendev.org/c/openstack/openstack-ansible-os_horizon/+/948411 | 09:31 |
| opendevreview | Ivan Anfimov proposed openstack/openstack-ansible-rabbitmq_server master: Remove tags from README https://review.opendev.org/c/openstack/openstack-ansible-rabbitmq_server/+/948410 | 09:32 |
| opendevreview | Ivan Anfimov proposed openstack/openstack-ansible-os_horizon master: Remove tags from README https://review.opendev.org/c/openstack/openstack-ansible-os_horizon/+/948411 | 09:32 |
| opendevreview | Ivan Anfimov proposed openstack/openstack-ansible-os_keystone master: Remove tags from README https://review.opendev.org/c/openstack/openstack-ansible-os_keystone/+/948412 | 09:36 |
| opendevreview | Ivan Anfimov proposed openstack/openstack-ansible-os_keystone master: Remove tags from README https://review.opendev.org/c/openstack/openstack-ansible-os_keystone/+/948412 | 09:36 |
| opendevreview | Ivan Anfimov proposed openstack/openstack-ansible-os_cinder master: Remove tags from README https://review.opendev.org/c/openstack/openstack-ansible-os_cinder/+/948413 | 09:37 |
| opendevreview | Ivan Anfimov proposed openstack/openstack-ansible-os_cinder master: Remove tags from README https://review.opendev.org/c/openstack/openstack-ansible-os_cinder/+/948413 | 09:38 |
| opendevreview | Ivan Anfimov proposed openstack/openstack-ansible-plugins master: Remove tags from README https://review.opendev.org/c/openstack/openstack-ansible-plugins/+/948414 | 09:40 |
| opendevreview | Ivan Anfimov proposed openstack/openstack-ansible-os_magnum master: Remove tags from README https://review.opendev.org/c/openstack/openstack-ansible-os_magnum/+/948415 | 09:40 |
| opendevreview | Ivan Anfimov proposed openstack/openstack-ansible-os_magnum master: Remove tags from README https://review.opendev.org/c/openstack/openstack-ansible-os_magnum/+/948415 | 09:41 |
| opendevreview | Ivan Anfimov proposed openstack/openstack-ansible-plugins master: Remove tags from README https://review.opendev.org/c/openstack/openstack-ansible-plugins/+/948414 | 09:41 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_keystone master: Use standalone httpd role https://review.opendev.org/c/openstack/openstack-ansible-os_keystone/+/948338 | 09:41 |
| jpw_ | is there something else I need to do to inform cinder that ceph is available. I'm using an existing cluster, i've created the pools, keyrings + perms and given all of the details to user_variables.yml. yet my `openstack volume service list` only shows the '__DEFAULT__' type which obvs won't work since i'm not deploying any storage_hosts | 10:16 |
| noonedeadpunk | jpw_: it's pretty much up to how you define `cinder_backends` variable | 10:21 |
| noonedeadpunk | ie example for Ceph we have: https://opendev.org/openstack/openstack-ansible/src/branch/master/etc/openstack_deploy/user_variables.yml.prod-ceph.example#L32-L41 | 10:21 |
| noonedeadpunk | even couple lines above that are also useful | 10:22 |
| jpw_ | and if working correctly would I expect to see a new volume type? how about "cinder-volumes" in `openstack volume service list`? | 10:25 |
| noonedeadpunk | jpw_: so you need to define `storage_hosts` for cinder-volumes to spawn | 10:42 |
| noonedeadpunk | Also, I think you need/want an env.d override to run cinder-volume in containers | 10:43 |
| noonedeadpunk | like that: https://opendev.org/openstack/openstack-ansible/src/branch/master/etc/openstack_deploy/env.d/cinder-volume.yml.container.example | 10:43 |
| jpw_ | i have the is_metal: false directive. I'll try adding some storage_hosts in | 10:53 |
| noonedeadpunk | storage_hosts do not produce ceph or anything - they're solely for cinder-volumes | 10:53 |
| noonedeadpunk | it's defined here: https://opendev.org/openstack/openstack-ansible/src/branch/master/inventory/env.d/cinder.yml#L38-L43 | 10:54 |
| noonedeadpunk | for ceph deployment it would be ceph-mon_hosts, cep[h-osd_hosts, etc | 10:54 |
| noonedeadpunk | https://opendev.org/openstack/openstack-ansible/src/branch/master/inventory/env.d/ceph.yml#L62-L92 | 10:54 |
| jpw_ | are they still required if it's an existing cluster? | 10:59 |
| jpw_ | an existing ceph cluster | 10:59 |
| noonedeadpunk | nope, they are not for sure | 11:01 |
| jpw_ | i'm going the no-ssh route so i've provided a cinder.conf some keyrings and some pool names. currently i've only got storage-infra_hosts defined. i'm trying again with storage_hosts too. | 11:01 |
| noonedeadpunk | only storage_hosts to deploy cinder-volume | 11:01 |
| jpw_ | thank you | 11:02 |
| noonedeadpunk | yes, right, no-ssh is the best way forward | 11:02 |
| noonedeadpunk | as well as managing ceph separately is a very wise choice as well | 11:02 |
| jpw_ | its not my first cloud :D | 11:06 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_keystone master: Use standalone httpd role https://review.opendev.org/c/openstack/openstack-ansible-os_keystone/+/948338 | 11:06 |
| noonedeadpunk | right, you just never know knowledgebase so often just start from the beginning to be on the safe side and not overlook smth very basic | 11:08 |
| noonedeadpunk | jpw_: btw, fwiw, you also can configure networking with OSA today completely. If you want to | 11:08 |
| jpw_ | im keen to give that a go | 11:09 |
| noonedeadpunk | ie https://docs.openstack.org/openstack-ansible/latest/user/network-arch/example.html#configuring-network-interfaces | 11:09 |
| jpw_ | i just want to get this demo out the door | 11:09 |
| jpw_ | i timed it yesterday i've got the whole deployment from blank devices to functioning (although not integrated, yet) ceph + openstack down to 1 hour 45 minutes | 11:10 |
| noonedeadpunk | it is very simplified version jsut as a showcase, actual copnfiguration might get very complex very far, especially when you want to do IP calculations and assignment and make it a very universal between different environments | 11:10 |
| noonedeadpunk | but if you configure that with PXE or smth like that - you might not need that actually.... | 11:11 |
| noonedeadpunk | kinda depends on need to change networking during runtime | 11:11 |
| noonedeadpunk | jpw_: also try out maxing our ANSIBLE_FORKS env var. | 11:12 |
| noonedeadpunk | as default is calculated based on amount of CPUs on deploy host, with max being 10 | 11:12 |
| noonedeadpunk | as this is what sshd would limit by default | 11:12 |
| noonedeadpunk | going slightly above might be beneficial for LXC scenario, especially step of initial LXC creation | 11:13 |
| jpw_ | i've gone pxe that leaves the hosts on a provisioning network and a playbook that takes a simple database of serial numbers, vlanid's to produce the the interfaces file. runtime reconfiguration hasn#t been considered yet but i'm expecting to be able to drop in any additional configs in to interfaces.d | 11:13 |
| jpw_ | the hosts get rebooted. i run some more prep on them and hand everything off to osa | 11:14 |
| noonedeadpunk | well, if you have jsut pxe as default - I'd really think about managing them with systemd_networkd role which this example uses | 11:14 |
| noonedeadpunk | as ifupdown I think is still very messy in terms of changes of networking during runtime... | 11:15 |
| jpw_ | i can give it a go. i'm currently using debian i don't think i've ever tried switching to systemd-networkd before | 11:15 |
| jpw_ | yeah it is somewhat minimal | 11:15 |
| noonedeadpunk | well, with lxc you're getting systemd-networkd anyway in there iirc | 11:16 |
| noonedeadpunk | (or maybe that's only inside containers, cant' recall) | 11:16 |
| jpw_ | i did see it being installed to the host too just to configure the lxc bridge | 11:17 |
| noonedeadpunk | yeah, ok, I think you're right | 11:17 |
| noonedeadpunk | it's also somehow handy on ubuntu, as netplan is using systemd-networkd as a backend anyway, and we can apply some specific overrides even to interfaces managed via netplan. But yers, ifupdown is very different | 11:18 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_keystone master: Use standalone httpd role https://review.opendev.org/c/openstack/openstack-ansible-os_keystone/+/948338 | 11:29 |
| jpw_ | specifying storage_hosts did the trick. thanks | 11:34 |
| noonedeadpunk | sure, no worries, ping us if anything :) | 12:00 |
| jrosser | i think for networking with PXE you should take some care to make sure that what you configure inside the PXEboot can then be built upon to become the host networking you want, rather than need to be replaced | 12:29 |
| jrosser | so if the boot interface eventually needs to be part of a bond, put in work to make the bond exist during PXEboot | 12:29 |
| jrosser | we currently don't do this and have an extra reboot needed as part of initial provisioning and thats quite awkward | 12:30 |
| jpw_ | i 100% will be revisiting that if this is selected for furthur development | 12:30 |
| jpw_ | admittedly im not sure what's great for pxw now days. we used to use puppetlabs-razor that worked really well until it was deprecated | 12:31 |
| jpw_ | s/pxw/pxe/ | 12:31 |
| jpw_ | ive just come up against CORS in firefox preventing image uploads. works ok in chrome. | 12:34 |
| jpw_ | nvm i see the headers are configurable. | 12:36 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_keystone master: Use standalone httpd role https://review.opendev.org/c/openstack/openstack-ansible-os_keystone/+/948338 | 12:45 |
| noonedeadpunk | #startmeeting openstack_ansible_meeting | 15:00 |
| opendevmeet | Meeting started Tue Apr 29 15:00:14 2025 UTC and is due to finish in 60 minutes. The chair is noonedeadpunk. Information about MeetBot at http://wiki.debian.org/MeetBot. | 15:00 |
| opendevmeet | Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. | 15:00 |
| opendevmeet | The meeting name has been set to 'openstack_ansible_meeting' | 15:00 |
| noonedeadpunk | #topic rollcall | 15:00 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_keystone master: Use standalone httpd role https://review.opendev.org/c/openstack/openstack-ansible-os_keystone/+/948338 | 15:00 |
| noonedeadpunk | o/ | 15:01 |
| NeilHanlon | o/ | 15:01 |
| NeilHanlon | (triple booked this morning 🙃) | 15:01 |
| noonedeadpunk | wow | 15:02 |
| noonedeadpunk | that's being harsh | 15:02 |
| NeilHanlon | yeah i am really mean to myself lol | 15:02 |
| NeilHanlon | i forgot about a webinar training i'd signed up for a month ago hah | 15:02 |
| NeilHanlon | heya DavidGomez_! :D | 15:03 |
| NeilHanlon | glad you could make it :) | 15:03 |
| DavidGomez_ | Happy to be here and hopefully help :) | 15:03 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_keystone master: Use standalone httpd role https://review.opendev.org/c/openstack/openstack-ansible-os_keystone/+/948338 | 15:04 |
| noonedeadpunk | hey o/ | 15:04 |
| jrosser | o/ hello | 15:04 |
| noonedeadpunk | #topic office hours | 15:04 |
| noonedeadpunk | So I think mainly we need to sync on state of things for releasing Epoxy | 15:04 |
| noonedeadpunk | Deprecation of the test repo | 15:05 |
| noonedeadpunk | Governance patch was merged: https://review.opendev.org/c/openstack/governance/+/947622 | 15:05 |
| noonedeadpunk | We have only requirements and releases patches left | 15:05 |
| noonedeadpunk | But as rest are merged - there's really a good progress made | 15:06 |
| noonedeadpunk | though - we were pointed to zuul errors which repo creates | 15:06 |
| noonedeadpunk | #link https://zuul.opendev.org/t/openstack/config-errors?project=openstack%2Fopenstack-ansible-tests&skip=0 | 15:06 |
| noonedeadpunk | I was gonna to check on that after the meeting | 15:07 |
| noonedeadpunk | HTPD role: https://review.opendev.org/q/topic:%22osa/httpd_role%22 | 15:07 |
| noonedeadpunk | So patch for Horizon looks good enough atm, and also keystone patch was passing as well | 15:07 |
| noonedeadpunk | But I'd really love to ask someone with a sandbox and a working federation to check it as well | 15:08 |
| noonedeadpunk | given how crucial keystone is... | 15:08 |
| jrosser | hrrm yes interesting | 15:08 |
| jrosser | tbh our staging env does not have federation enabled | 15:09 |
| jrosser | which is a known gap we have | 15:09 |
| noonedeadpunk | ouch | 15:09 |
| jrosser | how long do we have? | 15:09 |
| noonedeadpunk | well... I have one env with federation, but I don't have an actuall access to keycloack behind it to test anything... | 15:10 |
| noonedeadpunk | so the due date is 2nd of June or smth like that | 15:10 |
| noonedeadpunk | so we technically do have a month | 15:10 |
| noonedeadpunk | all in all | 15:10 |
| jrosser | maybe i can look at it next week | 15:11 |
| noonedeadpunk | but it would be really nice to get it out before the due date... | 15:11 |
| noonedeadpunk | that would be highly appreciated | 15:11 |
| noonedeadpunk | as we have no testing around federation at all | 15:11 |
| noonedeadpunk | Linux Bridge Removal | 15:12 |
| noonedeadpunk | I did not check on Zun yet to be honest... | 15:12 |
| NeilHanlon | my month will be probably pretty crazy with Rocky 9.5 and 10... | 15:12 |
| noonedeadpunk | we need 10 asap :) | 15:12 |
| noonedeadpunk | pretty much one of the first things I want to look at after we release Epoxy | 15:13 |
| noonedeadpunk | Talking about EL, I've fixed distro setup for cinder in https://review.opendev.org/c/openstack/openstack-ansible-os_cinder/+/947142 | 15:14 |
| noonedeadpunk | Unpacking images, which was used for Magnum coreos job was merged, so we're good in that regard | 15:15 |
| noonedeadpunk | Ceph/Ubuntu Noble - I've proposed a patch: https://review.opendev.org/c/openstack/openstack-ansible/+/948255 | 15:15 |
| noonedeadpunk | we actually were already having smth like that around antelope I believe | 15:15 |
| noonedeadpunk | I got a comment in one of bug reports, that it would be good to better document it | 15:16 |
| noonedeadpunk | But I am really not sure where and how | 15:16 |
| noonedeadpunk | So if anybody has an idea of better documentation - that will be very nice | 15:16 |
| jrosser | every so often we do get a super fast test node | 15:17 |
| noonedeadpunk | yeah, like ~30min is o_O | 15:17 |
| NeilHanlon | thats weird ... | 15:18 |
| noonedeadpunk | it makes me sometimes worried if everything there is even executed :D | 15:18 |
| noonedeadpunk | but these are AMD machines from what I've learned | 15:19 |
| noonedeadpunk | I think one last thing, which might worth looking into - is fixing Manila CI. As iirc Ceph does have some in-tree replacement for ganesha, which is pretty much mgr module | 15:20 |
| noonedeadpunk | wanna look into that and see if/how to implement that instead of installing ganesha separately | 15:21 |
| NeilHanlon | makes sense.. | 15:22 |
| NeilHanlon | btw I did want to introduce DavidGomez_ -- he's a coworker of mine who was interested in getting involved and helping with (among other things) Rocky 10 | 15:23 |
| noonedeadpunk | nice, that is always very welcome :) | 15:23 |
| noonedeadpunk | now we can bother him as well with our stupid questions?:) | 15:24 |
| DavidGomez_ | Hi all o/ what Neil said will help where I can and ask away I say | 15:25 |
| noonedeadpunk | \o/ | 15:25 |
| noonedeadpunk | these are great news | 15:26 |
| noonedeadpunk | On other thing - I think we need to step in and help fixing upgrade jobs for Swift role: https://review.opendev.org/c/openstack/openstack-ansible-os_swift/+/946990 | 15:27 |
| noonedeadpunk | it's cuurrently failing only on Rocky upgrade jobs | 15:27 |
| noonedeadpunk | I'm not sure if I will be able to check this week, but probably next week... | 15:27 |
| noonedeadpunk | I pretty much wonder what's the difference in there making rocky fail and specifically on upgrade... | 15:28 |
| noonedeadpunk | smth with SSHD and ssh certificates for sure... | 15:29 |
| noonedeadpunk | anyway | 15:29 |
| noonedeadpunk | that was pretty much all from my side | 15:29 |
| NeilHanlon | nothing more from me | 15:29 |
| noonedeadpunk | it would be really nice if we could come up with role branching patch by end of next week... | 15:31 |
| noonedeadpunk | but as of today - quite some things to land still | 15:31 |
| noonedeadpunk | not too much, but some | 15:31 |
| noonedeadpunk | if that's it, I'd love to conclude the meeting then :) | 15:35 |
| NeilHanlon | Sounds good! | 15:36 |
| noonedeadpunk | #endmeeting | 15:37 |
| opendevmeet | Meeting ended Tue Apr 29 15:37:16 2025 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | 15:37 |
| opendevmeet | Minutes: https://meetings.opendev.org/meetings/openstack_ansible_meeting/2025/openstack_ansible_meeting.2025-04-29-15.00.html | 15:37 |
| opendevmeet | Minutes (text): https://meetings.opendev.org/meetings/openstack_ansible_meeting/2025/openstack_ansible_meeting.2025-04-29-15.00.txt | 15:37 |
| opendevmeet | Log: https://meetings.opendev.org/meetings/openstack_ansible_meeting/2025/openstack_ansible_meeting.2025-04-29-15.00.log.html | 15:37 |
| * noonedeadpunk just learned about https://opendev.org/openstack/ansible-plugin-container-connection | 17:50 | |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_manila master: Switch Manila jobs to Noble https://review.opendev.org/c/openstack/openstack-ansible-os_manila/+/948474 | 18:43 |
| drarvese | Greetings all, I've been lurking and following things that jpw_ was doing yesterday. I'm in a similar situation struggling to get a new OVN installation setup. | 19:02 |
| drarvese | Here's a paste: https://paste.openstack.org/show/bCdE99YVzsWIbsxz1gHx/ -- does this look OK? | 19:02 |
| drarvese | It sounds like br-vlan isn't needed on the host and OSA will add the physical interface to the br-vlan bridge within OVN. Is that correct? | 19:03 |
| noonedeadpunk | drarvese: hey | 20:13 |
| noonedeadpunk | yes, that is correct | 20:13 |
| noonedeadpunk | this bridge does not need to be created as it will be an OVS bridge created by the role | 20:14 |
| drarvese | OK, that's good to know | 20:16 |
| opendevreview | Ivan Anfimov proposed openstack/openstack-ansible master: wip https://review.opendev.org/c/openstack/openstack-ansible/+/948111 | 20:17 |
| noonedeadpunk | drarvese: I think you can also check some configuration from our CI jobs to see the resulting config | 20:20 |
| noonedeadpunk | ie https://60667c112fa084ed347c-8bf91e93364f14a67a943838d2417a14.ssl.cf5.rackcdn.com/openstack/2f96f4d5eb934882a7a182f9aeccab8a/logs/etc/host/openstack_deploy/ | 20:20 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible master: Switch to using mariadb.org infra cache https://review.opendev.org/c/openstack/openstack-ansible/+/934038 | 20:26 |
| drarvese | Everything installs correctly but creating an instance times out trying to allocate the network. I can see the port get created and attached to bridge but it eventually times out | 20:29 |
| drarvese | This is a VLAN network FWIW. I'm unable to ping the gateway from within the network namespace that gets spawned | 20:32 |
| opendevreview | Ivan Anfimov proposed openstack/openstack-ansible master: docs: update openstackdocstheme to 3.5.0 https://review.opendev.org/c/openstack/openstack-ansible/+/948111 | 21:08 |
| opendevreview | Ivan Anfimov proposed openstack/openstack-ansible-os_horizon master: Remove tags from README https://review.opendev.org/c/openstack/openstack-ansible-os_horizon/+/948411 | 22:22 |
| opendevreview | Ivan Anfimov proposed openstack/openstack-ansible-os_trove master: Remove tags from README https://review.opendev.org/c/openstack/openstack-ansible-os_trove/+/948372 | 22:44 |
| opendevreview | Ivan Anfimov proposed openstack/openstack-ansible-os_keystone master: Remove tags from README https://review.opendev.org/c/openstack/openstack-ansible-os_keystone/+/948412 | 22:48 |
| opendevreview | Ivan Anfimov proposed openstack/openstack-ansible-plugins master: Remove tags from README https://review.opendev.org/c/openstack/openstack-ansible-plugins/+/948414 | 22:49 |
| opendevreview | Ivan Anfimov proposed openstack/openstack-ansible-os_magnum master: Remove tags from README https://review.opendev.org/c/openstack/openstack-ansible-os_magnum/+/948415 | 22:49 |
| opendevreview | Ivan Anfimov proposed openstack/openstack-ansible-rabbitmq_server master: Remove tags from README https://review.opendev.org/c/openstack/openstack-ansible-rabbitmq_server/+/948410 | 22:49 |
| opendevreview | Ivan Anfimov proposed openstack/openstack-ansible-os_cinder master: Remove tags from README https://review.opendev.org/c/openstack/openstack-ansible-os_cinder/+/948413 | 22:50 |
| opendevreview | Ivan Anfimov proposed openstack/openstack-ansible-ops master: Remove tags from README https://review.opendev.org/c/openstack/openstack-ansible-ops/+/948393 | 22:50 |
| opendevreview | Ivan Anfimov proposed openstack/openstack-ansible master: Remove tags from README https://review.opendev.org/c/openstack/openstack-ansible/+/948394 | 22:51 |
| opendevreview | Ivan Anfimov proposed openstack/openstack-ansible-os_horizon master: Remove tags from README https://review.opendev.org/c/openstack/openstack-ansible-os_horizon/+/948411 | 22:53 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!