| *** dpaclt has quit IRC | 00:17 | |
| *** watersj has quit IRC | 01:33 | |
| *** schwicht has quit IRC | 01:48 | |
| *** schwicht has joined #openstack-ansible | 01:51 | |
| *** spatel has joined #openstack-ansible | 01:56 | |
| *** tinwood has quit IRC | 02:12 | |
| *** KeithMnemonic has quit IRC | 02:14 | |
| *** tinwood has joined #openstack-ansible | 02:14 | |
| *** djhankb has quit IRC | 02:20 | |
| *** stingrayza has joined #openstack-ansible | 02:34 | |
| *** also_stingrayza has quit IRC | 02:38 | |
| *** djhankb has joined #openstack-ansible | 02:47 | |
| *** redrobot has quit IRC | 02:49 | |
| *** nurdie has joined #openstack-ansible | 02:57 | |
| *** itandops has quit IRC | 03:08 | |
| *** rh-jelabarre has quit IRC | 04:07 | |
| *** evrardjp has quit IRC | 04:33 | |
| *** evrardjp has joined #openstack-ansible | 04:33 | |
| *** spatel has quit IRC | 04:54 | |
| *** cshen has joined #openstack-ansible | 05:00 | |
| *** cshen has quit IRC | 05:05 | |
| *** nurdie has quit IRC | 05:20 | |
| *** udesale has joined #openstack-ansible | 05:29 | |
| *** raukadah is now known as chandankumar | 05:32 | |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible master: [WIP] Make upgrade jobs get role relevant versions https://review.opendev.org/710849 | 05:42 |
|---|---|---|
| *** cshen has joined #openstack-ansible | 05:45 | |
| *** schwicht has quit IRC | 06:06 | |
| openstackgerrit | Jonathan Rosser proposed openstack/ansible-role-python_venv_build stable/stein: Verify version number for virtualenv>=20.0.0 https://review.opendev.org/729844 | 06:13 |
| *** udesale has quit IRC | 06:31 | |
| *** udesale has joined #openstack-ansible | 06:33 | |
| *** udesale has quit IRC | 06:38 | |
| *** udesale has joined #openstack-ansible | 06:38 | |
| *** dpaclt has joined #openstack-ansible | 07:07 | |
| *** threestrands has quit IRC | 07:08 | |
| dpaclt | Hi All , Just wanted to reconfirm whether to define storage_hosts as controller/infrastructure_hosts if my backend is Ceph . | 07:13 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-tests master: Add Ubuntu Focal non voting check jobs https://review.opendev.org/729854 | 07:14 |
| janno | dpaclt: do you want to deploy the ceph backend through openstack-ansible or does the cluster already exist? | 07:15 |
| dpaclt | I want to deploy ceph backend through openstack-ansible ,dont have another cluster | 07:15 |
| janno | we deployed our ceph cluster with openstack-ansible we have groups ceph-mon_hosts and ceph-osd_hosts | 07:16 |
| janno | where ceph-mon_hosts are the infrastructure hosts and ceph-osd_hosts are the actual storage hosts | 07:17 |
| dpaclt | We have defined ceph-mon and ceph -osd hosts as you said ... | 07:17 |
| dpaclt | .Can see storage hosts defined as storage_hosts: *infrastructure_hosts in https://docs.openstack.org/openstack-ansible/latest/user/ceph/full-deploy.html | 07:18 |
| janno | ah, yes, you are right. we have storage_hosts defined as well | 07:19 |
| janno | storage_hosts: *infrastructure_hosts | 07:20 |
| *** NewJorg has quit IRC | 07:21 | |
| dpaclt | oh ok thanks for the confirmation. | 07:21 |
| dpaclt | I was intending to use native ceph as backend for manila ... so manila_data_hosts and manila_infra_hosts also should be *infrastructure_hosts right ? | 07:22 |
| *** NewJorg has joined #openstack-ansible | 07:23 | |
| dpaclt | So was wondering whether manila_data_hosts is infrastructure_hosts or ceph -osd hosts ? | 07:30 |
| *** tosky has joined #openstack-ansible | 07:31 | |
| noonedeadpunk | I think yes... But had only metal deployment on hands so it didn't matter in my case... | 07:34 |
| noonedeadpunk | jrosser: ok, seems that we need to disable ci anyway to merge https://review.opendev.org/#/c/728434/3 | 07:34 |
| jrosser | dpaclt: storage_hosts may be referring to API like cinder, not just the actual storage itself | 07:49 |
| jrosser | noonedeadpunk: o/ morning, let me look at the galera thing | 07:49 |
| dpaclt | jrosser ohh ok then it can be infrastructure node i guess | 07:49 |
| noonedeadpunk | Yeah, I think it all depends if driver_handles_share_servers is true or false | 07:50 |
| dpaclt | mine was driver_handles_share_servers is false | 07:51 |
| jrosser | noonedeadpunk: so yes this is circular isn't it? becasue the galera_server role uses the integrated repo in it's tests, which itself needs updates for the changes to galera_server | 07:53 |
| noonedeadpunk | yeah. and we're failing because of utility | 07:54 |
| jrosser | and theres no point adding depends-on becasue it just makes it worse | 07:54 |
| noonedeadpunk | So thinking about just dropping lxc jobs for now | 07:54 |
| *** cshen has quit IRC | 07:55 | |
| jrosser | and the metal jobs succeed becasue galera_server has previously run against the host which does what utility would have needed | 07:55 |
| jrosser | is that right? | 07:56 |
| noonedeadpunk | YEah, I think that's what happen | 07:57 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible master: Use galera server role to install galera client https://review.opendev.org/728434 | 07:57 |
| jrosser | ok so i had a few more things we should check, the patch to galera_server needs checking that everything (that we care about) from the galera_client role defaults can still be overriden | 07:58 |
| jrosser | that the vars from galera_server that actually apply to both maybe we change the name to make them galera_<foo> rather than galera_server_<foo> | 07:58 |
| jrosser | any role tests from client to move across, (the server ones look pretty good though) | 07:59 |
| jrosser | release note / any docs update | 07:59 |
| jrosser | that first one about checking anything we accidentally drop from galera_client would be great to get second eyes on | 08:00 |
| *** cshen has joined #openstack-ansible | 08:02 | |
| dpaclt | jrosser should we define manila_shares under group bind in br-storage | 08:06 |
| jrosser | dpaclt: i dont know - i've never deployed manila myself | 08:08 |
| jrosser | does the documentation for the manila ansible role help? | 08:09 |
| dpaclt | To be frank no.. i didnt get a proper documentation of manila | 08:10 |
| jrosser | it's also quite complex as irrc there are at least two different ways you can use it | 08:13 |
| jrosser | noonedeadpunk: can you review this as we have a few things still to merge into os_neutron that are now unblocked https://review.opendev.org/#/c/728804/2 | 08:14 |
| * jrosser removes vote from own patch :/ | 08:15 | |
| noonedeadpunk | Maybe you can vote as it's jamesdenton who patched the last time? | 08:16 |
| noonedeadpunk | I feel a bit uncomfortable is such cases tbh | 08:17 |
| jrosser | i'll get gshippey to take a look | 08:17 |
| *** mgariepy has quit IRC | 08:20 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-openstack_hosts master: Fix _package_repos for ubuntu focal https://review.opendev.org/729867 | 08:22 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-tests master: Add Ubuntu Focal non voting check jobs https://review.opendev.org/729854 | 08:23 |
| *** NewJorg has quit IRC | 08:23 | |
| jrosser | noonedeadpunk: i was also thinking about what to do for depends-on patches to the ansible openstack collection, and it's kinda ugly | 08:24 |
| jrosser | because if we set up a job to test a review in the collection with a depends-on, we are saying it's ok to merge an OSA patch when the collection patch merges | 08:25 |
| jrosser | and you would then expect that OSA works outside CI, maybe requiring a fix in the collection to have landed | 08:26 |
| jrosser | but thats not the case if we consume the collection from the published version in ansible galaxy, we'd be always on a released version | 08:27 |
| noonedeadpunk | yeah, I see what you mean. Like not being possible to install collaction just from git is smth I really miss | 08:30 |
| *** NewJorg has joined #openstack-ansible | 08:30 | |
| jrosser | i might look at using the collection from git, not using ansible-galaxy to install | 08:31 |
| jrosser | then treat it the same as one of our roles | 08:31 |
| *** aedc has joined #openstack-ansible | 08:31 | |
| *** aedc has quit IRC | 08:32 | |
| jrosser | thats not documented becasue the assumption is you use galaxy but the file tree it drops down is extremely similar to the original repo | 08:32 |
| noonedeadpunk | I was trying to look isntalling from git but kinda didn't find any docs allowing that to do | 08:43 |
| noonedeadpunk | but yes, using git would kinda resolve she issue and we can kinda just bump to sha and update it with release script | 08:44 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-lxc_hosts master: Use btrfs-progs instead of btrfs-tools https://review.opendev.org/729876 | 08:45 |
| *** cshen has quit IRC | 08:46 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-tests master: Add Ubuntu Focal non voting check jobs https://review.opendev.org/729854 | 08:46 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible master: Return lxc jobs https://review.opendev.org/729878 | 08:50 |
| *** mgariepy has joined #openstack-ansible | 09:04 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-lxc_hosts master: Do not install python2 packages on systems that should be python3 only https://review.opendev.org/729883 | 09:11 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-tests master: Add Ubuntu Focal non voting check jobs https://review.opendev.org/729854 | 09:12 |
| *** cshen has joined #openstack-ansible | 09:14 | |
| *** cshen has quit IRC | 09:19 | |
| openstackgerrit | Duncan Martin Walker proposed openstack/openstack-ansible-ops master: Ensure that the beat setup role runs when elk_beat_setup=true https://review.opendev.org/729654 | 09:20 |
| noonedeadpunk | jrosser: actually I think we can really use just galaxy version. It's anyway more flexible than we do have now | 09:20 |
| noonedeadpunk | Just didn't get about one sed here https://review.opendev.org/#/c/718431/3 | 09:21 |
| jrosser | yeah, looking at that it can be simplified | 09:27 |
| jrosser | i think i was probably thinking about what happens if you have a deploy host managing more than one deployment, which isnt something i ever do | 09:28 |
| *** gshippey has joined #openstack-ansible | 09:30 | |
| noonedeadpunk | but like we have deloyment host for several deployments, but they do share osa version | 09:31 |
| noonedeadpunk | Otherwise all we also need different venvs with different ansible versions wich is not the case at the moment | 09:32 |
| noonedeadpunk | and since you have the same ansible version, I think it's ok to have same collections version as well... | 09:32 |
| *** itandops has joined #openstack-ansible | 09:33 | |
| *** arxcruz is now known as arxcruz|off | 09:33 | |
| noonedeadpunk | btw yes, manila failed for centos7 ceph deployment | 09:35 |
| noonedeadpunk | because of that pyaml thing | 09:36 |
| noonedeadpunk | tbh, I'd drop centos 7 at that moment if only we had centos 8 implemented | 09:36 |
| jrosser | noonedeadpunk: yes centos7 has been on my mind also to drop now | 09:53 |
| jrosser | but i worry that if we make big patches for that, backporting centos-8 support is going to be very bad | 09:54 |
| noonedeadpunk | We can actually maybe ignore octopus issue for centos-7 then? | 09:54 |
| jrosser | for V we have no choice but to drop centos-7 i think | 09:54 |
| jrosser | it's just a question of exactly when we do it | 09:55 |
| noonedeadpunk | Oh, you're right, for U we may keep it | 09:55 |
| jrosser | for U we still seem to have it working by some miracle | 09:55 |
| jrosser | only distro jobs are going to be really difficult for U, like the magnum issue | 09:55 |
| noonedeadpunk | despite packages are for T | 09:56 |
| jrosser | i guess from a python pov we may just be building more from source during the wheel build | 09:56 |
| jrosser | for the source based installs i'm not sure that U distro packages are an issue at all | 09:57 |
| noonedeadpunk | yeah, for source based they do not matter I think | 09:57 |
| jrosser | the thing that will ultimately break centos-7 completely is a not-new-enough libvirt, something really fundamental | 09:58 |
| jrosser | something like that | 09:58 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible master: Update ceph release to octopus https://review.opendev.org/729331 | 10:14 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-os_neutron master: Cleanup after repo_build and pip_install retirement https://review.opendev.org/727401 | 10:24 |
| noonedeadpunk | jrosser: so, should be create branches out of https://review.opendev.org/#/c/729258/ or we want smth to bump here? | 10:27 |
| *** mgariepy has quit IRC | 10:30 | |
| *** dpaclt has quit IRC | 10:34 | |
| *** itandops has quit IRC | 10:34 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_neutron master: networking-ovn is now integrated into neutron https://review.opendev.org/728804 | 10:51 |
| openstackgerrit | Merged openstack/ansible-role-uwsgi master: Do not use bare variables in ansible conditionals https://review.opendev.org/727258 | 10:51 |
| noonedeadpunk | chandankumar: hey! have a question regarding rdo repos. Trying to add https://trunk.rdoproject.org/centos8-master/current-passed-ci/ as repo (used common one for centos 7 | 10:54 |
| noonedeadpunk | but get "Failed to download metadata for repo 'rdo-current-passed-ci'" | 10:54 |
| chandankumar | noonedeadpunk, is it coming everywhere? It might a network glitch | 11:01 |
| chandankumar | can you recheck if not then I will ask jpena to take a look | 11:02 |
| noonedeadpunk | chandankumar: I think it's probably the way we configure it but not sure | 11:02 |
| noonedeadpunk | As repo looks like having some new fromat | 11:02 |
| noonedeadpunk | while we stil define it in /etc/yum.repos.d/rdo.repo | 11:02 |
| noonedeadpunk | Along with https://trunk.rdoproject.org/centos8-master/deps/latest/ wich works | 11:03 |
| noonedeadpunk | As eventually link opens and available, but I think that dnf expects to see there metadata which is not just there | 11:03 |
| chandankumar | noonedeadpunk, can you join #rdo channel | 11:05 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-tests master: Add Ubuntu Focal non voting check jobs https://review.opendev.org/729854 | 11:12 |
| noonedeadpunk | jrosser: do you know where VERSION_ID comes from? | 11:14 |
| jrosser | where is that? | 11:14 |
| noonedeadpunk | https://review.opendev.org/#/c/728259/14/test-log-collect.sh | 11:14 |
| noonedeadpunk | and actually throughout our scripts | 11:14 |
| noonedeadpunk | But the thing is that all of them run in zuul run | 11:14 |
| jrosser | oh yes | 11:14 |
| jrosser | it's in the scripts library i think | 11:15 |
| noonedeadpunk | WHile test-log-collect is in post | 11:15 |
| noonedeadpunk | which means it's out of the context | 11:15 |
| noonedeadpunk | (probably) | 11:15 |
| noonedeadpunk | Just really dont see it's definition anywhere... | 11:16 |
| noonedeadpunk | Like start thinging it may come from zuul or whatever | 11:16 |
| jrosser | oh hmm the tests repo | 11:17 |
| jrosser | uh-oh | 11:18 |
| noonedeadpunk | oh | 11:18 |
| noonedeadpunk | it's just . /etc/os-release lol | 11:19 |
| jrosser | VERSION_ID comes from tox i think | 11:20 |
| noonedeadpunk | no, just sourcing /etc/os-release | 11:20 |
| jrosser | do you have a link to that? | 11:21 |
| jrosser | oh found it | 11:21 |
| noonedeadpunk | sure | 11:21 |
| noonedeadpunk | https://opendev.org/openstack/openstack-ansible-tests/src/branch/master/test-ansible-functional.sh#L91 | 11:21 |
| noonedeadpunk | like too easy lol | 11:21 |
| jrosser | https://zuul.opendev.org/t/openstack/build/d8af419677454d7b949e95bcb4916b18/log/job-output.txt#1429-1438 | 11:22 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-tests master: Deploy modern ara for py3 deployments https://review.opendev.org/728259 | 11:22 |
| jrosser | so we need that in the log collect script | 11:22 |
| noonedeadpunk | yeah, already patched | 11:22 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-lxc_hosts master: Add legacy config key mappings for lxc version 4 https://review.opendev.org/729919 | 11:24 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-tests master: Add Ubuntu Focal non voting check jobs https://review.opendev.org/729854 | 11:25 |
| jrosser | it will be good to drop centos -> drop lxc2 -> big simplification of lxc roles | 11:25 |
| *** watersj has joined #openstack-ansible | 11:26 | |
| watersj | what do you suggest when deploying, checkout specific branch stable/[version] or run from head? | 11:27 |
| jrosser | watersj: on the stable branches we make tagged released about every two weeks, those are your most stable points and provide well defined places to do minor version upgrades between | 11:28 |
| jrosser | stable/<branch> is the place where patches land before the next two-weekly tag on that branch so you might use that to pull in a backported bugfix faster than we have released a tag | 11:29 |
| jrosser | master is where the development for the next (Victoria) release cycle happens, so you would use that to contribute code or run a bleeding-edge lab, but not for production | 11:29 |
| watersj | jrosser, ty just wanted make sure | 11:31 |
| jrosser | no worries, as usual the answer is "it depends" :) | 11:31 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-openstack_hosts master: CentOS 8 support https://review.opendev.org/709455 | 11:34 |
| *** mgariepy has joined #openstack-ansible | 11:55 | |
| Brace | mgariepy: 210 vms, 850 networks, 440 routers (sorry for the lag in responding) | 12:00 |
| mgariepy | Brace, no worries. | 12:04 |
| mgariepy | are all of them backed by ceph volumes ? | 12:04 |
| mgariepy | or you have ephemeral stuff as well? | 12:05 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-tests master: Add Ubuntu Focal non voting check jobs https://review.opendev.org/729854 | 12:07 |
| *** dpaclt has joined #openstack-ansible | 12:08 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-lxc_hosts master: Add legacy config key mappings for lxc version 4 https://review.opendev.org/729919 | 12:12 |
| *** schwicht has joined #openstack-ansible | 12:12 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-lxc_hosts master: Add vars files and container prep script for ubuntu focal https://review.opendev.org/729937 | 12:15 |
| *** rh-jelabarre has joined #openstack-ansible | 12:16 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-tests master: Add Ubuntu Focal non voting check jobs https://review.opendev.org/729854 | 12:16 |
| *** cshen has joined #openstack-ansible | 12:33 | |
| noonedeadpunk | jrosser: https://review.opendev.org/#/c/728259/ ara works | 12:37 |
| noonedeadpunk | I think we can maerge it now | 12:38 |
| *** cshen has quit IRC | 12:38 | |
| noonedeadpunk | then we will unblock functional tests | 12:42 |
| Brace | it's all ceph | 12:47 |
| Brace | mgariepy: I don't think we have any ephemeral stuff | 12:54 |
| mgariepy | ok | 12:55 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-galera_server master: Combine galera client role into galera_server https://review.opendev.org/725903 | 12:55 |
| jrosser | noonedeadpunk: ara looks good, approved | 12:55 |
| openstackgerrit | amolkahat proposed openstack/openstack-ansible-os_tempest master: Added tempest ironic resources setup. https://review.opendev.org/720705 | 13:00 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-tests master: Add Ubuntu Focal non voting check jobs https://review.opendev.org/729854 | 13:01 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-tests master: Add Ubuntu Focal non voting check jobs https://review.opendev.org/729854 | 13:02 |
| openstackgerrit | Merged openstack/openstack-ansible-os_sahara master: Update docstheme for style https://review.opendev.org/729706 | 13:04 |
| *** rh-jelabarre has quit IRC | 13:11 | |
| jrosser | noonedeadpunk: there is another os_neutron patch from mnaser here https://review.opendev.org/#/c/708665/ - is that one important? | 13:12 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-os_neutron master: config: add region_name to nova https://review.opendev.org/708665 | 13:13 |
| noonedeadpunk | dunno but jsut fixed this nit | 13:14 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_neutron stable/stein: Add extra rootwrap filter https://review.opendev.org/704778 | 13:14 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_nova stable/stein: pre-install: use path and drop dest https://review.opendev.org/709074 | 13:15 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_mistral stable/stein: Add meta to mistral https://review.opendev.org/696981 | 13:16 |
| dpaclt | Hi all ,I am getting (32, 'EPIPE') error while uploading glance image,back end is ceph..any suggestions | 13:22 |
| dmsimard | noonedeadpunk: it works! no wait, it failed :p https://a5625bfb1c655cabd60a-b71c90f67122d6f7b38b8442bc3a114a.ssl.cf5.rackcdn.com/728259/15/check/openstack-ansible-functional-centos-8/b78f492/logs/ara-report/result/50.html | 13:22 |
| noonedeadpunk | but it's expected failure :p | 13:23 |
| *** cshen has joined #openstack-ansible | 13:24 | |
| dmsimard | I left a comment in the patch | 13:26 |
| dmsimard | I left another, actually | 13:27 |
| dmsimard | noonedeadpunk: no need to block the patch, can be improved in another patchset :p | 13:29 |
| noonedeadpunk | ok, won't | 13:29 |
| noonedeadpunk | but really need a follow-up one | 13:29 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-tests master: Add Ubuntu Focal non voting check jobs https://review.opendev.org/729854 | 13:32 |
| *** rh-jelabarre has joined #openstack-ansible | 13:35 | |
| *** spatel has joined #openstack-ansible | 13:37 | |
| *** spatel has quit IRC | 13:42 | |
| jrosser | noonedeadpunk: this is sad https://github.com/python-poetry/poetry/issues/1363 | 13:55 |
| jrosser | i'm currently not able to run a pure python 3 bionic role test becasue of that | 13:55 |
| jrosser | or else i'm missing something important | 13:56 |
| * noonedeadpunk hears about that libbrary first time | 13:56 | |
| noonedeadpunk | but bug looks pretty related yes | 13:56 |
| noonedeadpunk | https://github.com/python-poetry/poetry/issues/1363#issuecomment-555913200 that is sad | 13:57 |
| jrosser | like this http://paste.openstack.org/show/793842/ | 13:57 |
| jrosser | not sure how to raise a bug for that to be released | 14:05 |
| jrosser | because it's all focussed on the apt package which contains that python module | 14:05 |
| jrosser | rather than the python itself | 14:06 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-lxc_hosts master: Add vars files and container prep script for ubuntu focal https://review.opendev.org/729937 | 14:11 |
| noonedeadpunk | jrosser: so and isntallation of python3-apt do not help? Or it's already installed? | 14:12 |
| jrosser | trouble is this is running in a tox venv | 14:12 |
| jrosser | so the python3-apt package is installed, but outside the venv | 14:12 |
| noonedeadpunk | can think only about symlinking.... | 14:13 |
| jrosser | hmm yes i was just thinking the same | 14:13 |
| noonedeadpunk | THis is suuuuuper old https://answers.launchpad.net/ubuntu/+source/python-apt/+question/285522 | 14:13 |
| jrosser | i'm trying to debug something else so this is just a distraction for now - but i think is shows we still use python2 a lot | 14:13 |
| noonedeadpunk | with the same tox problem lol | 14:13 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-tests master: Simlify ara setup https://review.opendev.org/729975 | 14:15 |
| noonedeadpunk | wondering if selinux installation will work | 14:16 |
| jrosser | typo there "Simplify" | 14:18 |
| noonedeadpunk | yeah. I also thinking about setting some fixed versions of libselinux | 14:18 |
| noonedeadpunk | in case it will work at all | 14:18 |
| noonedeadpunk | Hope you got what I tried to say with so awfull grammar :p | 14:19 |
| *** udesale_ has joined #openstack-ansible | 14:33 | |
| *** udesale has quit IRC | 14:36 | |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-tests master: Deploy modern ara for py3 deployments https://review.opendev.org/728259 | 14:55 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-tests master: Deploy modern ara for py3 deployments https://review.opendev.org/728259 | 14:57 |
| *** gyee has joined #openstack-ansible | 15:04 | |
| *** dirk has quit IRC | 15:38 | |
| *** mgariepy has quit IRC | 15:47 | |
| *** admin0 has joined #openstack-ansible | 16:01 | |
| *** NewJorg has quit IRC | 16:09 | |
| *** NewJorg has joined #openstack-ansible | 16:10 | |
| *** dpaclt82 has joined #openstack-ansible | 16:10 | |
| *** dpaclt has quit IRC | 16:11 | |
| *** mgariepy has joined #openstack-ansible | 16:23 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-galera_server master: Bump MariaDB version to 10.4 https://review.opendev.org/717830 | 16:37 |
| *** cshen has quit IRC | 16:45 | |
| *** udesale_ has quit IRC | 16:47 | |
| gshippey | @noonedeadpunk are you happy for me to +1 workflow on this https://review.opendev.org/#/c/728434/ ?? | 17:00 |
| noonedeadpunk | gshippey: yeah I think it's good topic and we defenitely need this patch to merge to resolve dependency | 17:02 |
| noonedeadpunk | gshippey: but if you have some thoughs on it it would be really great to have such input | 17:02 |
| *** mgariepy has quit IRC | 17:05 | |
| *** mgariepy has joined #openstack-ansible | 17:05 | |
| gshippey | I think that ones good to go, going to do a double check of the combination patch! | 17:07 |
| *** cshen has joined #openstack-ansible | 17:14 | |
| *** cshen has quit IRC | 17:19 | |
| *** jawad_axd has quit IRC | 17:30 | |
| openstackgerrit | Merged openstack/openstack-ansible-lxc_hosts master: ensure 20listchanges is not in debian container https://review.opendev.org/718882 | 17:30 |
| *** jawad_axd has joined #openstack-ansible | 17:31 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_neutron master: Updated from OpenStack Ansible Tests https://review.opendev.org/727322 | 18:14 |
| watersj | getting this during my deployment when setup-openstack, TASK [python_venv_build : Install python packages into the venv] "stdout: Ignoring argparse: markers 'python_version == \"2.7\"' don't match your environment, full stacktrace is here https://pastebin.com/JXGpvAVD | 18:26 |
| watersj | ubuntu 18.04, stable/train | 18:26 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-tests master: Deploy modern ara for py3 deployments https://review.opendev.org/728259 | 18:35 |
| openstackgerrit | Merged openstack/openstack-ansible master: Use galera server role to install galera client https://review.opendev.org/728434 | 18:38 |
| jrosser | watersj: I don’t think that is the actual error, probably need to look in the more specific log file generated by the venv build | 18:51 |
| *** jawad_axd has quit IRC | 18:52 | |
| watersj | reference to missing pkg-config in keystone container. but i not overriding or specifying python | 18:52 |
| openstackgerrit | Merged openstack/ansible-role-python_venv_build stable/stein: Verify version number for virtualenv>=20.0.0 https://review.opendev.org/729844 | 18:55 |
| openstackgerrit | Merged openstack/openstack-ansible-os_neutron master: config: add region_name to nova https://review.opendev.org/708665 | 18:59 |
| *** mathlin_ has quit IRC | 19:00 | |
| *** jamesdenton has quit IRC | 19:08 | |
| watersj | Cannot find libsystemd or libsystemd-journal. then it fails and retries 5 times | 19:13 |
| watersj | within keystone container | 19:13 |
| *** jamesdenton has joined #openstack-ansible | 19:13 | |
| *** AJaeger has left #openstack-ansible | 19:13 | |
| jrosser | watersj: which operating system? | 19:14 |
| watersj | ubuntu 18.04 | 19:15 |
| watersj | should it not be python3 | 19:15 |
| watersj | ? | 19:15 |
| jrosser | i believe that those messages are just the pip resolver discounting candidate installs becasue they are for the wrong python version | 19:17 |
| jrosser | your issue is that systemd_python cannot be built, and the error suggests that the C library headers for libsystemd and libsystemd-journal cannot be found | 19:18 |
| jrosser | libsystemd-dev should be installed due to this https://github.com/openstack/openstack-ansible-os_keystone/blob/master/vars/debian.yml#L32 | 19:18 |
| watersj | appears to be ii libsystemd0:amd64 237-3ubuntu10.40 | 19:19 |
| jrosser | that is the library itself | 19:19 |
| watersj | maybe I need to go to specific version of stable/train? | 19:20 |
| jrosser | the development headers are always in -dev packages which must also be present to compile the python bindings for those libraries | 19:20 |
| jrosser | hmm | 19:21 |
| jrosser | do you have repo_server containers? | 19:23 |
| watersj | i do | 19:24 |
| *** mathlin has joined #openstack-ansible | 19:26 | |
| watersj | looking at deploy server os_keystone/vars/debian.yml: - libsystemd-dev | 19:26 |
| watersj | I reverted my test environment to a clean state and am trying to redeploy | 19:27 |
| watersj | previously keystone was working, so only thing change is possible later checkout of stable/train | 19:28 |
| jrosser | it should not be trying to build wheels in the keystone container | 19:29 |
| jrosser | that should be delegated to the repo container and happen earlier in the keystone deployment | 19:29 |
| jrosser | you can see in your log it downloads SQAlechemy and systemd-python as .tgz, rather than get the wheels from the repo server | 19:30 |
| jrosser | i do not think there is good value in switching about branches or tags of stable/train to "fix" this | 19:30 |
| *** mathlin_ has joined #openstack-ansible | 19:31 | |
| jrosser | the variable i pointed you to with libsystemd-dev in it is a list of packages which must be installed into the repo server for the wheel build | 19:31 |
| jrosser | not into the keystone container | 19:32 |
| jrosser | i would really recommend building an all-in-one in a VM which you can use as a reference, because that is what is used to validate the tagged releases that are made | 19:32 |
| *** mathlin has quit IRC | 19:33 | |
| watersj | ok | 19:33 |
| jrosser | you can also pick any of our CI jobs to look at | 19:33 |
| jrosser | see the link in the IRC topic https://bit.ly/2SAcGAn | 19:33 |
| jrosser | here are all the patches to the train branch https://review.opendev.org/#/q/project:openstack/openstack-ansible+branch:stable/train | 19:34 |
| *** cshen has joined #openstack-ansible | 19:34 | |
| jrosser | you will be able to find the most recent release we made, which would be this https://review.opendev.org/#/c/728696/ | 19:35 |
| jrosser | if you click one of the jobs (lxc, bionic would be representative) https://zuul.opendev.org/t/openstack/build/ec172dc2c4f9402aa88917a7ccb6f347 | 19:35 |
| jrosser | under the logs tab you can explore all the log files from the deployment | 19:36 |
| watersj | let me scale it back to AIO, then do multi-node | 19:37 |
| watersj | jrosser, ty | 19:37 |
| jrosser | for example, we can look at the keystone python venv build log https://zuul.opendev.org/t/openstack/build/ec172dc2c4f9402aa88917a7ccb6f347/log/logs/openstack/aio1_keystone_container-3c7b8ebe/python_venv_build.log.txt#6629-6658 | 19:37 |
| jrosser | and interestingly you can see there that the openstack CI python wheel server is providing the built wheel there rather than it being made on the OSA repo server | 19:38 |
| jrosser | so there could well be a bug there that is only manifesting itself outside the CI environment | 19:38 |
| jrosser | i have to go now but hopefully that is some going pointers for digging into what is broken | 19:39 |
| *** jamesdenton has quit IRC | 19:40 | |
| *** jamesdenton has joined #openstack-ansible | 19:43 | |
| *** yolanda has quit IRC | 19:43 | |
| *** arkan has joined #openstack-ansible | 20:04 | |
| arkan | Hi guys my name is arkan, I'm new to openstack | 20:05 |
| arkan | I want to install it to my 2 rack servers | 20:06 |
| arkan | I'm using openstack-ansible | 20:06 |
| arkan | I have some errors in this stage openstack-ansible setup-hosts.yml | 20:06 |
| arkan | can anyone help me please? | 20:07 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible master: Return lxc jobs https://review.opendev.org/729878 | 20:07 |
| noonedeadpunk | pretty early for errors. what are they? | 20:07 |
| arkan | I think I've done a wrong network configuration | 20:08 |
| arkan | I have 2 rack servers, each one has 4 NICs (1GB), I want to use one server for Controller and the other for Compute | 20:10 |
| arkan | I'm using one NIC from each server, and each NIC has sub interfaces for different vlans | 20:11 |
| arkan | those NICs (2 in total) are connected to my router | 20:11 |
| noonedeadpunk | jrosser: can you kindly check https://review.opendev.org/#/c/728259/18 one more time? | 20:12 |
| noonedeadpunk | I've implemented dmsimard's advices here all in all | 20:13 |
| jrosser | noonedeadpunk: sure | 20:13 |
| noonedeadpunk | not sure hwta this ara-mange error for centos 8 is :( | 20:13 |
| jrosser | noonedeadpunk: could you read back with watersj issue just now - I’m slightly worried about the repo server cleanup patches | 20:14 |
| noonedeadpunk | arkan: ok, sounds ok | 20:14 |
| arkan | yes | 20:14 |
| dmsimard | noonedeadpunk: where is the error ? | 20:15 |
| arkan | where you want me to paste the errors? | 20:16 |
| arkan | I'm not aware of the https://review.opendev.org, this is the first time when I'm seeing it | 20:16 |
| noonedeadpunk | dmsimard: https://zuul.opendev.org/t/openstack/build/6ef4ac7de9514c2bb61415b560be924d/log/job-output.txt#2458 | 20:17 |
| noonedeadpunk | arkan: you can paste here http://paste.openstack.org/ or wherever conmfortable | 20:17 |
| arkan | great | 20:18 |
| arkan | one moment, because the process is still running and I've used openstack-ansible setup-hosts.yml -vvvv | 20:18 |
| noonedeadpunk | jrosser: hm, you mean this one? https://opendev.org/openstack/openstack-ansible-repo_server/commit/53a8add55b480166ce5ded7df4e196f70d613c1e | 20:19 |
| arkan | @noonedeadpunk I will past my netplan there first and the config until the process finishes | 20:20 |
| arkan | @noondeadpunk http://paste.openstack.org/show/793859/ | 20:23 |
| noonedeadpunk | oh, I faound where this error comes from I guess | 20:24 |
| arkan | @nooneeadpunk http://paste.openstack.org/show/793860/ | 20:25 |
| arkan | @noonedeadpunk when the logs was running in my screen, I saw it was connectivity, it could not reach the internet | 20:26 |
| noonedeadpunk | I think you shouldn't have gateway for br-mgmt | 20:26 |
| noonedeadpunk | or not having for eno1 | 20:27 |
| arkan | @noonedeadpunk http://paste.openstack.org/show/793861/ | 20:27 |
| arkan | my eno1 is the only interface that is connected to the internet | 20:28 |
| arkan | through the server ip: 192.168.50.10 | 20:28 |
| noonedeadpunk | but I see gateway configured for br-mgmt as well | 20:29 |
| noonedeadpunk | so you might have 2 or wrong default route | 20:29 |
| arkan | yes, that is true | 20:30 |
| arkan | I will show you my router also | 20:30 |
| arkan | one moment | 20:30 |
| arkan | https://drive.google.com/drive/folders/1t71q1YlAz5JVohVFuWIzLNQ6j4GgqoL2?usp=sharing | 20:32 |
| arkan | I will put my config files there | 20:34 |
| noonedeadpunk | dmsimard: so we probably just not divide subunit report now? https://opendev.org/openstack/openstack-ansible-tests/src/branch/master/test-log-collect.sh#L262 | 20:39 |
| arkan | @noonedeadpunk I've created openstack folder with the 3 files 1.user_variables 2.openstack_user_config.yml 3.01-netcfg.yaml | 20:39 |
| noonedeadpunk | ok, so the question is - can you ping world, like pypi.org from hosts? | 20:40 |
| arkan | yes | 20:40 |
| noonedeadpunk | ok, cool. Then the most insteresting part is error:) | 20:41 |
| arkan | yes, I'm waiting for the process to finish, it took a lot of time | 20:41 |
| arkan | and It's still running | 20:41 |
| noonedeadpunk | ok, cool | 20:41 |
| *** also_stingrayza has joined #openstack-ansible | 20:41 | |
| noonedeadpunk | it just pretty late here.... | 20:41 |
| arkan | here in Romania it's 23:42 | 20:42 |
| arkan | and I'm eagerly want to see Horizon | 20:42 |
| arkan | I spent a few days trying to make this work, and I'm still continuing this road :)) | 20:43 |
| noonedeadpunk | So we're in the same timezone lol | 20:43 |
| arkan | wow nice | 20:43 |
| *** stingrayza has quit IRC | 20:45 | |
| dmsimard | noonedeadpunk: subunit does not exist in 1.x (yet) | 20:46 |
| dmsimard | wouldn't be terribly complicated to re-implement using the API, it's this part: https://github.com/ansible-community/ara/blob/acf7769f00de59c2d388409af62ccaa030b7bfc5/ara/cli/generate.py#L176-L291 | 20:48 |
| dmsimard | I'd like to keep subunit out of reqs.txt though | 20:49 |
| dmsimard | if it's something you need I can create an issue about it | 20:49 |
| noonedeadpunk | hard to say how much we need it tbh, we just always had it, so seems that we need to update our script to omit it for new ara | 20:51 |
| dmsimard | it may have been used for openstack-health at some point ? | 20:51 |
| dmsimard | I don't recall the details | 20:51 |
| dmsimard | subunit is arcane magic to me | 20:52 |
| noonedeadpunk | same for me :p | 20:52 |
| dmsimard | I remember needing to go through subunit's unit/functional/integration test to understand how the damn thing worked | 20:52 |
| noonedeadpunk | yeah, it's totally for openstack-health | 20:54 |
| noonedeadpunk | But I think we used a grafana for a while... | 20:54 |
| *** cshen has quit IRC | 20:54 | |
| noonedeadpunk | but anyway, yeah | 20:55 |
| jrosser | noonedeadpunk: do we merge it? :) | 20:55 |
| noonedeadpunk | great question lol | 20:55 |
| * jrosser only got half an eye on this subunit business | 20:55 | |
| noonedeadpunk | so it really neded for filling in data here http://status.openstack.org/openstack-health/#/ | 20:56 |
| *** arkan has quit IRC | 20:56 | |
| * noonedeadpunk start thinking if we should use new ara for centos 8 only | 20:57 | |
| *** arkan has joined #openstack-ansible | 21:04 | |
| arkan | @noonedeadpunk sorry I was disconnected | 21:04 |
| arkan | @noonedeadpunk I've put the errors in the files | 21:05 |
| arkan | https://drive.google.com/drive/folders/1t71q1YlAz5JVohVFuWIzLNQ6j4GgqoL2?usp=sharing | 21:05 |
| arkan | I've tried to paste it on paste.openstack.org but it does not work now | 21:06 |
| arkan | so I put the errors in the file in my drive | 21:06 |
| arkan | noonedeadpunk: it seems that it can not connect from inside the container | 21:11 |
| arkan | : Failed to fetch http://security.ubuntu.com/ubuntu/pool/main/p/python-apt/python-apt-common_1.6.5ubuntu0.2_all.deb Temporary failure resolving \'us.archive.ubuntu.com\' | 21:11 |
| noonedeadpunk | arkan: ok, so network from host works. | 21:12 |
| noonedeadpunk | but seems the issue is in lxc containers | 21:12 |
| arkan | noonedeadpunk: I think the network config. from inside the container has the problem | 21:12 |
| arkan | but from the host it can connect to the internet | 21:12 |
| noonedeadpunk | ok, so containers should connect to the internet through lxcbr0 by doing src-nat | 21:13 |
| noonedeadpunk | and they tend to set nameserver to the ip which is lxcbr | 21:14 |
| noonedeadpunk | this bridge is configured automatically | 21:14 |
| noonedeadpunk | But I still have doubts regarding having 2 default routes. | 21:14 |
| noonedeadpunk | what is the output of ip r on controller? | 21:15 |
| noonedeadpunk | `ip r` | 21:15 |
| *** gshippey has quit IRC | 21:16 | |
| arkan | noonedeadpunk: so how can I modify my netplan config to make it work? | 21:17 |
| arkan | should I remove the gateway from br-mgmt | 21:17 |
| arkan | should I remove the gateway from br-mgmt? | 21:17 |
| arkan | eno1 has the server ip (vlan 50 in router untagged) | 21:18 |
| noonedeadpunk | I'd say yes | 21:19 |
| noonedeadpunk | but not sure that issue is really exactly in it | 21:19 |
| arkan | br-mgmt: 172.29.236.10/22 (uses vlan10 tagged through the router) | 21:19 |
| noonedeadpunk | generally you shouldn't have 2 default rautes in any time | 21:19 |
| noonedeadpunk | *routes | 21:20 |
| noonedeadpunk | you can check current routes with `ip r` command | 21:20 |
| noonedeadpunk | so i'd use gateway for eno1 and not frm br-mgmt | 21:20 |
| arkan | root@controller1:~# ip route | 21:20 |
| arkan | default via 192.168.50.1 dev eno1 proto static | 21:20 |
| arkan | default via 172.29.236.1 dev br-mgmt proto static | 21:20 |
| arkan | 10.0.3.0/24 dev lxcbr0 proto kernel scope link src 10.0.3.1 linkdown | 21:20 |
| arkan | 172.29.236.0/22 dev br-mgmt proto kernel scope link src 172.29.236.10 | 21:20 |
| arkan | 172.29.240.0/22 dev br-vxlan proto kernel scope link src 172.29.240.10 | 21:20 |
| arkan | 192.168.50.0/24 dev eno1 proto kernel scope link src 192.168.50.10 | 21:20 |
| arkan | 192.168.50.0/24 dev br-flat proto kernel scope link src 192.168.50.100 | 21:20 |
| noonedeadpunk | huh, so traffic should go through eno1 which is correct | 21:22 |
| noonedeadpunk | can you ping router or world from inside container? | 21:22 |
| noonedeadpunk | you can access container with `lxc-attach -n <container_name>` | 21:22 |
| arkan | root@controller1:~# lxc list | 21:23 |
| arkan | +------+-------+------+------+------+-----------+ | 21:23 |
| arkan | | NAME | STATE | IPV4 | IPV6 | TYPE | SNAPSHOTS | | 21:23 |
| arkan | +------+-------+------+------+------+-----------+ | 21:23 |
| arkan | there is no container running | 21:23 |
| noonedeadpunk | lxc-ls -a | 21:23 |
| noonedeadpunk | lxc-ls --active | 21:24 |
| arkan | the command is not working | 21:24 |
| arkan | yes | 21:24 |
| arkan | now it listed the containers | 21:24 |
| arkan | no | 21:25 |
| arkan | root@controller1:~# lxc-attach -n controller1_rabbit_mq_container-0349abaa | 21:25 |
| arkan | root@controller1-rabbit-mq-container-0349abaa:~# | 21:25 |
| arkan | root@controller1-rabbit-mq-container-0349abaa:~# | 21:25 |
| arkan | root@controller1-rabbit-mq-container-0349abaa:~# | 21:25 |
| arkan | root@controller1-rabbit-mq-container-0349abaa:~# ping google.com | 21:25 |
| arkan | ping: google.com: Temporary failure in name resolution | 21:25 |
| arkan | root@controller1-rabbit-mq-container-0349abaa:~# | 21:25 |
| arkan | 21:25 | |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-tests master: Deploy modern ara for CentOS 8 deployments https://review.opendev.org/728259 | 21:25 |
| arkan | root@controller1-rabbit-mq-container-0349abaa:~# ip r | 21:25 |
| arkan | 172.29.236.0/22 dev eth1 proto kernel scope link src 172.29.239.228 | 21:25 |
| noonedeadpunk | so it's missing default route? | 21:26 |
| noonedeadpunk | that's weird | 21:26 |
| noonedeadpunk | what about `ip a`? | 21:26 |
| arkan | root@controller1-rabbit-mq-container-0349abaa:~# ip link | 21:26 |
| arkan | 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN mode DEFAULT group default qlen 1000 | 21:26 |
| arkan | link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 | 21:26 |
| arkan | 149: eth0@if150: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP mode DEFAULT group default qlen 1000 | 21:26 |
| arkan | link/ether 00:16:3e:f1:0a:7b brd ff:ff:ff:ff:ff:ff link-netnsid 0 | 21:26 |
| arkan | 151: eth1@if152: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP mode DEFAULT group default qlen 1000 | 21:26 |
| arkan | link/ether 00:16:3e:ab:22:48 brd ff:ff:ff:ff:ff:ff link-netnsid 0 | 21:26 |
| arkan | root@controller1-rabbit-mq-container-0349abaa:~# ip a | 21:27 |
| arkan | 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000 | 21:27 |
| arkan | link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 | 21:27 |
| arkan | inet 127.0.0.1/8 scope host lo | 21:27 |
| arkan | valid_lft forever preferred_lft forever | 21:27 |
| arkan | inet6 ::1/128 scope host | 21:27 |
| arkan | valid_lft forever preferred_lft forever | 21:27 |
| arkan | 149: eth0@if150: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000 | 21:27 |
| arkan | link/ether 00:16:3e:f1:0a:7b brd ff:ff:ff:ff:ff:ff link-netnsid 0 | 21:27 |
| arkan | inet6 fe80::216:3eff:fef1:a7b/64 scope link | 21:27 |
| arkan | valid_lft forever preferred_lft forever | 21:27 |
| arkan | 151: eth1@if152: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000 | 21:27 |
| arkan | link/ether 00:16:3e:ab:22:48 brd ff:ff:ff:ff:ff:ff link-netnsid 0 | 21:27 |
| arkan | inet 172.29.239.228/22 brd 172.29.239.255 scope global eth1 | 21:27 |
| arkan | valid_lft forever preferred_lft forever | 21:27 |
| arkan | inet6 fe80::216:3eff:feab:2248/64 scope link | 21:27 |
| arkan | valid_lft forever preferred_lft forever | 21:27 |
| arkan | root@controller1-rabbit-mq-container-0349abaa:~# | 21:27 |
| noonedeadpunk | oh my, please use some pastebin :) | 21:27 |
| noonedeadpunk | ok, so eth0 don't have an ip adress for some reason while should | 21:28 |
| *** cshen has joined #openstack-ansible | 21:28 | |
| arkan | https://pastebin.com/P9JEKzMu | 21:29 |
| arkan | yes, I don't know why | 21:29 |
| noonedeadpunk | yeah I got it - so problem is that eth0 don't have ip address | 21:29 |
| arkan | I was thinking that something wrong with my network config | 21:29 |
| noonedeadpunk | what bridges do you have on host? | 21:30 |
| noonedeadpunk | Yeah | 21:30 |
| noonedeadpunk | But I don't think it's smth with osa config | 21:30 |
| arkan | I think in my netplan | 21:30 |
| arkan | from what I've read in the doc | 21:30 |
| arkan | br-vlan does not need an ip address | 21:31 |
| arkan | but the problem that my router can not be configured without an ip address for that port | 21:31 |
| arkan | and my 2 servers are connected to the router | 21:31 |
| arkan | port 1 on router ---> for server 1 (controller) and port 2 on router ---> for (compute) | 21:32 |
| noonedeadpunk | so I'd try removing second defult route and re-create containers | 21:32 |
| arkan | what is the new netplan config that I should use? | 21:33 |
| arkan | you can download make change to the file from my drive | 21:33 |
| *** cshen has quit IRC | 21:33 | |
| noonedeadpunk | you can drop containers like this openstack-ansible /opt/openstack-ansible/playbooks/lxc-containers-destroy.yml --limit rabbitmq_all | 21:34 |
| arkan | ok | 21:34 |
| noonedeadpunk | abd create openstack-ansible /opt/openstack-ansible/playbooks/lxc-containers-create.yml --limit rabbitmq_all,lxc_hosts | 21:34 |
| noonedeadpunk | just drop gateway from br=mgmt | 21:34 |
| arkan | I will drop the gateway first from br-mgmt | 21:35 |
| arkan | I dropped the gateway from both controller and compute | 21:41 |
| arkan | and I ran the above code | 21:41 |
| arkan | I've received an error | 21:42 |
| arkan | https://pastebin.com/ybVmnkgL | 21:43 |
| arkan | and from inside the container | 21:49 |
| arkan | https://pastebin.com/r2EmiusH | 21:49 |
| arkan | noonedeadpunk: but how to reach the internet if it's tagged inside the route? | 21:52 |
| arkan | I can not see default gateway in the route | 21:53 |
| arkan | only untagged vlan (50) can get to the internet | 21:54 |
| jrosser | arkan: bear in mind it’s late :) but you are missing eth0 having an ip from inside your container | 21:55 |
| jrosser | eth0 will be assigned an address in the 10.x range by dnsmasq running on the host | 21:56 |
| arkan | yes, I will try to figure it out, till tomorrow, thanks a lot @noonedeadpunk @jrosser for your help | 21:57 |
| arkan | good night | 21:57 |
| jrosser | eth0 should be the default route in the container, and the host NATs eth0 on behalf of each container and the traffic goes out of the hosts default route | 21:57 |
| arkan | ok, I will check about it | 21:59 |
| arkan | thanks a lot | 21:59 |
| *** arkan has quit IRC | 22:02 | |
| *** rh-jelabarre has quit IRC | 22:09 | |
| *** itandops has joined #openstack-ansible | 22:24 | |
| *** itandops has quit IRC | 22:52 | |
| *** itandops has joined #openstack-ansible | 22:53 | |
| *** dpaclt82 has quit IRC | 23:14 | |
| *** NewJorg has quit IRC | 23:29 | |
| *** gyee has quit IRC | 23:46 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!