| opendevreview | Michal Nasiadka proposed openstack/kolla-ansible master: CI: Change cephadm jobs to voting and add them to gating https://review.opendev.org/c/openstack/kolla-ansible/+/912585 | 07:49 |
|---|---|---|
| mnasiadka | frickler: https://review.opendev.org/c/openstack/kolla/+/920279 - I guess I applied your comments ;-) | 07:56 |
| frickler | mnasiadka: but why do we need the requests pin again? I looked at that yesterday but couldn't understand it | 08:06 |
| mnasiadka | frickler: hmm, the jobs were failing without this - because after we unpin - we still don't have packaging in requirements (and newer docker doesn't have packaging in requirements) | 08:15 |
| mnasiadka | but it should work in the same change, let me try again | 08:15 |
| opendevreview | Michal Nasiadka proposed openstack/kolla master: Drop packaging requirement and unpin docker-py https://review.opendev.org/c/openstack/kolla/+/920279 | 08:16 |
| opendevreview | Michal Nasiadka proposed openstack/kolla master: Drop packaging requirement and unpin docker-py https://review.opendev.org/c/openstack/kolla/+/920279 | 08:19 |
| mnasiadka | frickler: I have a plan to rework kolla pre.yml to be a separate role that we could call from kolla-ansible - so that build process would be identical in both repos - now it's different | 08:22 |
| frickler | mnasiadka: ah, I see, yes, that would be a good idea | 08:22 |
| sylvr | What could cause nodes to be stuck in the "enroll" provisionning state ? And I tried to put idrac-redfish as the default driver for ironic but except 2 nodes, all of them are configured with ipmi driver, what could be my mistake ? | 08:29 |
| jovial[m] | sylvr, It normally means they are failing to validate. Possibly bad credentials or SSL certificate verification failing? Check the conductor logs. | 08:31 |
| mnasiadka | frickler: seems it's passing now | 08:47 |
| sylvr | jovial: checking /var/log/kolla/ironic/ironic.log and I can see "No conductor from group <none> found for driver ipmi" which doesn't surprise me as I don't want to use ipmi but idrac... but how can I reset all my nodes drivers ? | 08:48 |
| jovial[m] | sylvr: You can update the driver with baremetal node set, or just delete and re-enroll if they are not provisioned. | 08:49 |
| sylvr | okay ! | 08:49 |
| mnasiadka | frickler: https://review.opendev.org/c/openstack/kolla/+/920279 would need backporting to C | 08:50 |
| frickler | mnasiadka: ack | 08:55 |
| sylvr | jovial: couldn't do either of it since it couldn't find a conductor eligible or an available driver, I redeployed the seed with updated config in hope it will fix this | 08:59 |
| SvenKieske | +1 on the seperate role idea mnasiadka :) | 09:04 |
| SvenKieske | maybe https://review.opendev.org/c/openstack/kolla-ansible/+/913993 deserves RP+1; I think the change to the default behaviour is more of an accident and probably should be fixed | 09:07 |
| SvenKieske | mhm, don't we have a problem providing docker>=6 anymore on debian, thinking about it? https://review.opendev.org/c/openstack/ansible-collection-kolla/+/916258 | 09:11 |
| SvenKieske | okay I guess that is only relevant to non venv envs | 09:14 |
| SvenKieske | Thinking about it, it would be good to push everything into venvs, not because I dislike debian packages (I don't), but the benefit would be that we have the same versions, same setups process, same bugs, on every OS. | 09:15 |
| SvenKieske | minus distribution introduced venv bugs, of course ;) | 09:16 |
| mnasiadka | push everything? well, I was thinking on reworking the CI to use venvs always - but we can't really force users to do that :) | 09:30 |
| SvenKieske | we really should merge https://review.opendev.org/c/openstack/kolla-ansible/+/913908 btw for users using keystone federation, the backend seems totally borked there | 09:30 |
| SvenKieske | mnasiadka: ok, one step at a time then :) | 09:30 |
| SvenKieske | and no, I didn't want to do everything at once, I really like making smaller, contained changes, easier to debug and revert :) | 09:31 |
| SvenKieske | mnasiadka, frickler: I added https://bugs.launchpad.net/kolla-ansible/+bug/2058656 (patch is linked there) to the whiteboard, please review. I will be on PTO starting tomorrow up to (and including) monday next week. so I will not be present in kolla meeting | 09:42 |
| SvenKieske | but the bugreport is pretty evident imho and doesn't need me explaining it I think :) | 09:43 |
| SvenKieske | shout if this is still needed, I'm actually not sure anymore myself :D https://review.opendev.org/c/openstack/kolla/+/920261 | 09:56 |
| SvenKieske | frickler: your +2 is gone here :) https://review.opendev.org/c/openstack/kolla/+/920279 | 10:05 |
| frickler | SvenKieske: yes, I know, waiting for CI results | 10:07 |
| SvenKieske | yes, that seems like a good idea given the weird state of the last few days :) just thought you might have missed the new patchset | 10:08 |
| mnasiadka | frickler: https://review.opendev.org/c/openstack/kolla/+/920279 - CI results are in | 10:29 |
| frickler | ok, let's hope that this really puts an end to this issue | 10:48 |
| kevko | \o | 11:33 |
| tafkamax | Hi, I am setting up an openstack instance with kolla-ansible. And I've come into a bit of an 'question' with networking. I have 2 x Mellanox connectx6 with 2x25 ports. One is for openstack/ceph operations and the second is for tenant/provider network for VM-s. After configuring `neutron_external_interface=bond1` I get this state -... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/lBzZqlSAUYDsIEgiiiivDTnd>) | 11:39 |
| tafkamax | The network itself - provider or corps intranet, does not allow non-vlan tagged stuff. So I was thinking - do I need to add VLAN-s to the bond? E.g. bond0.10@bond0 for VLAN 10 access | 11:45 |
| kevko | Hi guys .. can anyone give some advice ? https://28d53a25c5675b454737-9cb29527c5d80ca3ff28f256812f71ef.ssl.cf2.rackcdn.com/920368/45/check/kolla-ansible-ubuntu/63c4f02/primary/logs/ansible/test-core-openstack << on the end i see return is 0 for test and then script exited with 1 ? | 11:59 |
| SvenKieske | kevko: Hi, looking into it via https://zuul.opendev.org/t/openstack/build/63c4f021dd444677b341ba6b80c5252d/log/primary/logs/ansible/test-core-openstack | 12:03 |
| SvenKieske | I guess you noticed tox linter is also yelling at you? :) | 12:04 |
| SvenKieske | kevko, maybe found it: "/run/cirros/datasource/data/user-data was not '#!' or executable" | 12:09 |
| SvenKieske | kevko: and way before that (before the log dump from cirros): | 12:10 |
| SvenKieske | "Failed to access server via SSH after 12 attempts" | 12:10 |
| SvenKieske | so afaik cirros can't configure dropbear/ssh because user-data is garbage and then ssh test to cirros fails | 12:10 |
| kevko | SvenKieske: yeaah ...tox is non interisting for now :) | 12:10 |
| kevko | hmmmm | 12:11 |
| SvenKieske | I don't know why the test runs onwards then thought :D | 12:11 |
| SvenKieske | shouldn't that halt all test execution? | 12:11 |
| opendevreview | Merged openstack/kolla master: Drop packaging requirement and unpin docker-py https://review.opendev.org/c/openstack/kolla/+/920279 | 12:12 |
| SvenKieske | it then goes on with "kolla-boot-test login: " | 12:12 |
| kevko | SvenKieske: well, it's lot of work to remove bashes | 12:13 |
| kevko | SvenKieske: for now i want to move bash setup to ansible ... | 12:13 |
| opendevreview | Michal Nasiadka proposed openstack/kolla stable/2024.1: Drop packaging requirement and unpin docker-py https://review.opendev.org/c/openstack/kolla/+/920628 | 12:13 |
| mnasiadka | jovial[m]: I assume Kayobe patch should now pass | 12:14 |
| mnasiadka | jovial[m]: It would be good to get it merged today and then post rc1 tomorrow | 12:14 |
| mnasiadka | I'm off Thu and Fri, so you would need to chase getting a final release by Monday | 12:14 |
| SvenKieske | kevko: sure :) It just returns 1 if function has an error and the bash test goes happily onwards to the next test lol: https://github.com/openstack/kolla-ansible/blob/master/tests/test-core-openstack.sh#L372 | 12:14 |
| SvenKieske | function is "test_ssh" | 12:14 |
| SvenKieske | when will we be able to switch to tempest again? | 12:15 |
| kevko | SvenKieske: again ? we never had tempest :D | 12:16 |
| kevko | SvenKieske: thank you ..i probably know where is problem :) | 12:17 |
| kevko | SvenKieske: as you know ...i had a chain where i have tempests implemented ... | 12:17 |
| kevko | but mnasiadka didn't want to merge my change for init-runonce bash script ...so i need to rewrite it :D | 12:18 |
| kevko | but challenge accepted | 12:18 |
| SvenKieske | kevko, let me test something.. | 12:18 |
| kevko | SvenKieske: where ? when ? | 12:18 |
| opendevreview | Michal Nasiadka proposed openstack/kolla stable/2024.1: CI: pin codespell to <2.3.0 https://review.opendev.org/c/openstack/kolla/+/920629 | 12:19 |
| opendevreview | Michal Nasiadka proposed openstack/kolla stable/2023.2: CI: pin codespell to <2.3.0 https://review.opendev.org/c/openstack/kolla/+/920630 | 12:19 |
| opendevreview | Michal Nasiadka proposed openstack/kolla stable/2024.1: Drop packaging requirement and unpin docker-py https://review.opendev.org/c/openstack/kolla/+/920628 | 12:20 |
| opendevreview | Michal Arbet proposed openstack/kolla-ansible master: Bash replacement https://review.opendev.org/c/openstack/kolla-ansible/+/920368 | 12:21 |
| opendevreview | Michal Arbet proposed openstack/kolla-ansible master: Bash replacement https://review.opendev.org/c/openstack/kolla-ansible/+/920368 | 12:24 |
| kevko | okay, let's see | 12:25 |
| SvenKieske | the set +erronexit should prevent the execution from going on, because test_ssh returns "1" in case of error | 12:26 |
| SvenKieske | it actually does, the rest of the output is the result from openstack --debug server show $foo | 12:32 |
| SvenKieske | so yeah, the cirros system can't get instance metadata via metadata service (I believe that is intentionally disabled in CI?) and the supplied metadata via script is malformed somehow. | 12:33 |
| kevko | SvenKieske: you gave right direction :) ..i've fixed a code | 12:33 |
| SvenKieske | nice :) | 12:33 |
| kevko | SvenKieske: nope, i think everything is working as expected ...i would say ... | 12:33 |
| SvenKieske | but it's hard to debug I have to agree, the outputs are confusing in that script :D | 12:33 |
| kevko | SvenKieske: but i reworked init-runonce | 12:33 |
| kevko | SvenKieske: first thing is that it's in ansible...but because it support parameter for user ..i need to generate and then upload ssh key to openstack as ssh-suffix where suffix = user | 12:34 |
| kevko | so in that case ~/.ssh/id_ecdsa_admin | 12:35 |
| kevko | on the end i want to test openstack under regular project ... | 12:35 |
| kevko | because now we are testing under admin which is bad from the beginning i would say ... | 12:35 |
| kevko | because of policy.yaml ....aaaaand in code from time to time it's harcoded is_admin function to check if admin ... | 12:36 |
| kevko | it was worse in previous version ..btu maybe still somewhere | 12:36 |
| SvenKieske | yeah, I agree better to test in normal project, like normal production workload is also tested in normal project I would say | 12:39 |
| kevko | SvenKieske: yep | 12:48 |
| kevko | SvenKieske: so i reworked to create external net/subnet in admin (shared) and admin-subnet by default ... BUT demo-net in demo project with separate ssh key and all with -suffix = project | 12:49 |
| kevko | so that's the reason probably | 12:49 |
| kevko | SvenKieske: and i also found a bug with venv debugged 2 days :D | 12:49 |
| kevko | SvenKieske: thanks, it helped :) https://zuul.opendev.org/t/openstack/stream/daffab8a4f6f49c1adcba356d090f0d1?logfile=console.log | 13:05 |
| SvenKieske | nice | 13:18 |
| opendevreview | Matúš Jenča proposed openstack/kolla-ansible master: Implement TLS for Redis https://review.opendev.org/c/openstack/kolla-ansible/+/909188 | 13:18 |
| tafkamax | <tafkamax> "The network itself - provider or..." <- Ok answering myself, as is the standard. This worked. As the trunk port native vlan is set to 1 currently, will test with some other native vlan soon. | 13:24 |
| opendevreview | Matúš Jenča proposed openstack/kolla-ansible master: Implement TLS for Redis https://review.opendev.org/c/openstack/kolla-ansible/+/909188 | 13:33 |
| opendevreview | Matúš Jenča proposed openstack/kolla-ansible master: Add Redis as caching backend for Heat https://review.opendev.org/c/openstack/kolla-ansible/+/909224 | 13:48 |
| opendevreview | Matúš Jenča proposed openstack/kolla-ansible master: Add Redis as caching backend for Placement https://review.opendev.org/c/openstack/kolla-ansible/+/909222 | 13:51 |
| opendevreview | Merged openstack/kolla stable/2024.1: CI: pin codespell to <2.3.0 https://review.opendev.org/c/openstack/kolla/+/920629 | 14:15 |
| opendevreview | Michal Nasiadka proposed openstack/kolla master: WIP: Switch to Ubuntu 24.04 https://review.opendev.org/c/openstack/kolla/+/907589 | 14:45 |
| mnasiadka | frickler: you probably will know - what's the plan regarding pbr and distutils missing in py312? | 15:20 |
| frickler | mnasiadka: just happened to do this yesterday and fungi has now added fixes https://review.opendev.org/c/openstack/pbr/+/920595 | 15:24 |
| mnasiadka | ah ok, kolla-build works with recent setuptools | 15:25 |
| fungi | mnasiadka: frickler: my change series is adding setuptools to pbr's testing. i think projects using pbr (or any setuptools extension) need setuptools preinstalled in venvs when using the venv module from python 3.12 | 15:26 |
| opendevreview | Michal Nasiadka proposed openstack/kolla master: WIP: Switch to Ubuntu 24.04 https://review.opendev.org/c/openstack/kolla/+/907589 | 15:26 |
| fungi | it's not a pbr-specific problem | 15:26 |
| mnasiadka | yeah, just noticed installing recent setuptools works | 15:27 |
| mnasiadka | (or at least gets past the pbr usage) | 15:27 |
| fungi | fwiw, our "modern" (pep 517) packaging example for pbr recommends listing setuptools explicitly as a build-system.requires for projects | 15:28 |
| mnasiadka | fungi: is that only on the ML (https://lists.openstack.org/archives/list/openstack-discuss@lists.openstack.org/message/HVFN5RBSHRTM3B2UUKPAWKH6H6AT6CYR/) or is there some doc entry in pbr docs? (trying to find it in pbr docs but it seems I'm blind) | 15:32 |
| mnasiadka | ah - here: https://docs.openstack.org/pbr/latest/user/using.html#pyproject-toml | 15:32 |
| opendevreview | Michal Nasiadka proposed openstack/kolla master: Add pyproject.toml to support pip 23.1 https://review.opendev.org/c/openstack/kolla/+/920684 | 15:37 |
| opendevreview | Michal Nasiadka proposed openstack/kolla master: WIP: Switch to Ubuntu 24.04 https://review.opendev.org/c/openstack/kolla/+/907589 | 15:38 |
| opendevreview | Michal Nasiadka proposed openstack/kolla master: WIP: Switch to Ubuntu 24.04 https://review.opendev.org/c/openstack/kolla/+/907589 | 15:38 |
| fungi | mnasiadka: yeah, in the docs (sorry, on a conference call at the moment) | 15:41 |
| fungi | though i guess that only addresses pbr when used at build/install time, not runtime. for that we're talking about adding setuptools as a dependency of pbr for python 3.12 and later, since that's the only way to get distutils | 15:42 |
| opendevreview | Michal Nasiadka proposed openstack/kolla master: WIP: Switch to Ubuntu 24.04 https://review.opendev.org/c/openstack/kolla/+/907589 | 15:59 |
| opendevreview | Kevin Tindall proposed openstack/kolla-ansible master: Add TLS proxy for novncproxy https://review.opendev.org/c/openstack/kolla-ansible/+/911141 | 16:38 |
| opendevreview | Michal Nasiadka proposed openstack/kolla master: WIP: Switch to Ubuntu 24.04 https://review.opendev.org/c/openstack/kolla/+/907589 | 17:25 |
| opendevreview | Michal Nasiadka proposed openstack/kolla master: WIP: Switch to Ubuntu 24.04 https://review.opendev.org/c/openstack/kolla/+/907589 | 18:10 |
| opendevreview | Rafal Lewandowski proposed openstack/kayobe master: Add ironic config files clean up tasks to overcloud deprovision https://review.opendev.org/c/openstack/kayobe/+/920546 | 18:30 |
| opendevreview | Merged openstack/kayobe master: Bump up Ansible supported versions to 8.x/9.x https://review.opendev.org/c/openstack/kayobe/+/910513 | 19:25 |
| opendevreview | Michal Arbet proposed openstack/kolla-ansible master: Bash replacement https://review.opendev.org/c/openstack/kolla-ansible/+/920368 | 22:55 |
| opendevreview | Merged openstack/kolla stable/2024.1: Drop packaging requirement and unpin docker-py https://review.opendev.org/c/openstack/kolla/+/920628 | 23:39 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!