| *** jamesdenton has joined #openstack-ansible | 01:17 | |
| *** jamesdenton has quit IRC | 01:22 | |
| *** jamesdenton has joined #openstack-ansible | 01:24 | |
| *** itandops has quit IRC | 02:28 | |
| *** gyee has quit IRC | 02:38 | |
| *** mmethot_ has joined #openstack-ansible | 03:54 | |
| *** mmethot has quit IRC | 03:56 | |
| *** schwicht has quit IRC | 04:06 | |
| *** schwicht has joined #openstack-ansible | 04:14 | |
| *** evrardjp has quit IRC | 04:36 | |
| *** evrardjp has joined #openstack-ansible | 04:36 | |
| *** KeithMnemonic has quit IRC | 05:14 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible master: Bump ansible version to 2.9.9 https://review.opendev.org/727925 | 05:28 |
|---|---|---|
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-tests master: Update ansible to 2.9.9 https://review.opendev.org/727926 | 05:30 |
| *** udesale has joined #openstack-ansible | 05:30 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible master: Update Magnum fedora image name https://review.opendev.org/727749 | 05:50 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_manila master: Ansible 'skipped' is a test not a filter https://review.opendev.org/727743 | 05:56 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-rsyslog_server master: [ussuri][goal] Updates for python 2.7 drop https://review.opendev.org/710680 | 05:59 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-rsyslog_server master: Updated from OpenStack Ansible Tests https://review.opendev.org/709519 | 05:59 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-plugins master: Ansible 'changed' is a test not a filter https://review.opendev.org/727932 | 06:02 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-plugins master: Updated from OpenStack Ansible Tests https://review.opendev.org/727334 | 06:03 |
| *** CeeMac has joined #openstack-ansible | 06:16 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-galera_server master: [WIP] combine galera client role into galera_server https://review.opendev.org/725903 | 06:20 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_octavia master: Replace usage of deprecated os_*_facts ansible modules https://review.opendev.org/727934 | 06:31 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_trove master: Replace usage of deprecated os_*_facts ansible modules https://review.opendev.org/727939 | 06:36 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_swift master: Replace usage of deprecated os_*_facts ansible modules https://review.opendev.org/727940 | 06:37 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_designate master: Replace usage of deprecated os_*_facts ansible modules https://review.opendev.org/727941 | 06:39 |
| noonedeadpunk | jrosser: have any thoughts on https://review.opendev.org/#/c/726228/ and topic overall? | 06:45 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-lxc_container_create master: Use lxc-3 syntax for tests https://review.opendev.org/665238 | 06:54 |
| chandankumar | jroll, noonedeadpunk morning | 06:59 |
| chandankumar | jrosser, morning | 06:59 |
| chandankumar | jroll, sorry | 06:59 |
| noonedeadpunk | o/ | 06:59 |
| noonedeadpunk | they should get used to I think :p | 06:59 |
| chandankumar | jrosser, I was trying to trigger octavia test here https://review.opendev.org/727856 but it is not working | 06:59 |
| chandankumar | noonedeadpunk, hehe | 06:59 |
| chandankumar | here is the series https://review.opendev.org/#/q/topic:octavia_ostempest+(status:open+OR+status:merged) | 07:00 |
| noonedeadpunk | chandankumar: yeah as you need to do that in integrated repo | 07:00 |
| chandankumar | the same is working on tripleo side | 07:00 |
| noonedeadpunk | like https://opendev.org/openstack/openstack-ansible/src/branch/master/tests/roles/bootstrap-host/templates/user_variables_manila.yml.j2 | 07:00 |
| chandankumar | noonedeadpunk, integrated repo means openstack-ansible or opestack-ansible-tests | 07:00 |
| chandankumar | noonedeadpunk, Ah interesting | 07:01 |
| chandankumar | , let me do that | 07:01 |
| noonedeadpunk | so just put another template and add it to the list https://opendev.org/openstack/openstack-ansible/src/branch/master/tests/roles/bootstrap-host/tasks/prepare_aio_config.yml#L158 | 07:02 |
| noonedeadpunk | variables you mentioned are for functional tests which were ran by tox which is almost not the case nowadays | 07:03 |
| *** fghaas has joined #openstack-ansible | 07:16 | |
| jrosser | chandankumar: it is quite possible we already are running octavia tests | 07:17 |
| jrosser | noonedeadpunk: good morning! | 07:17 |
| jrosser | many many things getting fixed - this is good | 07:17 |
| chandankumar | jrosser, I grepped for octavia tests but not found | 07:19 |
| openstackgerrit | Chandan Kumar (raukadah) proposed openstack/openstack-ansible master: Enable octavia tempest tests https://review.opendev.org/727949 | 07:19 |
| jrosser | yes i'm just taking a look now | 07:19 |
| jrosser | heres the tempest log from a recently merged patch to os_octavia, and yes it doesnt run any octavia tests there | 07:20 |
| jrosser | https://zuul.opendev.org/t/openstack/build/83d38a71b7894617abc30749c2c234b7/log/logs/openstack/aio1-utility/tempest_run.log.txt | 07:20 |
| openstackgerrit | Chandan Kumar (raukadah) proposed openstack/openstack-ansible-os_octavia master: [WIP] enable octavia tempest tests https://review.opendev.org/727856 | 07:21 |
| chandankumar | jrosser, noonedeadpunk https://review.opendev.org/727949 and https://review.opendev.org/727856 will trigger octavia tests or I am still missing something | 07:21 |
| chandankumar | or I need to enable octavia somewhere in metal jobs? | 07:22 |
| jrosser | from osa side we already enable the octavia tempest plugin https://zuul.opendev.org/t/openstack/build/83d38a71b7894617abc30749c2c234b7/log/job-output.txt#15959 | 07:22 |
| jrosser | chandankumar: a couple of things - i don't think the condition for creating the network is needed in os_tempest for us https://review.opendev.org/#/c/727854/ | 07:23 |
| jrosser | neutron is always present when testing octavia | 07:23 |
| chandankumar | jrosser, but when I enable neutron it create public and private networks both | 07:24 |
| chandankumar | for octavia we need only public network | 07:24 |
| jrosser | ah ok, does it cause issues if the private network gets created there, or you just want to keep things as minimal as needed? | 07:25 |
| jrosser | then the question about triggering octavia tests in the openstack-ansible integrated repo is here https://github.com/openstack/openstack-ansible/blob/master/zuul.d/playbooks/pre-gate-scenario.yml#L39-L42 | 07:26 |
| chandankumar | jrosser, yes I want to keep it minimal as much as posible | 07:26 |
| jrosser | we create a veriable called "scenario" and the zuul pre-job generates the senario string and will add "octavia" if the test is running for os_octavia role | 07:26 |
| noonedeadpunk | chandankumar: placed some comments | 07:26 |
| jrosser | so that should just all happen for you | 07:27 |
| chandankumar | jrosser, let me do one more testing in TQ side, rethink on my public and private network hypothesis | 07:27 |
| jrosser | sure, no worries - tbh this feels quite close on the OSA side, just needs the tests enabling i think | 07:28 |
| noonedeadpunk | chandankumar: so actually https://review.opendev.org/#/c/727856/ is not needed as patch - we can still use it as noop patch | 07:29 |
| chandankumar | noonedeadpunk, yes, we can make it as a noop patch | 07:29 |
| jrosser | noonedeadpunk: i was thinking that a bunch of roles we still carry old functional tests - i guess this is an example of it | 07:29 |
| noonedeadpunk | yeah, was thinking about cleaning that up as well | 07:30 |
| jrosser | some we still use, like keystone and neutron have functional + integrated | 07:30 |
| jrosser | but a big bunch are now integrated only | 07:30 |
| noonedeadpunk | but probably need to leave that for V as well... | 07:31 |
| jrosser | yep, oh did you see i found another magnum thing - they have a default flavor name now in the tempest plugin | 07:31 |
| *** tosky has joined #openstack-ansible | 07:31 | |
| noonedeadpunk | Actually we use linters checks locally everywhre | 07:31 |
| noonedeadpunk | but we can use integrated now as well even for it | 07:32 |
| *** rpittau|afk is now known as rpittau | 07:33 | |
| noonedeadpunk | jrosser: no, I didn't.... | 07:33 |
| noonedeadpunk | but see now | 07:33 |
| noonedeadpunk | So you set master_flavor_id, does it also need flavor_id for pods? | 07:34 |
| jrosser | err :) maybe | 07:35 |
| jrosser | oh i see flavor_id vs master_flavor_id | 07:36 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible master: Update Magnum fedora image name https://review.opendev.org/727749 | 07:36 |
| jrosser | yeah thats going to break | 07:36 |
| noonedeadpunk | oh, I didn't check ci lol | 07:37 |
| jrosser | this is like errrrr what? https://opendev.org/openstack/magnum-tempest-plugin/src/branch/master/magnum_tempest_plugin/config.py#L67 | 07:37 |
| jrosser | how are you supposed to know the uuid.... | 07:37 |
| noonedeadpunk | you can set it while creating flavor | 07:38 |
| noonedeadpunk | disregrd | 07:38 |
| *** dpaclt has joined #openstack-ansible | 07:38 | |
| noonedeadpunk | I mixed with flavor lol | 07:38 |
| jrosser | maybe it's a magical field which will take name or uuid | 07:39 |
| noonedeadpunk | It doesn't look like it tbh... but who knows | 07:44 |
| *** born2bake has joined #openstack-ansible | 08:49 | |
| *** sshnaidm|afk is now known as sshnaidm | 09:11 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_trove master: Replace usage of deprecated os_*_facts ansible modules https://review.opendev.org/727939 | 09:25 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_designate master: Replace usage of deprecated os_*_facts ansible modules https://review.opendev.org/727941 | 09:26 |
| openstackgerrit | Merged openstack/openstack-ansible-plugins master: Ansible 'changed' is a test not a filter https://review.opendev.org/727932 | 09:28 |
| openstackgerrit | Merged openstack/openstack-ansible-plugins master: Updated from OpenStack Ansible Tests https://review.opendev.org/727334 | 09:29 |
| *** amalrajgenocidex has joined #openstack-ansible | 09:36 | |
| amalrajgenocidex | Hi, Is openstack-ansible Ussuri stable on ubuntu at the moment? | 09:37 |
| amalrajgenocidex | I was going for a new deployment and wanted to decide the version to use | 09:37 |
| jrosser | amalrajgenocidex: the core openstack components for Ussuri are just released | 09:57 |
| jrosser | but deployment tooling is in a difficult spot because they have to validate that the official Ussuri release is working with that deployment, and that takes some time | 09:59 |
| jrosser | openstack-ansible and other deployment tools make their release later than the core components | 09:59 |
| amalrajgenocidex | jrosser Thank you. I will go with train then :) | 10:02 |
| *** rpittau is now known as rpittau|bbl | 10:18 | |
| noonedeadpunk | spotz: jrosser so, really would love to hear thoughs about https://review.opendev.org/#/c/726228/ as it's our U blocker tbh... | 10:28 |
| noonedeadpunk | we really need to branch and don't want to backport all this stuff later... | 10:29 |
| noonedeadpunk | Or maybe we should backport? | 10:29 |
| *** viks____ has joined #openstack-ansible | 10:30 | |
| *** fghaas has left #openstack-ansible | 10:33 | |
| chandankumar | jrosser, https://storage.bhs.cloud.ovh.net/v1/AUTH_dcaab5e32b234d56b626f72581e3644c/zuul_opendev_logs_05a/727856/2/check/openstack-ansible-deploy-aio_metal-centos-7/05a1800/logs/openstack/aio1-utility/tempest_run.log.txt | 10:33 |
| noonedeadpunk | yeah, look like temepst runs octavia tests lol | 10:34 |
| chandankumar | yesh but it failed | 10:35 |
| noonedeadpunk | also in https://storage.bhs.cloud.ovh.net/v1/AUTH_dcaab5e32b234d56b626f72581e3644c/zuul_opendev_logs_05a/727856/2/check/openstack-ansible-deploy-aio_metal-centos-7/05a1800/logs/host/octavia-housekeeping.service.journal-09-01-33.log.txt Flow 'octavia-create-amphora-flow': octavia.common.exceptions.ComputeBuildException: Failed to build compute instance due to: Invalid key_name provided. (HTTP 400) | 10:37 |
| noonedeadpunk | Yeah and see policy warnings in api | 10:39 |
| noonedeadpunk | but I don't see that we provide policy... I think we expect them to be benerated by octavia.. | 10:41 |
| noonedeadpunk | btw, I think we should probably bother regarding moving policies.json to be yaml instead.... | 10:42 |
| openstackgerrit | Merged openstack/openstack-ansible-lxc_container_create master: Use lxc-3 syntax for tests https://review.opendev.org/665238 | 10:50 |
| openstackgerrit | Merged openstack/ansible-role-systemd_networkd master: Do not use deprecated ansible 'bare' variables https://review.opendev.org/727105 | 10:55 |
| openstackgerrit | Merged openstack/openstack-ansible-lxc_container_create master: Small syntax update https://review.opendev.org/665418 | 11:00 |
| *** dpaclt has quit IRC | 11:05 | |
| openstackgerrit | Merged openstack/openstack-ansible master: Install python3 instead of python2 during host bootstrapping. https://review.opendev.org/723276 | 11:07 |
| *** dpaclt has joined #openstack-ansible | 11:13 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_rally master: Updated from OpenStack Ansible Tests https://review.opendev.org/727327 | 11:18 |
| *** dpaclt has quit IRC | 11:37 | |
| *** udesale has quit IRC | 11:44 | |
| *** udesale has joined #openstack-ansible | 11:58 | |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible master: [WIP] Resolve issue with undefined facts while running with tags https://review.opendev.org/728052 | 12:04 |
| noonedeadpunk | jrosser: have a minute for discussion of ^ ? | 12:04 |
| *** udesale_ has joined #openstack-ansible | 12:04 | |
| noonedeadpunk | like having several options to resolve this thing, so'd better ask if you ok with this one or I need to do try another ones before changing all playbooks | 12:05 |
| noonedeadpunk | so as alternative to setting `always` tag on upper play we can use this https://opendev.org/openstack/openstack-ansible/src/branch/master/playbooks/containers-nspawn-create.yml#L16-L18 or this https://opendev.org/openstack/openstack-ansible/src/branch/master/playbooks/ceph-rgw-install.yml#L28-L37 way | 12:06 |
| noonedeadpunk | just want to pick one and follow it wherever possible | 12:07 |
| * jrosser meetings :( | 12:08 | |
| jrosser | i'll be about an hour | 12:09 |
| noonedeadpunk | sure | 12:09 |
| *** rpittau|bbl is now known as rpittau | 12:15 | |
| *** rh-jelabarre has joined #openstack-ansible | 12:22 | |
| *** amalrajgenocidex has quit IRC | 12:22 | |
| *** jbadiapa has joined #openstack-ansible | 12:25 | |
| spotz | noonedeadpunk: looking | 13:00 |
| spotz | noonedeadpunk: So technically the last patch is mine so unless it's minor like a typo I usually don't review. But I can if you want or you can by the same thought it's my patch | 13:01 |
| noonedeadpunk | spotz: I'm totally fine with your edit:) | 13:04 |
| noonedeadpunk | but I'm kinda laways feel unconfident in such cases when to vote or not | 13:05 |
| spotz | Ok I'll throw a vote up if you've read it | 13:08 |
| spotz | Ok voted, if jrosser or no one else votes by the end of the day lets workflow it | 13:09 |
| chandankumar | jrosser, hello | 13:10 |
| chandankumar | jrosser, https://31f867d906c85b3d450a-9945c6e94a1b9feee2753a6244e7fd57.ssl.cf2.rackcdn.com/726752/8/check/tripleo-ci-centos-8-scenario010-standalone/7081ef6/job-output.txt | 13:10 |
| chandankumar | jrosser, if we enable neutron private network creation fails ConflictException: 409: Client Error for url: http://192.168.24.1:9696/v2.0/networks.json, Unable to create the network. The tunnel ID 1 is in use. | 13:10 |
| guilhermesp | spotz: noonedeadpunk will take a look in a bit too at this topic | 13:10 |
| spotz | Thanks guilhermesp | 13:11 |
| guilhermesp | it seems fine to me, however i have a tittle question: do we want to keep "contribute" and "contributing" here? https://review.opendev.org/#/c/726228/5/doc/source/contributor/index.rst | 13:20 |
| noonedeadpunk | that's damn good question | 13:20 |
| noonedeadpunk | Like we have to have contributing stuck to the template | 13:20 |
| noonedeadpunk | while contribute is wider and better imo... | 13:20 |
| noonedeadpunk | s/stuck/stick/' | 13:21 |
| spotz | I questioned that mentally myself but 2 different docs | 13:21 |
| noonedeadpunk | Yeah... Like at first I was trying to adjust contributing but like data there by template is taken from at least 3 different docs... | 13:22 |
| noonedeadpunk | and dunno if combining them is good idea considering their size... | 13:23 |
| jamesdenton | good morning | 13:26 |
| guilhermesp | hum i see... yeah but i guess taking the guidelines , i'd vote to remove that line from index :P | 13:26 |
| noonedeadpunk | any suggestions where to place it's content? | 13:29 |
| noonedeadpunk | maybe we can swap them in the list and it will look more logically | 13:30 |
| noonedeadpunk | like place contributing.rst the first one. Like - So you want to contribute, then onboarding, bug handling, etc | 13:31 |
| noonedeadpunk | eventually we have too much contrubutor docs already for this template | 13:32 |
| noonedeadpunk | I'd rather do not inlcude that new contributing.rst into the tree lol | 13:34 |
| noonedeadpunk | but then error raises so not an option | 13:34 |
| guilhermesp | oh noonedeadpunk i see your point... | 13:37 |
| guilhermesp | im not seeing the benefits of swapping them into the list.. and we do need contributing as well, i'd say for us to keep as is, unless someone else has a good idea about it | 13:40 |
| noonedeadpunk | I don't ;( | 13:41 |
| *** KeithMnemonic has joined #openstack-ansible | 13:43 | |
| guilhermesp | but yeah, in general the patch looks good, do we want more eyes on it or let's just start merging? | 13:44 |
| noonedeadpunk | maybe jrosser have some thoughs | 13:45 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_trove master: Replace usage of deprecated os_*_facts ansible modules https://review.opendev.org/727939 | 13:46 |
| jrosser | hello | 13:46 |
| * jrosser unmeetings | 13:46 | |
| noonedeadpunk | guilhermesp: as once this is merged we have https://review.opendev.org/#/q/topic:project-ptl-and-contrib-docs+(status:open)+owner:noonedeadpunk@ya.ru for review :p | 13:47 |
| jrosser | chandankumar: i think this has set the tunnel id to 1, so that could be overridden becasue I guess that in the tripleo deploy something else has already used tunnel_id=1? | 13:49 |
| jrosser | chandankumar: https://github.com/openstack/openstack-ansible-os_tempest/blob/master/defaults/main.yml#L121 | 13:49 |
| jrosser | noonedeadpunk: what do we need to look at? just had 5hrs meetings so not sure which way is up right now | 13:50 |
| guilhermesp | noonedeadpunk: yep sure, just ping me if i dont see here | 13:50 |
| noonedeadpunk | jrosser: at the moment we were discussing https://review.opendev.org/#/c/726228/ and how ugly having contributing.rst and contribute.rst | 13:51 |
| noonedeadpunk | second thing I'll ask later:) | 13:51 |
| mgariepy | morning everyone | 13:52 |
| mgariepy | qq, centos8 stuff, does it need to be backported to train ? | 13:52 |
| mgariepy | or only u would be good enough ? | 13:53 |
| noonedeadpunk | mgariepy: it needs :((( | 13:53 |
| mgariepy | hmm, forget that. it's not much more work.. as we need a release that is able to do both.. | 13:53 |
| noonedeadpunk | as only train has 7 and 8 rdo packages | 13:53 |
| noonedeadpunk | so this will be kinda migratrion release | 13:54 |
| mgariepy | for lxc part, there is a repo for lxc3 on centos7 and 8. | 13:55 |
| mgariepy | lxc2 is only for centos6 and 7. | 13:55 |
| noonedeadpunk | so for 8 we should use lxc3. for centos7 not sure it matters | 13:56 |
| mgariepy | i guess we don;t need to bother with migration of lxc2>lxc3 since we can stick with lxc2. for 7. | 13:56 |
| noonedeadpunk | we can probably leave it as is to have less work on our hands | 13:56 |
| mgariepy | centos need full re-install afaik. | 13:56 |
| jrosser | noonedeadpunk: contribute and contributing is ugly to have both | 13:56 |
| noonedeadpunk | jrosser: totally agree but out of thoughts.... | 13:56 |
| noonedeadpunk | Like including all of our contribute docs into the one contributing.rst would be as ugly as having both | 13:57 |
| jrosser | is this becasue the community goal mandates a particular filename? i saw this right at the start and thought it was unfortunate | 13:57 |
| noonedeadpunk | Yep, they require specific name and specific template to be used | 13:57 |
| mgariepy | master is U right now ? | 13:58 |
| noonedeadpunk | `There exists project specific information about contribution to that project in each project’s documentation in the form of a doc/source/contributor/contributing.rst following the template in the cookiecutter repo [3].` | 13:58 |
| noonedeadpunk | mgariepy: yep | 13:58 |
| mgariepy | for how long ? | 13:58 |
| mgariepy | do we have an eta ? | 13:58 |
| jrosser | noonedeadpunk: i guess it depends on the intent - like if no projects have contributor guidelines then adding them is cool, just that we happened to have pretty good ones just in a filename a chouple of chars different | 13:58 |
| noonedeadpunk | want to branch asap tbh | 13:59 |
| mgariepy | ok | 13:59 |
| noonedeadpunk | But it won't be too much hard to backport stuff to U | 14:02 |
| mgariepy | yep indeed. | 14:02 |
| mgariepy | does it worth it to put the effort on centos ? | 14:03 |
| mgariepy | is there a lot of ppl using centos that we know of ? | 14:06 |
| noonedeadpunk | Yeah | 14:06 |
| noonedeadpunk | There're pretty much ppl on centos | 14:06 |
| mgariepy | k | 14:08 |
| akahat | jrosser, o/ | 14:08 |
| jrosser | akahat: hello | 14:08 |
| akahat | jrosser, need to talk about this review: https://review.opendev.org/#/c/720705/ | 14:09 |
| akahat | jrosser, i'm still unclear what exactly breaking. | 14:09 |
| noonedeadpunk | openstack-ansible-os_ironic role | 14:09 |
| noonedeadpunk | akahat: https://review.opendev.org/#/c/727067/ | 14:10 |
| akahat | noonedeadpunk, okay. | 14:11 |
| akahat | johanssone, noonedeadpunk suggestions? | 14:11 |
| akahat | johanssone, please ignore :) | 14:13 |
| akahat | jrosser, ^ | 14:13 |
| jrosser | akahat: you saw that i made a depends-on patch for the openstack-ansible ironic tests? https://review.opendev.org/#/c/727067/ | 14:13 |
| johanssone | akahat: Kind of figured :). No worries. | 14:14 |
| akahat | jrosser, yes. I saw it. And it's network conflicts. | 14:14 |
| akahat | :) | 14:14 |
| jrosser | i did this to demonstrate that you patch makes a bunch of assumptions that are valid for the tripleo deployment, but not for the openstack-ansible one | 14:14 |
| jrosser | so as i was saying to chandankumar in out meeting on tuesday we should take care to only include things in os_tempest that absolutely need to be there | 14:15 |
| jrosser | just one example, your test is assuming OVS, ours is assuming linuxbridge, so we must abstract that out as a minimum | 14:16 |
| jrosser | and i think that there is a more wide question for me, is it right that the tempest role is responsible for setting up the environment for ironic? i a a bit uncomfortable with that becasue it feels like part of the deployment, not part of the test runner | 14:17 |
| akahat | jrosser, got your point. So what you suggest.. what exactly i do here? | 14:22 |
| jrosser | in the openstack-ansible case specific resources like images and networks for special stuff like trove, octavia and ironic are internal to the deployment of those components, not in the tempest role | 14:25 |
| jrosser | becaue the cloud operator would need the ironic provisioning network regardless of running tempest or not | 14:25 |
| jrosser | but unfortunately i have no idea how these things are handled in tripleo to be able to comment on how things work there | 14:26 |
| akahat | jrosser, okay. thanks for the clarity. :) | 14:31 |
| noonedeadpunk | also I think that we need make os_tempest to be able to run against not only CI but also prod deployments, like in cases when ppl want just to have tempest to check their deployemnt integrity | 14:32 |
| akahat | that makes seance. | 14:34 |
| noonedeadpunk | jrosser: we need to come up with smth regarding https://review.opendev.org/#/c/726228/ - like I'm very open to advices | 14:46 |
| jrosser | :( | 14:47 |
| jrosser | noonedeadpunk: is it all the patches you are concerned about, or that one specifically? | 14:50 |
| noonedeadpunk | like the rest of that topic ore okeyinsh probably - just having link to contributing.rst. | 14:50 |
| jrosser | i just looked at os_glance and that seems like a good tidy up | 14:51 |
| jrosser | point to the integrated repo contributor guide and delete a load of stuff | 14:51 |
| noonedeadpunk | jrosser: yeah, they're probably ok. Maybe to minimalistinc, but that's tempalte:) | 14:52 |
| noonedeadpunk | also wanted to talks about https://review.opendev.org/#/c/728052/ | 14:52 |
| noonedeadpunk | Like I've just started workig and decided to talk about approach | 14:53 |
| jrosser | so i just looked at the docs preview https://6fff890ed033289a8522-64990ae6344a6a389f4859507c11a876.ssl.cf2.rackcdn.com/726228/5/check/openstack-tox-docs/eddfb53/docs/ | 14:53 |
| jrosser | where even is the contributor guide :) ? | 14:53 |
| noonedeadpunk | alternative ways to do this are here https://opendev.org/openstack/openstack-ansible/src/branch/master/playbooks/containers-nspawn-create.yml#L16-L18 or this https://opendev.org/openstack/openstack-ansible/src/branch/master/playbooks/ceph-rgw-install.yml#L28-L37\ | 14:53 |
| openstackgerrit | Marc Gariépy (mgariepy) proposed openstack/openstack-ansible-lxc_hosts master: [WIP] centos-8 support https://review.opendev.org/728108 | 14:54 |
| noonedeadpunk | jrosser: oh... I've placed it in Developer Documentation | 14:54 |
| noonedeadpunk | really, we can probably move it to upper level? | 14:54 |
| jrosser | well that is actually developer stuff | 14:54 |
| jrosser | but we could have that other new page, the more generic one at the top level | 14:54 |
| *** evrardjp has quit IRC | 14:55 | |
| jrosser | this one https://review.opendev.org/#/c/726228/5/doc/source/contributor/contributing.rst | 14:55 |
| noonedeadpunk | yeah, nice idea | 14:55 |
| jrosser | then i think that makes some sense | 14:55 |
| noonedeadpunk | ok, cool! | 14:55 |
| *** fresta has quit IRC | 14:56 | |
| jrosser | it should appear in the clue box top left with "user guide" "operations guide" ... | 14:56 |
| jrosser | *blue box | 14:56 |
| *** evrardjp has joined #openstack-ansible | 14:56 | |
| openstackgerrit | Duncan Martin Walker proposed openstack/openstack-ansible-ops master: Enable plugin install and removal from mirrors https://review.opendev.org/728110 | 14:59 |
| *** cp- has quit IRC | 15:02 | |
| *** gyee has joined #openstack-ansible | 15:04 | |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible master: [ussuri][goal] Update contributor documentation https://review.opendev.org/726228 | 15:05 |
| mgariepy | dmsimard, do you have a workaround for this ? 2020-05-14 15:01:38.636254 | centos-8 | ERROR: ara 0.16.7 has requirement SQLAlchemy<1.3.0, but you'll have sqlalchemy 1.3.16 which is incompatible. | 15:05 |
| mgariepy | from : https://storage.gra.cloud.ovh.net/v1/AUTH_dcaab5e32b234d56b626f72581e3644c/zuul_opendev_logs_8b1/728108/1/check/openstack-ansible-functional-centos-8/8b15aec/job-output.txt | 15:06 |
| mgariepy | do we need a old version of ara ? | 15:06 |
| dmsimard | mgariepy: it could probably be unpinned ? https://github.com/ansible-community/ara/blob/acf7769f00de59c2d388409af62ccaa030b7bfc5/requirements.txt#L5-L7 | 15:06 |
| mgariepy | noonedeadpunk, jrosser ? | 15:07 |
| dmsimard | I haven't tested 0.16 on centos-8 tbh | 15:07 |
| *** cp- has joined #openstack-ansible | 15:07 | |
| mgariepy | why not install a newer release ? | 15:08 |
| noonedeadpunk | thats tough question lol | 15:08 |
| jrosser | mgariepy: noonedeadpunk and I hold great affection for the old version :) | 15:08 |
| noonedeadpunk | we can probably go on if centos 8 requires it | 15:08 |
| noonedeadpunk | but yes) | 15:08 |
| jrosser | dmsimard: sorry about liking the old version so much :)) | 15:08 |
| noonedeadpunk | I still like old one more lol | 15:08 |
| mgariepy | lol | 15:09 |
| *** fresta has joined #openstack-ansible | 15:09 | |
| dmsimard | mgariepy: it's been tested in various ways like https://review.opendev.org/#/c/682587/ , https://review.opendev.org/#/c/696634/ and https://review.opendev.org/#/c/682962/ | 15:09 |
| dmsimard | sorry about the new UI :( lol | 15:09 |
| jrosser | it's no criticism btw - i use it all the time and it's awesome | 15:09 |
| dmsimard | jrosser: no offense taken | 15:10 |
| noonedeadpunk | actually last version I saw was much better - at least it was usable | 15:10 |
| noonedeadpunk | Like having playbooks names, having reverse sorting | 15:10 |
| *** fresta has quit IRC | 15:11 | |
| dmsimard | https://api.demo.recordsansible.org/?name=&path=openstack-ansible&label= | 15:11 |
| dmsimard | there's no search in the static html version though | 15:11 |
| noonedeadpunk | yeah:( | 15:12 |
| *** fresta has joined #openstack-ansible | 15:12 | |
| noonedeadpunk | Like I've missed what forced you to move to static version? | 15:12 |
| dmsimard | when openstack-infra moved logs to swift | 15:13 |
| dmsimard | we lost the ability to dynamically render ansible.sqlite files off of a central log server | 15:13 |
| noonedeadpunk | oh, sorry, I meant more about design change between <1 and after 1 | 15:14 |
| dmsimard | two things, 1.x was re-written from scratch in django (0.x was flask) and 1.x doesn't require additional javascript dependencies | 15:15 |
| dmsimard | installing 0.x pulled in like 50 dependencies, 1.x pulls in like 5 :) | 15:15 |
| dmsimard | there's ara-web for javascript persons but it's still very WIP and doesn't get much contributions | 15:15 |
| dmsimard | https://web.demo.recordsansible.org/ (it's not very pretty) | 15:16 |
| dmsimard | I am /not/ a javascript person | 15:17 |
| noonedeadpunk | yeah, I can totally get your point | 15:18 |
| noonedeadpunk | I can hardly do anything when it comes to UI as well... | 15:18 |
| jrosser | noonedeadpunk: docs look good - slight funny formatting with the link to lauchpad https://6ea4468309ffe2c3ce04-f26b85fb61fdc96fdf90ad547deeb51e.ssl.cf5.rackcdn.com/726228/6/check/openstack-tox-docs/19b4426/docs/contributor/contributing.html | 15:27 |
| jrosser | but other than that i think we should call that done | 15:28 |
| jrosser | then you want to talk about https://review.opendev.org/#/c/728052/ ? | 15:29 |
| noonedeadpunk | eventually they have it set as a list as well https://opendev.org/openstack/cookiecutter/raw/branch/master/%7b%7bcookiecutter.repo_name%7d%7d/doc/source/contributor/contributing.rst | 15:29 |
| noonedeadpunk | jrosser: yep | 15:30 |
| jrosser | ok i think we call the docs done - too many hurting brain cells | 15:31 |
| *** spatel has joined #openstack-ansible | 15:31 | |
| noonedeadpunk | so regarding https://review.opendev.org/#/c/728052/ | 15:31 |
| noonedeadpunk | there are several ways to implement that | 15:31 |
| noonedeadpunk | as it takes literally each playbook we have I think I need to discuss it first | 15:32 |
| noonedeadpunk | before continuing | 15:32 |
| spatel | jamesdenton: are you around? | 15:33 |
| noonedeadpunk | jrosser: I stopped at some point so to check if I don't miss smth while selecting an approach and can continue | 15:34 |
| noonedeadpunk | so option 1 - use 'always' tag for the play. Then facts are gathered and role is run with it's own tag (if provided). The downside that this will include pre-steps and post steps also "always" I think | 15:36 |
| noonedeadpunk | maybe I should to mark all pre steps with role tag as well... | 15:36 |
| noonedeadpunk | second option we have here https://opendev.org/openstack/openstack-ansible/src/branch/master/playbooks/containers-nspawn-create.yml#L16-L18 | 15:37 |
| noonedeadpunk | so we do extra play which aims only to gather facts against hosts we're going run the role | 15:37 |
| *** andrelf has joined #openstack-ansible | 15:37 | |
| noonedeadpunk | and the third one is here https://opendev.org/openstack/openstack-ansible/src/branch/master/playbooks/ceph-rgw-install.yml#L28-L37 which eventually do setup in pre-step (dunno why but not liking that way) | 15:38 |
| noonedeadpunk | but maybe second option is really better.... | 15:39 |
| *** dpaclt has joined #openstack-ansible | 15:45 | |
| *** fresta has quit IRC | 15:52 | |
| *** fresta has joined #openstack-ansible | 15:53 | |
| jrosser | noonedeadpunk: i read the ansible issue there, and that is a big mess | 15:53 |
| noonedeadpunk | it was resolved in another one, but the idea is that ansible fixed "issue" when facts gathering was made even for play with specific tags | 15:54 |
| noonedeadpunk | there's a bucnh of issues submitted regardign this topic | 15:55 |
| jrosser | and then on the question of if pre-tasks should be "always" thats also important, quite often required setup is in the pre_taskss | 15:55 |
| noonedeadpunk | https://github.com/ansible/ansible/issues/57529#issuecomment-500886760 | 15:55 |
| noonedeadpunk | so if we go the way I started then I can just remove always tag from pre-task (as it's already present on upper level) and set like "cinder" tag on tasks without always | 15:57 |
| noonedeadpunk | to preserve behaviour | 15:57 |
| noonedeadpunk | not sure it's corect at the moment | 15:58 |
| *** rpittau is now known as rpittau|afk | 16:01 | |
| jrosser | wht does this do then https://review.opendev.org/#/c/728052/1/playbooks/containers-lxc-host.yml | 16:02 |
| noonedeadpunk | yeah, that does what I was talking about) | 16:03 |
| jrosser | "always gather facts when osa_gather_facts is true" | 16:03 |
| *** gregwork has joined #openstack-ansible | 16:03 | |
| noonedeadpunk | I mean I'm not sure that cureently we don't miss a tag somewhere | 16:04 |
| jrosser | i'm certain | 16:04 |
| jrosser | gshippey looks very worried every time i suggest using --limit when doing upgrades | 16:04 |
| noonedeadpunk | So if I'll do that it will preserve behaviour and solve our issue with facts gathering | 16:04 |
| jrosser | it's similar danger that something may go wrong becasue of misplaced tags | 16:05 |
| noonedeadpunk | limit is ok) but --tag should concern lol | 16:05 |
| jrosser | --limit without localhost or some other host you forgot in the list will go wrong | 16:05 |
| noonedeadpunk | ah, yes, that's true | 16:06 |
| noonedeadpunk | but it's more predictable lol | 16:06 |
| noonedeadpunk | than just willing to update glance configs) | 16:06 |
| jrosser | this looks understandable https://review.opendev.org/#/c/728052/1/playbooks/containers-lxc-host.yml | 16:07 |
| noonedeadpunk | so like https://opendev.org/openstack/openstack-ansible/src/branch/master/playbooks/containers-nspawn-create.yml#L16-L18 seems eathier to implement but looks not so plain | 16:07 |
| jrosser | though are pre-tasks then "always" ? | 16:07 |
| noonedeadpunk | if pre-tasks are without tag then always | 16:08 |
| noonedeadpunk | otherwise - only with the matching tag | 16:08 |
| jrosser | so in that file it is not really useful to have the lxc-hosts tag? | 16:09 |
| jrosser | if we consider rsyslog-client to be legacy and should be deleted then it's on all the tasks | 16:10 |
| jamesdenton | spatel whats up? | 16:10 |
| noonedeadpunk | the thing I found kind of broken is that while using tag lxc_hosts-install https://opendev.org/openstack/openstack-ansible-lxc_hosts/src/branch/master/tasks/main.yml#L59 it won't include any of pre-tasks | 16:10 |
| jrosser | oh well is that using tags as a shortcut instead of calling that playbook only | 16:11 |
| noonedeadpunk | So I was considering tags as running only some specific part of the playbook | 16:11 |
| noonedeadpunk | like we do have <service>-config to adjust configuration and restart services without running python_venv_build, adding rabbitmq, etc | 16:12 |
| spatel | jamesdenton: soon i am going to build 200 compute cloud so thinking to use openvswitch and i have no idea if openvswitch can scale that level | 16:12 |
| jamesdenton | sure it can | 16:13 |
| spatel | I really need to get DPDK because SRIOV isn't fun at all (specially in NUMA ) | 16:13 |
| noonedeadpunk | but if we're talking about lxc-hosts - yes, I think it's just for shortcut | 16:13 |
| jamesdenton | i would refrain from deploying DVR | 16:13 |
| spatel | DVR is very complex :( | 16:13 |
| jamesdenton | DPDK will still be NUMA sensitive, too | 16:13 |
| noonedeadpunk | like you run setup-hosts.yml --tags lxc-hosts and this will launch only containers creation or smth | 16:14 |
| spatel | but DPDK support bonding on host and security group | 16:14 |
| jamesdenton | i would lab that out before doing a 200 node deployment | 16:15 |
| jamesdenton | just to make sure you get a) the performance you're looking for and b) the components you want working (i.e. openvswitch firewall) don't have caveats | 16:15 |
| spatel | jamesdenton: i am only worried about some kind of unknown bugs :( | 16:16 |
| spatel | jamesdenton: i will do labbing before final deployment | 16:17 |
| jamesdenton | well, i can yell you there are some known bugs with openvswitch firewall and dvr deployments today (even if DVR routers are not configured). but, they're being worked on. i'm curious to see how you get on with that configuration (dpdk, bonds, etc). keep us posted | 16:19 |
| *** sshnaidm is now known as sshnaidm|afk | 16:21 | |
| spatel | jamesdenton: sure, i will | 16:22 |
| spatel | In currently most standard deployment are 3 node controller, can i go with 5 node, any issue there (i don't think so but incase anyone disagree with me) | 16:24 |
| jrosser | spatel: if you think you are at the point of needing to scale out the controllers you could consider splitting them into API servers and rabbit/galera/memcached instead of just make 5 of each | 16:26 |
| noonedeadpunk | jrosser: oh... https://review.opendev.org/#/c/728052/1/playbooks/containers-lxc-host.yml is not working for pre tasks | 16:26 |
| noonedeadpunk | so they're still always | 16:27 |
| jrosser | noonedeadpunk: i think whats difficult here is that we have a tags setup that comes from when ansible worked differently | 16:27 |
| jrosser | and it just seems that the rules have changed | 16:27 |
| noonedeadpunk | yep, that is true | 16:28 |
| *** udesale has quit IRC | 16:28 | |
| *** udesale_ has quit IRC | 16:28 | |
| jrosser | i think what people really care about is the 'just do config' | 16:28 |
| *** andrelf has quit IRC | 16:28 | |
| spatel | jrosser: you are saying make 3 node for API and 3 other node for rabbit/mysql/memcached (so total 6 nodes) | 16:28 |
| noonedeadpunk | at least that's what I care :p | 16:28 |
| jrosser | spatel: you can make it however you want :) but you must have a reason to want 5 instead of 3? | 16:29 |
| jrosser | and you've kind of got two choices, just make everything horizontally wider or start separating things away from each other if you're exceeding the capacity of one controller | 16:30 |
| jrosser | depends on your motivation for needing to do it | 16:30 |
| jamesdenton | :thumbsup: | 16:30 |
| spatel | jrosser: my problem is we don't know when company say we need to launch new product and now we need capacity so lets add 500 more compute nodes in cluster that point i don't want to do re-engineering (so want to keep controller plane strong enough) | 16:30 |
| spatel | currently i am running 400 compute nodes with 3 controller with lots of tweaking on timers etc.. to make things happy :( | 16:31 |
| *** evrardjp has quit IRC | 16:33 | |
| *** evrardjp has joined #openstack-ansible | 16:33 | |
| noonedeadpunk | jrosser: ok, thanks for discussion, I understood that I really made a mistake | 16:34 |
| spatel | I think i would go with 6 nodes plan that would be good to keep rabbitMQ outside of API plane. | 16:34 |
| jrosser | noonedeadpunk: i'm not really having very much clarity about what the right answer is unfortunately :( | 16:34 |
| jrosser | but, anything that can make simpler is good - and if we have tags that don't work there is an argument to remove them | 16:35 |
| jrosser | then we can re-add specific things for "config only" when we understand better | 16:35 |
| jrosser | spatel: really it depends on your application and rate of churn, sounds like you have optimised rabbit a lot so you could move that aside, and if you spend a lot of CPU on neutron agents for the control plane you can also move that away too | 16:37 |
| jrosser | *L3 stuff i mean | 16:37 |
| spatel | I am using vlan base provider so i don't have any virtual routers etc.. but i have noticed neutron and rabbitmq are big time CPU eater (mysql really doing nothing in my case) | 16:38 |
| spatel | jrosser: look at my CPU load after tuning rabbitmq/neutron - https://imgur.com/a/grRttjH | 16:40 |
| jrosser | then maybe there is your answer - dedicated nodes for rabbitmq would benefit your workload | 16:40 |
| jrosser | and release CPU time on the controller for other things | 16:40 |
| spatel | ++ | 16:41 |
| spatel | jrosser: any advantage to use metal for rabbitmq instead of LXC ? | 16:48 |
| openstackgerrit | Merged openstack/openstack-ansible master: Increase disk size for manila service VM during tests https://review.opendev.org/727827 | 17:10 |
| noonedeadpunk | what are healthcheck-* playbooks o_O | 17:17 |
| noonedeadpunk | it looks like early implementation of tempest lol | 17:18 |
| jrosser | i think it's tempestless deployment sanity check | 17:20 |
| jrosser | evrardjp did that i think | 17:20 |
| * noonedeadpunk wondering if anybody use them | 17:20 | |
| noonedeadpunk | I see | 17:20 |
| noonedeadpunk | Just the first time I did so close look through all of our playbooks | 17:21 |
| evrardjp | probably no one | 17:23 |
| evrardjp | but that should be removed from the docs :p | 17:24 |
| noonedeadpunk | is it even in docs?:) | 17:24 |
| evrardjp | though it helped me quite a few times for people having problem on this channel: Just run these playbooks, report to me what's the output | 17:24 |
| evrardjp | I think it was | 17:24 |
| noonedeadpunk | I wasn't going to drop them I was just surprised having them lol | 17:24 |
| evrardjp | not sure what the current state though :p | 17:24 |
| evrardjp | how are folks doing in here? | 17:25 |
| evrardjp | is everything okay for you? | 17:25 |
| *** andrelf has joined #openstack-ansible | 17:25 | |
| evrardjp | noonedeadpunk: oh you do whatever you want with them, I don't really have a stake in that game anymore :D | 17:25 |
| * noonedeadpunk totally failed releasing | 17:25 | |
| evrardjp | it's hard to keep up when you have thousand of things to do :) | 17:26 |
| noonedeadpunk | like we should have been branched | 17:26 |
| noonedeadpunk | but generally pretty fine:) | 17:26 |
| noonedeadpunk | exhausted because of all these covid limitations, you know) | 17:26 |
| *** yolanda has quit IRC | 17:28 | |
| openstackgerrit | Merged openstack/openstack-ansible-haproxy_server master: Only restart rsyslog when it is installed https://review.opendev.org/727851 | 17:36 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible master: Resolve issue with undefined facts while running with tags https://review.opendev.org/728052 | 17:53 |
| noonedeadpunk | jrosser: that's what I get from this tag thing ^ | 17:53 |
| noonedeadpunk | guilhermesp: spotz +w for https://review.opendev.org/#/c/726228/ ?:) | 17:56 |
| spotz | noonedeadpunk: done | 17:57 |
| noonedeadpunk | thanks) | 17:57 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/ansible-hardening master: WIP Centos-8 support https://review.opendev.org/709460 | 18:00 |
| guilhermesp | cool col noonedeadpunk looking | 18:07 |
| guilhermesp | good it seems spotz alraedy +w :P | 18:07 |
| noonedeadpunk | yeah:) | 18:07 |
| noonedeadpunk | sorry for disturbing:) | 18:07 |
| spotz | :) Hey when I'm pinged and awake I try to be quick:) | 18:08 |
| guilhermesp | no worries at all. i guess we have then a lot of review for roles | 18:08 |
| noonedeadpunk | Really appreciated :) | 18:08 |
| noonedeadpunk | jrosser: can you recall what we've decided with lsync on centos8? Seems just commenting out for now? https://review.opendev.org/#/c/709532/3/vars/redhat-8.yml | 18:09 |
| *** andrelf has quit IRC | 18:15 | |
| openstackgerrit | Merged openstack/openstack-ansible master: [ussuri][goal] Update contributor documentation https://review.opendev.org/726228 | 18:15 |
| noonedeadpunk | ok, so contributing page is live:) https://docs.openstack.org/openstack-ansible/latest/contributor/contributing.html | 18:21 |
| noonedeadpunk | we can look through https://review.opendev.org/#/q/topic:project-ptl-and-contrib-docs+(status:open) now:) | 18:21 |
| jrosser | noonedeadpunk: I looked at iSync and found a method to use centos 7 epel for just lsync, on centos 8 | 18:29 |
| mgariepy | (╯°□°)╯︵ ┻━┻ el8.. | 18:29 |
| jrosser | argh *lsync not iSync | 18:29 |
| noonedeadpunk | xD | 18:29 |
| * jrosser blames iPhone | 18:29 | |
| noonedeadpunk | I just tested build from soure and it worked as well... | 18:30 |
| mgariepy | of course it work, it's not the reason for removing it ! :P haha | 18:31 |
| jrosser | https://www.liquidweb.com/kb/how-to-sync-two-centos-8-servers-using-file-replication/ | 18:31 |
| noonedeadpunk | AH I read it today:) | 18:31 |
| jrosser | noonedeadpunk: I also built it from source :) | 18:31 |
| jrosser | the build was easy and as the repo server needs a tool chain anyway it’s not so bad | 18:32 |
| noonedeadpunk | using epel7 is too hacky I think... | 18:32 |
| noonedeadpunk | like may break thing later | 18:33 |
| * jrosser didn’t get to look at galera role again | 18:35 | |
| jrosser | today has evaporated in meetings | 18:35 |
| *** dpaclt has quit IRC | 18:36 | |
| openstackgerrit | Chandan Kumar (raukadah) proposed openstack/openstack-ansible master: Enable octavia tempest tests https://review.opendev.org/727949 | 18:40 |
| openstackgerrit | Chandan Kumar (raukadah) proposed openstack/openstack-ansible master: Enable octavia tempest tests https://review.opendev.org/727949 | 18:41 |
| openstackgerrit | Chandan Kumar (raukadah) proposed openstack/openstack-ansible-os_octavia master: [WIP] enable octavia tempest tests https://review.opendev.org/727856 | 18:42 |
| mgariepy | also, no btrfs on centos8. | 19:03 |
| noonedeadpunk | seems it's easiers to name what is present | 19:04 |
| mgariepy | lol | 19:04 |
| mgariepy | the list would be shorter for sure,\ | 19:04 |
| *** NewJorg has quit IRC | 19:04 | |
| mgariepy | not fun at all | 19:05 |
| *** NewJorg has joined #openstack-ansible | 19:10 | |
| openstackgerrit | Marc Gariépy (mgariepy) proposed openstack/openstack-ansible-tests master: [WIP] centos8 https://review.opendev.org/728259 | 19:12 |
| openstackgerrit | Marc Gariépy (mgariepy) proposed openstack/openstack-ansible-lxc_hosts master: [WIP] centos-8 support https://review.opendev.org/728108 | 19:13 |
| openstackgerrit | Marc Gariépy (mgariepy) proposed openstack/openstack-ansible-lxc_hosts master: [WIP] centos-8 support https://review.opendev.org/728108 | 19:14 |
| mgariepy | at the speed i progress i doubt i'll have it done before the end of the year. | 19:21 |
| noonedeadpunk | Actually I think basic source roles should work nicely | 19:21 |
| mgariepy | lol | 19:22 |
| noonedeadpunk | And I'll also try to help a bitwith that | 19:22 |
| mgariepy | you think.. | 19:22 |
| mgariepy | lol | 19:22 |
| noonedeadpunk | Ah, yeah, distro packages.... | 19:22 |
| mgariepy | i'd like to have lxc so can test stuff more quickly. | 19:22 |
| noonedeadpunk | I pray lol | 19:22 |
| mgariepy | in the 12 lines of patches i made today, i had like 4-5 roadblock | 19:23 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-repo_server master: WIP Centos-8 support https://review.opendev.org/709532 | 19:23 |
| noonedeadpunk | that is annoying | 19:23 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible master: [WIP] Centos-8 support https://review.opendev.org/689629 | 19:27 |
| noonedeadpunk | ok, so the last time for metal we were stuck on galera https://zuul.opendev.org/t/openstack/build/897deaaccb324348a6251aaa8a675ff1/log/job-output.txt#12130 | 19:28 |
| openstackgerrit | Marc Gariépy (mgariepy) proposed openstack/openstack-ansible-tests master: [WIP] centos8 https://review.opendev.org/728259 | 19:30 |
| openstackgerrit | Marc Gariépy (mgariepy) proposed openstack/openstack-ansible-galera_server master: WIP Centos-8 support https://review.opendev.org/724721 | 19:51 |
| mgariepy | yep galera was not working. my patch doesn't apply anymore. but the galera-4 stuff was in it. | 19:58 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/ansible-hardening master: Remove ⌘ symbol from docs https://review.opendev.org/728278 | 19:58 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-repo_server master: WIP Centos-8 support https://review.opendev.org/709532 | 20:02 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible master: [WIP] Centos-8 support https://review.opendev.org/689629 | 20:03 |
| openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/openstack-ansible-galera_server master: WIP Centos-8 support https://review.opendev.org/724721 | 20:11 |
| noonedeadpunk | mgariepy: I dropped depends-on for galera patch as it was in conflict because of that | 20:13 |
| noonedeadpunk | like anyway galera bump to 10.4 is known to be broken at the moment because of the client | 20:14 |
| jrosser | noonedeadpunk: i am probably not going to be able to look at the galera role combining till next week | 20:25 |
| jrosser | i think there is just maybe a silly error of some kind in the role | 20:25 |
| jrosser | but the tests are still trying to use galera_client and i didnt have time to take that out | 20:26 |
| jrosser | not stopping using galera_client is the biggest thing to fix next | 20:26 |
| noonedeadpunk | ok, got it. will try to look at it tomorrow | 20:29 |
| jrosser | also mnaser guilhermesp there are a bunch of small patches for deprecations and bug fixes to review if you’ve got a moment | 20:32 |
| guilhermesp | sure thing jrosser i will take a look today | 20:47 |
| *** born2bake has quit IRC | 21:24 | |
| *** spatel has quit IRC | 21:33 | |
| openstackgerrit | Merged openstack/openstack-ansible-tests master: Update ansible to 2.9.9 https://review.opendev.org/727926 | 22:12 |
| *** jbadiapa has quit IRC | 22:25 | |
| *** tosky has quit IRC | 23:05 | |
| *** schwicht has quit IRC | 23:06 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_rally master: Cleanup after repo_build and pip_install retirement https://review.opendev.org/727414 | 23:11 |
| *** schwicht has joined #openstack-ansible | 23:15 | |
| openstackgerrit | Merged openstack/openstack-ansible master: Bump ansible version to 2.9.9 https://review.opendev.org/727925 | 23:53 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!