| *** weifan has joined #openstack-ansible | 00:13 | |
| *** weifan has quit IRC | 00:17 | |
| *** gyee has quit IRC | 00:19 | |
| *** yolanda has quit IRC | 00:20 | |
| *** weifan has joined #openstack-ansible | 00:25 | |
| *** weifan has quit IRC | 00:30 | |
| *** BjoernT has joined #openstack-ansible | 00:45 | |
| *** BjoernT has quit IRC | 00:47 | |
| *** BjoernT has joined #openstack-ansible | 00:48 | |
| *** BjoernT has quit IRC | 00:48 | |
| *** BjoernT has joined #openstack-ansible | 00:49 | |
| *** BjoernT has quit IRC | 00:49 | |
| *** BjoernT has joined #openstack-ansible | 00:49 | |
| *** BjoernT has joined #openstack-ansible | 00:50 | |
| *** BjoernT has quit IRC | 00:50 | |
| *** BjoernT has joined #openstack-ansible | 00:51 | |
| *** BjoernT has quit IRC | 00:51 | |
| *** BjoernT has joined #openstack-ansible | 00:54 | |
| *** BjoernT has joined #openstack-ansible | 00:54 | |
| *** BjoernT has quit IRC | 00:54 | |
| *** tinwood has quit IRC | 02:10 | |
| *** tinwood has joined #openstack-ansible | 02:12 | |
| *** mrda has quit IRC | 02:25 | |
| *** mrda has joined #openstack-ansible | 02:27 | |
| *** udesale has joined #openstack-ansible | 03:57 | |
| *** joshualyle has joined #openstack-ansible | 04:20 | |
| *** ChipOManiac has joined #openstack-ansible | 04:38 | |
| *** BjoernT has joined #openstack-ansible | 04:58 | |
| *** BjoernT has quit IRC | 05:01 | |
| *** jbadiapa has quit IRC | 05:02 | |
| *** BjoernT has joined #openstack-ansible | 05:02 | |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-os_ceilometer master: Fix gnocchi db population https://review.opendev.org/671988 | 05:10 |
|---|---|---|
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-os_ceilometer master: service_setup: refactor service setup to a single file https://review.opendev.org/670473 | 05:11 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible master: [WIP] Drop creation of log directories https://review.opendev.org/671783 | 05:12 |
| *** zbr has joined #openstack-ansible | 05:45 | |
| openstackgerrit | pengyuesheng proposed openstack/openstack-ansible-os_neutron master: Blacklist sphinx 2.1.0 (autodoc bug) https://review.opendev.org/672866 | 05:59 |
| openstackgerrit | pengyuesheng proposed openstack/openstack-ansible-os_neutron master: Bump the openstackdocstheme extension to 1.20 https://review.opendev.org/672867 | 06:07 |
| openstackgerrit | pengyuesheng proposed openstack/openstack-ansible-os_octavia master: Blacklist sphinx 2.1.0 (autodoc bug) https://review.opendev.org/672868 | 06:10 |
| openstackgerrit | pengyuesheng proposed openstack/openstack-ansible-os_octavia master: Bump the openstackdocstheme extension to 1.20 https://review.opendev.org/672869 | 06:14 |
| *** miloa has joined #openstack-ansible | 06:18 | |
| openstackgerrit | pengyuesheng proposed openstack/openstack-ansible-os_gnocchi master: Blacklist sphinx 2.1.0 (autodoc bug) https://review.opendev.org/672871 | 06:18 |
| *** jbadiapa has joined #openstack-ansible | 06:24 | |
| *** raukadah is now known as chandankumar | 06:51 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible master: Imported Translations from Zanata https://review.opendev.org/672894 | 07:18 |
| *** kopecmartin|off is now known as kopecmartin | 07:26 | |
| *** cshen has joined #openstack-ansible | 07:32 | |
| *** tosky has joined #openstack-ansible | 07:34 | |
| openstackgerrit | pengyuesheng proposed openstack/openstack-ansible-tests master: Blacklist sphinx 2.1.0 (autodoc bug) https://review.opendev.org/672901 | 07:34 |
| openstackgerrit | pengyuesheng proposed openstack/openstack-ansible-tests master: Blacklist sphinx 2.1.0 (autodoc bug) https://review.opendev.org/672903 | 07:38 |
| *** cshen has quit IRC | 07:43 | |
| *** rpittau|afk is now known as rpittau | 07:44 | |
| *** pcaruana has joined #openstack-ansible | 07:44 | |
| *** mkuf has joined #openstack-ansible | 07:46 | |
| openstackgerrit | pengyuesheng proposed openstack/openstack-ansible-tests master: Bump the openstackdocstheme extension to 1.20 https://review.opendev.org/672904 | 07:50 |
| openstackgerrit | pengyuesheng proposed openstack/openstack-ansible-os_neutron master: Bump the openstackdocstheme extension to 1.20 https://review.opendev.org/672867 | 07:52 |
| openstackgerrit | pengyuesheng proposed openstack/openstack-ansible-os_octavia master: Bump the openstackdocstheme extension to 1.20 https://review.opendev.org/672869 | 07:53 |
| *** cshen has joined #openstack-ansible | 07:56 | |
| noonedeadpunk | do we have broken nova role, aren't we? | 08:15 |
| noonedeadpunk | while config is in place https://opendev.org/openstack/openstack-ansible-os_nova/src/branch/master/templates/nova.conf.j2#L264 it's not rendered correctly http://logs.openstack.org/83/671783/8/check/openstack-ansible-deploy-aio_lxc-ubuntu-bionic/be4e974/logs/etc/openstack/aio1_nova_api_container-2e62b5d5/nova/nova.conf.txt.gz | 08:18 |
| *** dkopper has joined #openstack-ansible | 08:19 | |
| * noonedeadpunk wondering if ansible 2.8 might be related | 08:19 | |
| mkuf | Hey there, I'm trying to redeploy a lxc-container on one of my infra-nodes but it seems networking isn't set up correctly by lxc-containers-create. http://paste.openstack.org/show/754883/ Any idea what might cause this issue? I'm using 18.1.8 on Ubuntu 16.04 | 08:21 |
| *** dkopper has left #openstack-ansible | 08:23 | |
| jrosser | noonedeadpunk: also config_template work recently may have an effect there | 08:26 |
| noonedeadpunk | has it been merged? | 08:28 |
| *** siqbal has joined #openstack-ansible | 08:29 | |
| noonedeadpunk | yeah... | 08:29 |
| noonedeadpunk | cloudnull: can you check this out please? (regarding nova config) | 08:30 |
| *** ivve has joined #openstack-ansible | 08:37 | |
| *** hamzaachi has joined #openstack-ansible | 08:40 | |
| *** hamzaachi has quit IRC | 08:41 | |
| *** hamzaachi has joined #openstack-ansible | 08:42 | |
| *** hamzaachi has quit IRC | 08:43 | |
| *** hamzaachi has joined #openstack-ansible | 08:43 | |
| *** hamzaachi has quit IRC | 09:01 | |
| *** hamzaachi has joined #openstack-ansible | 09:01 | |
| *** joshualyle has quit IRC | 09:23 | |
| *** hamzaachi has quit IRC | 09:28 | |
| *** hamzaachi has joined #openstack-ansible | 09:28 | |
| *** hamzaachi has quit IRC | 09:30 | |
| *** hamzaachi has joined #openstack-ansible | 09:30 | |
| *** siqbal has quit IRC | 09:31 | |
| *** siqbal has joined #openstack-ansible | 09:31 | |
| *** arxcruz is now known as arxcruz|off | 09:32 | |
| *** jbadiapa has quit IRC | 09:39 | |
| jrosser | noonedeadpunk: i think the same is happening here for image_ref and image_ref_alt? http://logs.openstack.org/04/672904/1/check/openstack-ansible-functional-ubuntu-bionic/33ee10f/logs/etc/openstack/keystone1/tempest/tempest.conf.txt.gz | 09:57 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible master: Bump sha of heat https://review.opendev.org/672943 | 10:15 |
| *** mkuf has quit IRC | 10:25 | |
| noonedeadpunk | jrosser: yep, I think yes.. | 10:27 |
| noonedeadpunk | so I guess we may either try to debug this or just revert.... | 10:28 |
| noonedeadpunk | waiting for cloudnull is still an option:) | 10:28 |
| jrosser | yeah send up the bat-signal for cloudnull and see what happens :) | 10:31 |
| *** hamzaachi has quit IRC | 10:31 | |
| *** brtknr has quit IRC | 10:36 | |
| *** brtknr has joined #openstack-ansible | 10:37 | |
| *** ChipOManiac has quit IRC | 10:39 | |
| *** udesale has quit IRC | 10:44 | |
| noonedeadpunk | returning to nova - seems that it theoretically could be just empty https://opendev.org/openstack/openstack-ansible-os_nova/src/branch/master/defaults/main.yml#L277 | 10:44 |
| *** udesale has joined #openstack-ansible | 10:45 | |
| *** siqbal has quit IRC | 10:47 | |
| noonedeadpunk | so I guess that problem is that's it just missing "=", which might be dropped by config_template... | 10:48 |
| openstackgerrit | Namrata proposed openstack/openstack-ansible stable/stein: Update docs & upgrade scripts for Stein https://review.opendev.org/672719 | 10:49 |
| noonedeadpunk | but why in the world it passed gates... | 10:51 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_heat master: Fix keystone endpoint for heat servers https://review.opendev.org/672948 | 10:55 |
| jrosser | noonedeadpunk: well config_template has its own testsuite i think | 10:56 |
| jrosser | i wonder what the behaviour was before for an empty value, if it omitted the config option | 10:56 |
| noonedeadpunk | it's pretty easy to check. | 10:57 |
| admin0 | hi all .. is just adding a nova_override to have the virt_type=lxc enough to setup a new compute node as lxc node ? | 10:57 |
| noonedeadpunk | http://logs.openstack.org/83/671783/3/check/openstack-ansible-deploy-aio_metal-ubuntu-bionic/7902cfd/logs/etc/host/nova/nova.conf.txt.gz | 10:57 |
| noonedeadpunk | it was left as is | 10:58 |
| jrosser | http://logs.openstack.org/47/672347/1/gate/openstack-ansible-deploy-aio_metal-debian-stable/dc30489/logs/etc/host/nova/nova.conf.txt.gz | 10:58 |
| jrosser | yes so we have lost the = | 10:58 |
| noonedeadpunk | I kinda try to look through change one more time... | 10:59 |
| *** ksdean has quit IRC | 10:59 | |
| noonedeadpunk | but probably it's worth dropping param completely? | 10:59 |
| jrosser | there might be a case for needing foo = "" | 11:01 |
| noonedeadpunk | jrosser: I guess this might be the case https://opendev.org/openstack/ansible-config_template/src/branch/master/action/config_template.py#L226 | 11:01 |
| *** brtknr has quit IRC | 11:02 | |
| *** brtknr has joined #openstack-ansible | 11:02 | |
| *** brtknr has quit IRC | 11:02 | |
| noonedeadpunk | lets check this out:) | 11:02 |
| *** brtknr has joined #openstack-ansible | 11:03 | |
| *** brtknr has quit IRC | 11:06 | |
| ioni | hey guys | 11:09 |
| ioni | i don't quite understand something | 11:09 |
| *** brtknr has joined #openstack-ansible | 11:10 | |
| ioni | in the docs, is mentioned that compute nodes needs to have an ip on br-storage | 11:10 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/ansible-config_template master: Fix writing of keys without values https://review.opendev.org/672952 | 11:10 |
| noonedeadpunk | jrosser: let's see what happens... ^ | 11:10 |
| ioni | after deploying a ceph cluster, cinder and so on, computes nodes are using br-mgmt ips to connect to ceph | 11:10 |
| ioni | is there a problem in the docs? | 11:10 |
| ioni | https://docs.openstack.org/openstack-ansible/stein/user/ceph/full-deploy.html | 11:10 |
| jrosser | ioni: i would say that is a long standing error with the way the integration of OSA and ceph-ansible has been done | 11:11 |
| ioni | jrosser, so what should I do? | 11:11 |
| *** kplant has joined #openstack-ansible | 11:11 | |
| ioni | exchange the public_network ? | 11:11 |
| ioni | monitor_address_block: "{{ cidr_networks.container }}" | 11:12 |
| ioni | public_network: "{{ cidr_networks.container }}" | 11:12 |
| ioni | cluster_network: "{{ cidr_networks.storage }}" | 11:12 |
| jrosser | if you can have a go at getting the storage traffic actually onto the storage network that would be really cool | 11:12 |
| ioni | jrosser, do you have a bug that I can follow? | 11:15 |
| ioni | i have the impression is not as simple just to reconfigure public_network | 11:15 |
| ioni | since nova is using ceph mon to actually connect to ceph | 11:15 |
| ioni | ceph mon has only one interface into br-mgmt | 11:16 |
| jrosser | ioni: right, so thats probably the root cause really | 11:16 |
| jrosser | that the mon container need an interface on the storage network | 11:16 |
| * jrosser away for a bit | 11:16 | |
| ioni | alright | 11:16 |
| openstackgerrit | Namrata proposed openstack/openstack-ansible-plugins master: Removing unused filters from osa plugins https://review.opendev.org/672955 | 11:23 |
| *** jbadiapa has joined #openstack-ansible | 11:37 | |
| *** irclogbot_3 has quit IRC | 11:53 | |
| *** irclogbot_0 has joined #openstack-ansible | 11:54 | |
| openstackgerrit | Namrata proposed openstack/openstack-ansible-plugins master: Removing unused filters from osa plugins https://review.opendev.org/672955 | 11:55 |
| chandankumar | jrosser: Hello | 11:56 |
| jrosser | Hello | 11:57 |
| chandankumar | jrosser: Need some help here https://review.opendev.org/#/c/672231/ on this review | 11:57 |
| chandankumar | jrosser: I am not sure what is needed to fix that issue | 11:57 |
| jrosser | Ok | 11:57 |
| jrosser | So the reason that I -1 is that the commit message talks about many many thing including stackviz | 11:58 |
| jrosser | And then the set_fact is when: on stackviz too | 11:59 |
| openstackgerrit | Namrata proposed openstack/openstack-ansible-plugins master: Removing unused filters from osa plugins https://review.opendev.org/672955 | 11:59 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/ansible-config_template master: Fix writing of keys without values https://review.opendev.org/672952 | 11:59 |
| jrosser | But unless I am misunderstanding, this is a very general problem where the handler from the venv build take will be run at the end of the play that included os_tempest | 12:00 |
| jrosser | And the handler brings with it a conditional from os_tempest using a var that is not in scope of the play | 12:00 |
| jrosser | So this isn’t actually to do with stackviz? But is a potential issue for any play in a project that tries to use os_tempest as an include_role? | 12:01 |
| jrosser | noonedeadpunk: interested in your opinion on this too ^^ | 12:02 |
| jrosser | So I was thinking that instead, os_tempest should set_fact of the install method var always and in a very prominent place, and include a very clear explanation of why in the code | 12:03 |
| noonedeadpunk | so, I'd probably do set_fact: tempest_install_method: "source" when tempest_install_method is not defined. And did it right before Install Tempest packages from PIP | 12:06 |
| noonedeadpunk | or just placed tempest_install_method into defaults for example... | 12:06 |
| noonedeadpunk | which is probably the best place for it, since it's needed anyway | 12:07 |
| *** hwoarang has quit IRC | 12:08 | |
| *** jbadiapa has quit IRC | 12:08 | |
| *** jbadiapa has joined #openstack-ansible | 12:08 | |
| jrosser | The thing with having it in the defaults is that the role defaults are not in scope of a handler executed at the end of a play that includes the role | 12:09 |
| jrosser | In OSA we have install method as some top level group var or override so it exists everywhere | 12:10 |
| noonedeadpunk | so probably the same approach might be done with stackviz? | 12:11 |
| *** ksdean has joined #openstack-ansible | 12:12 | |
| noonedeadpunk | but otherwise I agree with your suggestion jrosser | 12:12 |
| *** hwoarang has joined #openstack-ansible | 12:13 | |
| jrosser | The root cause is we use that noop handler in the venv build role as a kind of callback | 12:13 |
| jrosser | And in os_tempest we make the venv build conditional | 12:14 |
| jrosser | And so the conditional test gets leaked out into the play via the handler | 12:14 |
| jrosser | chandankumar: ^ does this all make sense? | 12:15 |
| chandankumar | jrosser: yes, make sense | 12:16 |
| jrosser | So I think we make a task that unconditionally does set_fact, right before the pip packages install | 12:17 |
| jrosser | And we can redo the commit message to summarise what we have said just now, I can help with that if you need? | 12:18 |
| noonedeadpunk | probably only when tempest_install_method is not defined? | 12:18 |
| jrosser | Well - I dunno | 12:18 |
| noonedeadpunk | since unconditionally means that we may accidentally redefine it | 12:18 |
| jrosser | oh well isnt it more subtle - set_fact to its own value | 12:20 |
| jrosser | defaulted to 'source' | 12:20 |
| jrosser | you see it is already defined in the role defaults https://github.com/openstack/openstack-ansible-os_tempest/blob/master/defaults/main.yml#L25 | 12:21 |
| jrosser | so any task in the os_tempest role is guaranteed to have a value | 12:21 |
| noonedeadpunk | ah, I see | 12:22 |
| jrosser | we need to 'promote' the value out of the role defaults into a hostvar with set_fact, so the contition in the handler can find it at the end of the play, outside of os_tempest | 12:22 |
| jrosser | so actually there isnt a default needed, just set_fact to itself | 12:23 |
| noonedeadpunk | jrosser: are you sure that we need empty "option ="?:) http://logs.openstack.org/52/672952/2/check/openstack-ansible-functional-debian-stable/e737714/job-output.txt.gz#_2019-07-26_12_15_48_860902 | 12:29 |
| jrosser | noonedeadpunk: i well i don't really know, other than thats what it did before? | 12:30 |
| openstackgerrit | Chandan Kumar (raukadah) proposed openstack/openstack-ansible-os_tempest master: Set fact tempest_install_method var for meta handler https://review.opendev.org/672231 | 12:30 |
| chandankumar | jrosser: ^^ | 12:30 |
| noonedeadpunk | dunno either... | 12:30 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/ansible-config_template master: Fix writing of keys without values https://review.opendev.org/672952 | 12:34 |
| *** waverider has joined #openstack-ansible | 12:34 | |
| *** brtknr has quit IRC | 13:01 | |
| *** brtknr has joined #openstack-ansible | 13:03 | |
| *** brtknr has quit IRC | 13:05 | |
| *** brtknr has joined #openstack-ansible | 13:05 | |
| *** brtknr has quit IRC | 13:05 | |
| *** brtknr has joined #openstack-ansible | 13:06 | |
| *** brtknr has quit IRC | 13:06 | |
| *** brtknr has joined #openstack-ansible | 13:06 | |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/ansible-config_template master: Fix writing of keys without values https://review.opendev.org/672952 | 13:09 |
| *** ygk_12345 has joined #openstack-ansible | 13:10 | |
| *** grabes has joined #openstack-ansible | 13:17 | |
| *** jawad_axd has joined #openstack-ansible | 13:24 | |
| *** stingrayza_ has joined #openstack-ansible | 13:28 | |
| *** stingrayza has quit IRC | 13:28 | |
| *** goldyfruit has joined #openstack-ansible | 13:31 | |
| openstackgerrit | Florian Haas proposed openstack/openstack-ansible stable/stein: Fix systemd-journal-remote socket dependency https://review.opendev.org/672980 | 13:34 |
| *** jawad_axd has quit IRC | 13:37 | |
| *** waverider has quit IRC | 13:39 | |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/ansible-config_template master: Fix writing of keys without values https://review.opendev.org/672952 | 13:40 |
| *** Blinkiz has quit IRC | 13:47 | |
| *** ygk_12345 has quit IRC | 14:01 | |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/ansible-config_template master: Fix writing of keys without values https://review.opendev.org/672952 | 14:05 |
| *** radeks has joined #openstack-ansible | 14:07 | |
| ksdean | hi, Anyone got any idea why I cant deploy horizon from source. | 14:07 |
| ksdean | ? | 14:07 |
| ksdean | using kolla | 14:07 |
| cloudnull | noonedeadpunk what's going on ? | 14:08 |
| noonedeadpunk | cloudnull: http://logs.openstack.org/83/671783/8/check/openstack-ansible-deploy-aio_lxc-ubuntu-bionic/be4e974/logs/etc/openstack/aio1_nova_api_container-2e62b5d5/nova/nova.conf.txt.gz | 14:08 |
| cloudnull | ksdean this channel is largely for the openstack-ansible project you might want to reach out in the #openstack-kolla channel | 14:08 |
| noonedeadpunk | so I'm experimentally trying to patch this:) | 14:08 |
| ksdean | ahh ok cool | 14:09 |
| ksdean | thanks | 14:09 |
| cloudnull | np, have a good one ksdean | 14:09 |
| cloudnull | the config looks good :) | 14:09 |
| cloudnull | is there an issue with it, that Im not seeing ? | 14:09 |
| noonedeadpunk | so disk_cachemodes has no "=" | 14:09 |
| noonedeadpunk | which makes config not working one for nova | 14:10 |
| noonedeadpunk | I'm trying to address https://review.opendev.org/672952 and got pretty good results (but not really) | 14:10 |
| noonedeadpunk | like extra "\n" http://logs.openstack.org/52/672952/5/check/openstack-ansible-python3-ubuntu-bionic/916a2b7/job-output.txt.gz#_2019-07-26_13_52_25_703629 | 14:11 |
| *** ansmith has quit IRC | 14:11 | |
| *** goldenfri has quit IRC | 14:11 | |
| cloudnull | so the item has a None value | 14:11 |
| noonedeadpunk | yep, and you're returning just key | 14:12 |
| cloudnull | is disk_cachemodes being set as an override ? | 14:12 |
| noonedeadpunk | So the problem is here https://opendev.org/openstack/ansible-config_template/src/branch/master/action/config_template.py#L226 | 14:12 |
| noonedeadpunk | No, it's just the way config is being renderred even without overrides I guess | 14:13 |
| cloudnull | we had similar logic even before the patch - https://opendev.org/openstack/ansible-config_template/src/commit/73aa099f0a4871dc6ca63a5365b1010645dc0c72/action/config_template.py#L150-L159 | 14:14 |
| cloudnull | which is supposed to allow us to support options without values, like the ones found in my.cnf | 14:14 |
| *** hamzaachi has joined #openstack-ansible | 14:14 | |
| noonedeadpunk | So maybe I'm not right in seeing what exactly leads to this thing, but smth defenitely does... | 14:16 |
| cloudnull | hum ... | 14:18 |
| noonedeadpunk | and gates are broken, like https://review.opendev.org/#/c/671988/ | 14:19 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/ansible-config_template master: Fix writing of keys without values https://review.opendev.org/672952 | 14:25 |
| *** hamzaachi has quit IRC | 14:34 | |
| *** hamzaachi has joined #openstack-ansible | 14:34 | |
| *** miloa has quit IRC | 14:38 | |
| *** hamzaachi has quit IRC | 14:40 | |
| cloudnull | noonedeadpunk testing that and a few other things locally | 14:42 |
| *** hamzaachi has joined #openstack-ansible | 14:42 | |
| *** hamzaachi has quit IRC | 14:42 | |
| *** hamzaachi has joined #openstack-ansible | 14:42 | |
| cloudnull | noonedeadpunk do you by chance that the config file source handy? | 14:47 |
| cloudnull | I think its reading in that option as a none type | 14:47 |
| noonedeadpunk | cloudnull: not really. I'm even testing in CI... | 14:47 |
| cloudnull | ok | 14:47 |
| noonedeadpunk | but it's supposed to be empty string.... | 14:48 |
| cloudnull | ++ | 14:48 |
| *** chandankumar is now known as raukadah | 14:48 | |
| cloudnull | I think we need to get the module to respect that | 14:48 |
| cloudnull | noonedeadpunk I think we just need to adjust the item check | 14:49 |
| cloudnull | https://review.opendev.org/#/c/672952/7/action/config_template.py | 14:49 |
| cloudnull | checking to see if `item is not None` | 14:50 |
| *** openstackgerrit has quit IRC | 14:51 | |
| noonedeadpunk | ok, lets try this out | 14:51 |
| cloudnull | I think that solves it. | 14:54 |
| noonedeadpunk | I've just thought that you do strip() somewhere but didn't check that | 14:54 |
| cloudnull | at least it does locally . | 14:54 |
| *** openstackgerrit has joined #openstack-ansible | 14:55 | |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/ansible-config_template master: Fix writing of keys without values https://review.opendev.org/672952 | 14:55 |
| noonedeadpunk | but that version was really very close as well:) | 14:55 |
| noonedeadpunk | it wasn't respecting extra "\n" but still | 14:56 |
| *** kopecmartin is now known as kopecmartin|off | 14:56 | |
| cloudnull | ++ | 14:57 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/ansible-config_template master: Fix writing of keys without values https://review.opendev.org/672952 | 14:58 |
| cloudnull | ^ noonedeadpunk just fixing a nit | 14:58 |
| cloudnull | figured it better to fix before the job runs :) | 14:58 |
| noonedeadpunk | it was kinda part of the test :P | 14:58 |
| cloudnull | https://review.opendev.org/#/c/672952/8/tests/templates/test_default_section.ini | 14:59 |
| cloudnull | oh ?! | 14:59 |
| cloudnull | my bad | 14:59 |
| noonedeadpunk | since https://opendev.org/openstack/openstack-ansible-os_nova/src/branch/master/templates/nova.conf.j2#L264 will have whitespace:) | 15:00 |
| cloudnull | hum . i think the reader will pick that up as an empty string | 15:00 |
| * cloudnull trying locally | 15:00 | |
| noonedeadpunk | yeah, so I suggested about strip() usage on values and keys (but still didn't checked if it's true) | 15:01 |
| noonedeadpunk | (I thought that it was used) | 15:04 |
| cloudnull | so the latest review passes the test :) | 15:11 |
| cloudnull | which is a good thing :D | 15:11 |
| * cloudnull running tests locally | 15:11 | |
| cloudnull | oh wait, im an idiot, i didnt run the checkout. | 15:13 |
| cloudnull | running again | 15:13 |
| noonedeadpunk | CI will have result really soon:) | 15:13 |
| noonedeadpunk | need to add extra "\n" to expected file http://logs.openstack.org/52/672952/9/check/openstack-ansible-functional-ubuntu-bionic/e814d09/job-output.txt.gz#_2019-07-26_15_15_25_557627 | 15:17 |
| noonedeadpunk | but yeah, approach generally works | 15:18 |
| cloudnull | I was seeing the same thing. | 15:21 |
| cloudnull | I think that's a bug | 15:21 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/ansible-config_template master: Fix writing of keys without values https://review.opendev.org/672952 | 15:21 |
| noonedeadpunk | yeah, but it's not annoying:) | 15:21 |
| cloudnull | when I rerun a bunch of times, it adds \n to the modified sections | 15:21 |
| noonedeadpunk | so that my version was dropping this "\n" but was also taking extra ones. But since this way is much more cleaner, I'd leave it as is | 15:22 |
| cloudnull | so if we use the same file as input and output it adds \n, then \n\n, then \n\n\n and so forth | 15:22 |
| noonedeadpunk | wow, than it's not so good | 15:22 |
| cloudnull | looking at that now | 15:22 |
| cloudnull | in the osa case its not so bad | 15:23 |
| cloudnull | the base file always comes from a template | 15:23 |
| cloudnull | in tripleo its a problem | 15:23 |
| cloudnull | well a cosmetic problem | 15:23 |
| cloudnull | but a problem none-the-less | 15:23 |
| noonedeadpunk | probably it's worth returning to https://review.opendev.org/#/c/672952/7 and just drop these extra "\n" from expected files? | 15:23 |
| cloudnull | I think we get that test to pass. get the gates unbroken | 15:24 |
| noonedeadpunk | (and add "else: return str()") | 15:24 |
| cloudnull | and I'll build on-top | 15:24 |
| noonedeadpunk | these 2 options will unbroke gates... but ok. | 15:25 |
| *** cshen has quit IRC | 15:29 | |
| noonedeadpunk | so probably smth like "elif option == "\n": return option" will do the trick? | 15:33 |
| cloudnull | ++ | 15:37 |
| cloudnull | that might be | 15:37 |
| *** jbadiapa has quit IRC | 15:39 | |
| noonedeadpunk | cloudnull: again the same... http://logs.openstack.org/52/672952/10/check/openstack-ansible-functional-ubuntu-bionic/4880e91/job-output.txt.gz#_2019-07-26_15_37_29_570989 | 15:42 |
| noonedeadpunk | So I'll try to add elif... | 15:42 |
| noonedeadpunk | and start testing locally as well:) | 15:44 |
| noonedeadpunk | do you just ./runtest.sh functional? | 15:46 |
| noonedeadpunk | or probably there's an option to use python debbuger?:)) | 15:47 |
| cloudnull | I use pdb | 15:52 |
| *** joshualyle has joined #openstack-ansible | 15:52 | |
| cloudnull | but to test it started, I simply clone the repo and run the local tests script | 15:52 |
| *** joshualyle has quit IRC | 15:57 | |
| *** rpittau is now known as rpittau|afk | 15:57 | |
| *** gyee has joined #openstack-ansible | 15:59 | |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/ansible-config_template master: Fix writing of keys without values https://review.opendev.org/672952 | 16:10 |
| *** ksdean has quit IRC | 16:14 | |
| cloudnull | noonedeadpunk that last review works | 16:16 |
| cloudnull | just need to remove the extra space | 16:16 |
| cloudnull | https://review.opendev.org/#/c/672952/edit/tests/files/test_default_section.ini.expected | 16:16 |
| cloudnull | i have to run, but will be back in about 1 hour | 16:17 |
| mnaser | it might be good for us to run osa jobs against openstack/ansible-config_template :) | 16:17 |
| cloudnull | ++ | 16:17 |
| cloudnull | its running most of the setup | 16:17 |
| cloudnull | container creates and things as is | 16:17 |
| cloudnull | however, a nova / neutron job might go a long way | 16:18 |
| openstackgerrit | Chandan Kumar (raukadah) proposed openstack/openstack-ansible-os_tempest master: Set fact tempest_install_method var for meta handler https://review.opendev.org/672231 | 16:18 |
| raukadah | jrosser: ^^ feel free to take over it | 16:18 |
| noonedeadpunk | cloudnull: for me without this space task fails | 16:25 |
| noonedeadpunk | so elif probably doesn't do what's expected... | 16:26 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/ansible-config_template master: Fix writing of keys without values https://review.opendev.org/672952 | 16:32 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/ansible-config_template master: Fix writing of keys without values https://review.opendev.org/672952 | 16:32 |
| openstackgerrit | Chandan Kumar (raukadah) proposed openstack/openstack-ansible-os_tempest master: Set fact tempest_install_method var for meta handler https://review.opendev.org/672231 | 16:40 |
| raukadah | noonedeadpunk: I kept your default also and set_facts it might solve both the problem | 16:41 |
| noonedeadpunk | raukadah: you have indent problem | 16:41 |
| noonedeadpunk | in when | 16:41 |
| openstackgerrit | Guilherme Steinmuller Pimentel proposed openstack/openstack-ansible-os_ironic stable/stein: Fix ironic package list https://review.opendev.org/673017 | 16:41 |
| raukadah | stupid typos | 16:42 |
| noonedeadpunk | just a sec | 16:42 |
| openstackgerrit | Chandan Kumar (raukadah) proposed openstack/openstack-ansible-os_tempest master: Set fact tempest_install_method var for meta handler https://review.opendev.org/672231 | 16:42 |
| raukadah | noonedeadpunk: feel free to update the patch, | 16:43 |
| noonedeadpunk | I guess jrosser was saying about smth like this http://paste.openstack.org/show/754901/ | 16:43 |
| * jrosser back | 16:43 | |
| jrosser | curious it needs the default :) but this whole thing is curious | 16:44 |
| raukadah | noonedeadpunk: but it will start installing stuff from pip which we donot want | 16:44 |
| raukadah | they will create a mix of both which always screws | 16:44 |
| noonedeadpunk | raukadah: only when tempest_install_method is not defined I guess | 16:45 |
| raukadah | set_facts should be happen after pip install | 16:45 |
| jrosser | raukadah: i just don't understand now | 16:45 |
| jrosser | the role defaults has install method being source | 16:45 |
| raukadah | let me try that | 16:45 |
| jrosser | if you want distro, then you've surely got to be overriding that in the playbook | 16:46 |
| jrosser | or vars somewhere | 16:46 |
| jrosser | so if you've got a higher level override, how do we get into this situation where the handler runs somewhere that the override is not defined? | 16:47 |
| raukadah | jrosser: donot know, need to debug locally | 16:50 |
| noonedeadpunk | I guess it's just friday evening:) | 16:50 |
| *** KeithMnemonic has joined #openstack-ansible | 16:50 | |
| raukadah | jrosser: in playbook, it is already defined to use tempest_install_method as distro | 16:50 |
| jrosser | raukadah: can you point me to the code where os_tempest is included? | 16:50 |
| raukadah | jrosser: tempest playbook https://opendev.org/openstack/tripleo-quickstart-extras/src/branch/master/playbooks/tempest.yml called here https://opendev.org/openstack/tripleo-quickstart-extras/src/branch/master/playbooks/multinode-standalone.yml#L53 | 16:52 |
| jrosser | raukadah: ok so this line https://opendev.org/openstack/tripleo-quickstart-extras/src/branch/master/playbooks/multinode-standalone.yml#L54 | 16:54 |
| *** joshualyle has joined #openstack-ansible | 16:55 | |
| jrosser | thats where the var tempest_install_method is defined, and it's scope is only for that task | 16:55 |
| jrosser | so when the handler runs at the and of the play it fails because the var is not in scope | 16:55 |
| *** joshualyle has quit IRC | 16:55 | |
| raukadah | do I need to make it caceable? | 16:56 |
| jrosser | you could try moving the defintion of the var up to the vars: section at the top of the play | 16:56 |
| jrosser | and then it will be in scope for the whole of the play, including when the handlers run | 16:56 |
| *** goldenfri has joined #openstack-ansible | 16:56 | |
| raukadah | jrosser: yes, doing that | 16:57 |
| *** ansmith has joined #openstack-ansible | 17:03 | |
| *** EmilienM has joined #openstack-ansible | 17:04 | |
| EmilienM | o/ | 17:05 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/ansible-config_template master: Fix writing of keys without values https://review.opendev.org/672952 | 17:06 |
| raukadah | jrosser: noonedeadpunk https://review.opendev.org/#/c/673021/ | 17:07 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/ansible-config_template master: Fix writing of keys without values https://review.opendev.org/672952 | 17:07 |
| raukadah | testing here https://review.opendev.org/#/c/667219/ | 17:09 |
| raukadah | jrosser: noonedeadpunk thanks for all the help on this :-) | 17:10 |
| raukadah | Happy weekend | 17:10 |
| jrosser | raukadah: you too :) | 17:11 |
| *** weifan has joined #openstack-ansible | 17:28 | |
| noonedeadpunk | jrosser cloudnull: https://review.opendev.org/672952 is finally going to pass, so kindly check it right after tripleo finish it's test | 17:29 |
| noonedeadpunk | so have great weekends everyone:) | 17:30 |
| *** udesale has quit IRC | 17:31 | |
| *** electrofelix has quit IRC | 17:38 | |
| *** hwoarang has quit IRC | 17:42 | |
| *** hwoarang_ has joined #openstack-ansible | 17:43 | |
| guilhermesp | cloudnull: still around? if so, did you test this in a metal environment? https://github.com/openstack/openstack-ansible-os_nova/blob/f11199f27e49e1ebe23e181ab7d6046fba428197/tasks/nova_virt_detect.yml#L54 | 17:48 |
| *** hwoarang_ has quit IRC | 17:54 | |
| *** hwoarang has joined #openstack-ansible | 17:56 | |
| *** chason has quit IRC | 17:56 | |
| *** hwoarang has quit IRC | 18:03 | |
| *** hwoarang has joined #openstack-ansible | 18:04 | |
| *** poopcat has joined #openstack-ansible | 18:04 | |
| *** joshualyle has joined #openstack-ansible | 18:09 | |
| *** joshualyle has quit IRC | 18:12 | |
| *** weifan has quit IRC | 18:16 | |
| *** weifan has joined #openstack-ansible | 18:17 | |
| *** weifan has quit IRC | 18:17 | |
| *** hamzaachi has quit IRC | 18:18 | |
| *** hwoarang has quit IRC | 18:32 | |
| *** hwoarang has joined #openstack-ansible | 18:32 | |
| *** weifan has joined #openstack-ansible | 18:34 | |
| *** weifan has quit IRC | 18:34 | |
| cloudnull | guilhermesp I'm not 100% sure. | 18:35 |
| cloudnull | I think so ? | 18:35 |
| cloudnull | I believe RAX is running ironic on baremetal still | 18:35 |
| cloudnull | cjloader ^ | 18:35 |
| guilhermesp | huum so ironic_compute_containers are on metal, in controller nodes. But seems that even defining https://github.com/cloudnull/os-ansible-deployment/blob/master/group_vars/ironic_compute.yml or in host_vars, the var nova_virt_type is not being loaded | 18:36 |
| guilhermesp | so tasks like that onehttps://github.com/openstack/openstack-ansible-os_nova/blob/f11199f27e49e1ebe23e181ab7d6046fba428197/tasks/nova_virt_detect.yml#L54 are ommited | 18:36 |
| *** weifan has joined #openstack-ansible | 18:38 | |
| guilhermesp | it starts skipping right here https://github.com/openstack/openstack-ansible-os_nova/blob/f11199f27e49e1ebe23e181ab7d6046fba428197/tasks/main.yml#L28 so which means the host_vars or groups vars defining nova_virt_type: ironic is not there | 18:40 |
| *** weifan has quit IRC | 18:42 | |
| guilhermesp | seems we have something similar here http://eavesdrop.openstack.org/irclogs/%23openstack-ansible/%23openstack-ansible.2017-08-17.log.html#t2017-08-17T05:48:34 | 18:50 |
| cloudnull | hum I'm really not sure. | 18:54 |
| * cloudnull doesn't have an osa ironic lab up at the moment | 18:54 | |
| cloudnull | I was kinda hoping cjloader was around and could check | 18:54 |
| *** weifan has joined #openstack-ansible | 19:07 | |
| *** weifan has quit IRC | 19:08 | |
| *** weifan has joined #openstack-ansible | 19:08 | |
| *** weifan has quit IRC | 19:13 | |
| spotz | cloudnull: I know he peeked in the other day:( | 19:21 |
| *** weifan has joined #openstack-ansible | 19:32 | |
| *** weifan has quit IRC | 19:36 | |
| *** weifan has joined #openstack-ansible | 19:36 | |
| *** weifan has quit IRC | 19:37 | |
| *** weifan has joined #openstack-ansible | 19:37 | |
| *** weifan has quit IRC | 19:38 | |
| *** weifan has joined #openstack-ansible | 19:38 | |
| openstackgerrit | Merged openstack/ansible-config_template master: Fix writing of keys without values https://review.opendev.org/672952 | 19:38 |
| *** weifan has quit IRC | 19:39 | |
| *** weifan has joined #openstack-ansible | 19:39 | |
| *** weifan has quit IRC | 19:39 | |
| *** weifan has joined #openstack-ansible | 19:40 | |
| *** brtknr has quit IRC | 19:40 | |
| *** brtknr has joined #openstack-ansible | 19:40 | |
| *** weifan has quit IRC | 19:40 | |
| *** BjoernT has joined #openstack-ansible | 19:51 | |
| *** BjoernT has quit IRC | 19:55 | |
| *** allanb has quit IRC | 19:56 | |
| cjloader | cloudnull: guilhermesp yes we are | 19:59 |
| *** kplant has quit IRC | 20:00 | |
| *** goldyfruit has quit IRC | 20:02 | |
| *** goldyfruit has joined #openstack-ansible | 20:04 | |
| cjloader | guilhermesp: what do you need checked specifically? | 20:05 |
| cjloader | we are using metal ironic osa | 20:05 |
| guilhermesp | hey cjloader | 20:05 |
| guilhermesp | yeah we are trying to deploy it on metal as well | 20:06 |
| guilhermesp | seems that nova_virt_type is not being evaluated | 20:06 |
| guilhermesp | even here is not being executed https://github.com/openstack/openstack-ansible-os_nova/blob/f11199f27e49e1ebe23e181ab7d6046fba428197/tasks/nova_virt_detect.yml#L48 | 20:06 |
| guilhermesp | i tried to define host_vars/ctl with nova_virt_type: ironic but seems that has no effect | 20:07 |
| cjloader | how recent is the deployment? | 20:07 |
| guilhermesp | stein | 20:07 |
| cjloader | i'll look at my lab | 20:07 |
| cjloader | hold on | 20:07 |
| guilhermesp | I appreciate that cjloader | 20:07 |
| cjloader | /etc/openstack_deploy/openstack_user_config.yml:111: nova_virt_type: ironic | 20:10 |
| cjloader | guilhermesp: ^ | 20:10 |
| guilhermesp | yep. I have tried that. Actually I wanted to make it per host inside host_vars | 20:11 |
| guilhermesp | as ironic_compute is going to run within the controllers | 20:11 |
| *** weifan has joined #openstack-ansible | 20:11 | |
| cjloader | container_vars: | 20:11 |
| cjloader | nova_virt_type: ironic | 20:11 |
| cjloader | add that in host def | 20:12 |
| cjloader | in your openstack_user_config | 20:12 |
| guilhermesp | ok let me try | 20:12 |
| cjloader | that's how my lab is setup | 20:13 |
| guilhermesp | so but this is going to run under all compute hosts right? | 20:13 |
| guilhermesp | we have some kvm/qemu computes | 20:13 |
| cjloader | unless you define it in a separate non kvm/qemu computes | 20:14 |
| guilhermesp | yeah, that's why I was trying to define it under hosts_vars. | 20:14 |
| guilhermesp | I will replace in my host_vars file the content there by the content you use in o_u_c | 20:15 |
| cjloader | i put infra01: blah blah container_vars: ... | 20:15 |
| cjloader | etc | 20:16 |
| *** weifan has quit IRC | 20:16 | |
| guilhermesp | I see... yeah under hosts_vars doesn't work. I will define it per node in o_u_c and try it again | 20:16 |
| cjloader | my setup assumes no outside kvm/qemu hosts | 20:16 |
| cjloader | guilhermesp: let me know, brb | 20:20 |
| guilhermesp | yeah that's weird. Same thing. All debugging shows that tasks that requires nova_virt_type to be define are being ommited saying that nova_virt_type is undefined | 20:20 |
| guilhermesp | thanks cjloader :) | 20:20 |
| *** brtknr_ has joined #openstack-ansible | 20:20 | |
| cjloader | http://paste.openstack.org/show/754909/ | 20:21 |
| cjloader | guilhermesp: ^ paste from my o_u_c | 20:22 |
| * guilhermesp looking | 20:22 | |
| *** brtknr has quit IRC | 20:22 | |
| cjloader | cc: cloudnull ^ | 20:24 |
| guilhermesp | yeah no effect.. seems that tasks are being ommited still https://pasteboard.co/IpOrQb0.png | 20:24 |
| cjloader | do you have a compute_hosts block? | 20:26 |
| * guilhermesp that's how it is now http://paste.openstack.org/show/754910/ | 20:27 | |
| cjloader | it might being overridden | 20:27 |
| cjloader | my setup has 0 computes | 20:28 |
| cjloader | might want to set an override | 20:30 |
| guilhermesp | but I'd assume at least nova_virt_type would be set as when it is not defined, sets defaults to qemu | 20:30 |
| guilhermesp | https://github.com/openstack/openstack-ansible-os_nova/blob/f11199f27e49e1ebe23e181ab7d6046fba428197/tasks/nova_virt_detect.yml#L48 | 20:30 |
| cjloader | set a debug: var={{ nova_virt_type }} | 20:31 |
| cjloader | and lets see what its being set to | 20:31 |
| guilhermesp | all right "ironic": "VARIABLE IS NOT DEFINED!" | 20:33 |
| cjloader | that help? | 20:34 |
| guilhermesp | not that much https://github.com/openstack/openstack-ansible-os_nova/blob/f11199f27e49e1ebe23e181ab7d6046fba428197/defaults/main.yml#L192 | 20:35 |
| cjloader | okay | 20:35 |
| cjloader | are you mixing hosts -- some kvm/qemu? | 20:37 |
| guilhermesp | the two I have up are qemu | 20:38 |
| cjloader | so tell me your lab setup | 20:39 |
| guilhermesp | we have two qemu computes. Are are deploying everything with no containers at all. All serives are set to metal = true in env.d/metal. So that's basically this. | 20:40 |
| guilhermesp | ironic creates a container to simulate a compute_hosts. In that case, as everything is on metal, controllers will play that role | 20:41 |
| cjloader | no controllers? | 20:41 |
| cjloader | oh | 20:41 |
| guilhermesp | yep we have 3 controllers ( ctl1,ctl2 and ctl3 ) | 20:41 |
| guilhermesp | it's pretty much that paste here http://paste.openstack.org/show/754910/ | 20:42 |
| guilhermesp | qemu kvms goes to compute_hosts | 20:42 |
| cjloader | we have 3 ctl and 2 infra (haproxy/keepalived) and no computes | 20:42 |
| *** weifan has joined #openstack-ansible | 20:42 | |
| cjloader | and use ironic to spin up computes | 20:43 |
| guilhermesp | i see. In that case we are gonna proide vms and baremetal instances | 20:43 |
| guilhermesp | provide* | 20:43 |
| cjloader | generally I thought nova assigned computes to the ctl pln if no computes available | 20:46 |
| cjloader | I could be wrong | 20:47 |
| *** goldyfruit has quit IRC | 20:48 | |
| guilhermesp | i think acually wherever you define it in o_u_c | 20:49 |
| cjloader | sorry if i'm not able to help guilhermesp | 20:55 |
| guilhermesp | that's ok cjloader ! I appreciate your infos! I will let you know if i figure that out | 20:55 |
| cjloader | might ask in the ironic channel? | 20:56 |
| *** kplant has joined #openstack-ansible | 20:58 | |
| *** goldyfruit has joined #openstack-ansible | 21:05 | |
| guilhermesp | I noticed that when I remove ironic from that list https://github.com/openstack/openstack-ansible-os_nova/blob/f11199f27e49e1ebe23e181ab7d6046fba428197/defaults/main.yml#L192 it breaks with another error http://paste.openstack.org/show/754911/ | 21:05 |
| guilhermesp | "ironic" is not defined anyware anymore in my config dir | 21:05 |
| jrosser | noonedeadpunk: we are still a bit broken even though the config_template change merged, see #infra | 21:17 |
| *** KeithMnemonic has quit IRC | 21:18 | |
| openstackgerrit | Mohammed Naser proposed openstack/openstack-ansible master: wip: fail when roles arent cloned by zuul https://review.opendev.org/673109 | 21:21 |
| jrosser | mnaser: https://opendev.org/openstack/openstack-ansible/src/branch/master/scripts/get-ansible-role-requirements.yml#L76 | 21:26 |
| jrosser | ^ that just looks odd | 21:26 |
| mnaser | jrosser: pretty much its saying in that all the roles that *dont* exist (as in they were not cloned by zuul), to clone them manually | 21:27 |
| jrosser | well i have to say this has always been confusing | 21:28 |
| jrosser | ok yes i see they should be symlinked into place in earlier tasks | 21:29 |
| jrosser | hence don't clone | 21:30 |
| openstackgerrit | Mohammed Naser proposed openstack/openstack-ansible master: wip: fail when roles arent cloned by zuul https://review.opendev.org/673109 | 21:30 |
| mnaser | jrosser: yep so zuul manually clones things | 21:30 |
| mnaser | so we actually want to check for the things that are cloned and symlink in /etc/ansible/roles | 21:30 |
| mnaser | i just realized that code won't work properly because we have somethings we pull in from github, yikes | 21:31 |
| jrosser | which should be this https://opendev.org/openstack/openstack-ansible/src/branch/master/scripts/get-ansible-role-requirements.yml#L50-L59 | 21:31 |
| jrosser | right so it's always a mix | 21:31 |
| jrosser | link the things we're provided, clone the rest | 21:31 |
| mnaser | jrosser: yep | 21:32 |
| mnaser | i have an idea tho | 21:32 |
| mnaser | jrosser: i think ill only select values with 'opendev' in them but ill wait for the CI to run | 21:34 |
| jrosser | you can probably json_query the list down quite neatly | 21:35 |
| jrosser | anyway, !computer now | 21:36 |
| mnaser | ave a good weeknd jrosser | 21:36 |
| jrosser | thanks, you too | 21:36 |
| openstackgerrit | Mohammed Naser proposed openstack/openstack-ansible master: wip: fail when roles arent cloned by zuul https://review.opendev.org/673109 | 21:40 |
| noonedeadpunk | jrosser: yeah, I see tasks are failing, but now I rly don't understand why | 21:46 |
| noonedeadpunk | since config_template is covered with common test.. | 21:47 |
| cjloader | mnaser: btw i'm working on the combined ironic/inspector role | 21:48 |
| cjloader | close -- but was wondering, what the best way to get it in? | 21:48 |
| cjloader | master, backport to stein? | 21:49 |
| spotz | unless it's something removed in master yes, otherwise always master then backport | 21:50 |
| spotz | cjloader: ^ | 21:50 |
| cjloader | ok ty | 21:51 |
| cjloader | i have it working in stein... | 21:51 |
| cjloader | fun | 21:52 |
| noonedeadpunk | looked up for details in infra chanel, but didn't fully realized the root cause:( | 21:55 |
| *** brtknr_ has quit IRC | 22:01 | |
| noonedeadpunk | so what I get is that during checks some repos are taken on out of date state due to which checks are not so relevant. | 22:04 |
| noonedeadpunk | since they are taken not from zuul but from mirrors | 22:05 |
| *** gyee has quit IRC | 22:39 | |
| *** goldyfruit has quit IRC | 22:52 | |
| *** weifan has quit IRC | 22:53 | |
| *** gyee has joined #openstack-ansible | 22:55 | |
| *** weifan has joined #openstack-ansible | 22:59 | |
| *** weifan has quit IRC | 23:03 | |
| *** tosky has quit IRC | 23:31 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!
