| *** klamath has quit IRC | 00:07 | |
| *** dxiri_ has joined #openstack-ansible | 00:07 | |
| *** dxiri has quit IRC | 00:07 | |
| *** victorhe has quit IRC | 00:10 | |
| *** victorhe has joined #openstack-ansible | 00:11 | |
| *** jperry has quit IRC | 00:15 | |
| *** spotz is now known as spotz_zzz | 00:28 | |
| *** weezS has quit IRC | 00:32 | |
| *** vnogin has quit IRC | 00:32 | |
| *** sdake has quit IRC | 00:43 | |
| *** dxiri_ has quit IRC | 00:44 | |
| *** sdake has joined #openstack-ansible | 00:45 | |
| *** yolanda has joined #openstack-ansible | 00:56 | |
| *** agrebennikov has quit IRC | 01:00 | |
| *** dxiri has joined #openstack-ansible | 01:01 | |
| *** f13o has joined #openstack-ansible | 01:03 | |
| *** sdake has quit IRC | 01:05 | |
| *** dxiri has quit IRC | 01:06 | |
| *** sdake has joined #openstack-ansible | 01:06 | |
| *** f13o has quit IRC | 01:08 | |
| *** jlockwood has quit IRC | 01:08 | |
| *** sdake has quit IRC | 01:27 | |
| *** sdake has joined #openstack-ansible | 01:29 | |
| *** victorhe has quit IRC | 01:34 | |
| *** woodard has quit IRC | 01:36 | |
| *** woodard has joined #openstack-ansible | 01:36 | |
| *** dxiri has joined #openstack-ansible | 01:38 | |
| *** dxiri has quit IRC | 01:42 | |
| *** havocbmx-LT has quit IRC | 01:42 | |
| *** qiliang27 has joined #openstack-ansible | 01:48 | |
| *** sdake has quit IRC | 01:50 | |
| *** omiday has quit IRC | 01:51 | |
| *** sdake has joined #openstack-ansible | 01:53 | |
| *** sdake has quit IRC | 02:14 | |
| *** sdake has joined #openstack-ansible | 02:15 | |
| *** chhavi has joined #openstack-ansible | 02:15 | |
| *** omiday has joined #openstack-ansible | 02:16 | |
| *** jwitko has quit IRC | 02:19 | |
| *** dxiri has joined #openstack-ansible | 02:23 | |
| *** dxiri has quit IRC | 02:23 | |
| *** dxiri has joined #openstack-ansible | 02:24 | |
| *** dxiri_ has joined #openstack-ansible | 02:26 | |
| *** fguillot has quit IRC | 02:27 | |
| *** dxiri_ has quit IRC | 02:28 | |
| *** dxiri_ has joined #openstack-ansible | 02:29 | |
| *** dxiri has quit IRC | 02:29 | |
| *** jamielennox is now known as jamielennox|away | 02:30 | |
| *** thorst has joined #openstack-ansible | 02:32 | |
| *** thorst has quit IRC | 02:32 | |
| *** thorst has joined #openstack-ansible | 02:33 | |
| *** sdake has quit IRC | 02:36 | |
| *** jwitko has joined #openstack-ansible | 02:37 | |
| *** sdake has joined #openstack-ansible | 02:37 | |
| *** chhavi has quit IRC | 02:39 | |
| *** thorst has quit IRC | 02:40 | |
| *** thorst has joined #openstack-ansible | 02:41 | |
| *** thorst has quit IRC | 02:45 | |
| *** asettle has joined #openstack-ansible | 02:45 | |
| *** asettle has quit IRC | 02:50 | |
| *** chhavi has joined #openstack-ansible | 02:56 | |
| *** sdake has quit IRC | 02:57 | |
| *** jamielennox|away is now known as jamielennox | 02:59 | |
| *** sdake has joined #openstack-ansible | 02:59 | |
| *** dxiri has joined #openstack-ansible | 03:08 | |
| *** thorst has joined #openstack-ansible | 03:09 | |
| *** thorst has quit IRC | 03:10 | |
| *** dxiri_ has quit IRC | 03:12 | |
| *** Mudpuppy_ has quit IRC | 03:13 | |
| *** javeriak has joined #openstack-ansible | 03:17 | |
| *** sdake has quit IRC | 03:19 | |
| *** sdake has joined #openstack-ansible | 03:21 | |
| *** kylek3h_ has quit IRC | 03:24 | |
| *** kylek3h has joined #openstack-ansible | 03:25 | |
| *** omiday has quit IRC | 03:26 | |
| *** kylek3h has quit IRC | 03:29 | |
| *** thorst has joined #openstack-ansible | 03:36 | |
| *** thorst has quit IRC | 03:37 | |
| openstackgerrit | Chhavi Agarwal proposed openstack/openstack-ansible-os_trove: Trove service user configuration should be aligned https://review.openstack.org/401054 | 03:39 |
|---|---|---|
| *** sdake has quit IRC | 03:41 | |
| *** sdake has joined #openstack-ansible | 03:43 | |
| *** Mudpuppy has joined #openstack-ansible | 03:56 | |
| *** chhavi has quit IRC | 04:00 | |
| *** hw_wutianwei has joined #openstack-ansible | 04:02 | |
| *** markvoelker has quit IRC | 04:02 | |
| *** sdake has quit IRC | 04:03 | |
| *** sdake has joined #openstack-ansible | 04:05 | |
| openstackgerrit | Merged openstack/openstack-ansible-repo_build: Remove join filter from pre-build pip module tasks https://review.openstack.org/400360 | 04:06 |
| openstackgerrit | Merged openstack/openstack-ansible-os_nova: Remove join filter from optional pip module tasks https://review.openstack.org/400356 | 04:14 |
| *** sdake has quit IRC | 04:25 | |
| *** kylek3h has joined #openstack-ansible | 04:25 | |
| *** sdake has joined #openstack-ansible | 04:28 | |
| *** kylek3h has quit IRC | 04:30 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_neutron: Remove join filter from optional pip module tasks https://review.openstack.org/400346 | 04:35 |
| *** mhayden has quit IRC | 04:38 | |
| openstackgerrit | Merged openstack/openstack-ansible: Remove join filter from pip module tasks https://review.openstack.org/400362 | 04:44 |
| *** sdake has quit IRC | 04:49 | |
| *** dxiri has quit IRC | 04:49 | |
| *** mhayden has joined #openstack-ansible | 04:50 | |
| *** jwitko has quit IRC | 04:51 | |
| *** sdake has joined #openstack-ansible | 04:53 | |
| *** havocbmx-LT has joined #openstack-ansible | 04:53 | |
| *** havocbmx-LT has quit IRC | 04:54 | |
| *** jwitko has joined #openstack-ansible | 04:59 | |
| *** dxiri has joined #openstack-ansible | 05:01 | |
| *** poopcat has quit IRC | 05:03 | |
| *** victorhe has joined #openstack-ansible | 05:05 | |
| *** dxiri has quit IRC | 05:05 | |
| *** Jeffrey4l has quit IRC | 05:06 | |
| *** victorhe has quit IRC | 05:06 | |
| *** victorhe has joined #openstack-ansible | 05:07 | |
| *** victorhe has quit IRC | 05:16 | |
| *** javeriak has quit IRC | 05:21 | |
| *** chhavi has joined #openstack-ansible | 05:24 | |
| *** kylek3h has joined #openstack-ansible | 05:26 | |
| *** kylek3h has quit IRC | 05:31 | |
| *** sdake_ has joined #openstack-ansible | 05:31 | |
| *** sdake has quit IRC | 05:35 | |
| *** victorhe has joined #openstack-ansible | 05:37 | |
| *** javeriak has joined #openstack-ansible | 05:38 | |
| *** victorhe has quit IRC | 05:42 | |
| *** Jack_Iv has joined #openstack-ansible | 05:43 | |
| *** hybridpollo has quit IRC | 05:43 | |
| *** cathrichardson has quit IRC | 05:49 | |
| *** cathrichardson has joined #openstack-ansible | 05:50 | |
| *** jwitko has quit IRC | 06:02 | |
| *** markvoelker has joined #openstack-ansible | 06:02 | |
| *** qiliang27 has quit IRC | 06:04 | |
| *** markvoelker has quit IRC | 06:08 | |
| *** michaelgugino has quit IRC | 06:23 | |
| *** janki has joined #openstack-ansible | 06:26 | |
| *** kylek3h has joined #openstack-ansible | 06:27 | |
| *** kylek3h has quit IRC | 06:31 | |
| *** jwitko has joined #openstack-ansible | 06:34 | |
| openstackgerrit | Chhavi Agarwal proposed openstack/openstack-ansible-os_trove: Trove service user configuration should be aligned https://review.openstack.org/401054 | 06:36 |
| *** javeriak has quit IRC | 06:40 | |
| *** oanson has joined #openstack-ansible | 06:43 | |
| openstackgerrit | Steve Lewis (stevelle) proposed openstack/openstack-ansible: Inventory refactor SystemExit uses removed https://review.openstack.org/400453 | 06:44 |
| *** Jack_Iv has quit IRC | 06:52 | |
| *** Jack_Iv has joined #openstack-ansible | 06:53 | |
| *** Mudpuppy has quit IRC | 06:56 | |
| *** h5t4 has joined #openstack-ansible | 07:06 | |
| *** Mudpuppy has joined #openstack-ansible | 07:27 | |
| *** kylek3h has joined #openstack-ansible | 07:28 | |
| *** gouthamr has joined #openstack-ansible | 07:30 | |
| *** Mudpuppy has quit IRC | 07:32 | |
| *** kylek3h has quit IRC | 07:32 | |
| *** tshtilma has joined #openstack-ansible | 07:34 | |
| *** survietamine has joined #openstack-ansible | 07:42 | |
| *** drifterza has joined #openstack-ansible | 07:44 | |
| *** chhavi has quit IRC | 07:46 | |
| *** victorhe has joined #openstack-ansible | 07:58 | |
| *** victorhe has quit IRC | 08:02 | |
| *** berendt has joined #openstack-ansible | 08:12 | |
| *** derekjhyang has quit IRC | 08:21 | |
| *** kylek3h has joined #openstack-ansible | 08:28 | |
| openstackgerrit | Steve Lewis (stevelle) proposed openstack/openstack-ansible: Inventory refactor SystemExit uses removed https://review.openstack.org/400453 | 08:33 |
| *** kylek3h has quit IRC | 08:33 | |
| *** victorhe has joined #openstack-ansible | 08:34 | |
| *** victorhe has quit IRC | 08:38 | |
| openstackgerrit | Jimmy McCrory proposed openstack/openstack-ansible-galera_server: [WIP] Install MariaDB 10.1 server https://review.openstack.org/401123 | 08:43 |
| openstackgerrit | Omer Anson proposed openstack/openstack-ansible-os_neutron: [WIP] Dragonflow https://review.openstack.org/391524 | 08:43 |
| neith | evrardjp: awesome talk yesterday! | 08:45 |
| prometheanfire | evrardjp: you gave a talk? | 08:48 |
| *** oneswig has joined #openstack-ansible | 08:57 | |
| *** karimb has joined #openstack-ansible | 09:00 | |
| *** gouthamr has quit IRC | 09:04 | |
| *** asettle has joined #openstack-ansible | 09:05 | |
| openstackgerrit | Jimmy McCrory proposed openstack/openstack-ansible-galera_server: [WIP] Install MariaDB 10.1 server https://review.openstack.org/401123 | 09:07 |
| *** chhavi has joined #openstack-ansible | 09:16 | |
| *** gtrxcb has joined #openstack-ansible | 09:28 | |
| evrardjp | prometheanfire: yes | 09:29 |
| *** kylek3h has joined #openstack-ansible | 09:29 | |
| *** rpittau has joined #openstack-ansible | 09:29 | |
| *** markvoelker has joined #openstack-ansible | 09:32 | |
| *** karimb has quit IRC | 09:33 | |
| *** kylek3h has quit IRC | 09:34 | |
| *** tshtilma has quit IRC | 09:34 | |
| *** tshtilma has joined #openstack-ansible | 09:34 | |
| *** markvoelker has quit IRC | 09:37 | |
| *** karimb has joined #openstack-ansible | 09:37 | |
| andymccr | ok! i think my job to remove trusty gating from master is in :D (whilst making xenial voting on master, but non-voting on newton!) | 09:45 |
| drifterza | Hello all | 09:45 |
| *** haad1 has joined #openstack-ansible | 09:47 | |
| evrardjp | andymccr: onwards! | 09:49 |
| andymccr | evrardjp: i dont think it worked looking at the gates running :P | 09:50 |
| andymccr | might need help figuring out why hahaha | 09:50 |
| evrardjp | finishing my bug answer, and then I'll help you | 09:50 |
| andymccr | thanks! i think i got the nv vs voting incorrect | 09:50 |
| andymccr | hmm nope that looks ok | 09:52 |
| drifterza | evrardjp, how was your talk yesterday? | 09:53 |
| *** victorhe has joined #openstack-ansible | 09:54 | |
| *** gtrxcb has quit IRC | 09:55 | |
| evrardjp | It was really nice! I thank Rackspace for sending me, and the OpenStack Day France for having me. It was my first one, so a little stressed out before... I hope it didn't impact too much the presentation. | 09:56 |
| evrardjp | I would have wished to have more time for questions, but I got questions afterwards, so it's fine. I did a full house, with ppl standing in the room, so that feels good too. | 09:57 |
| evrardjp | I hope I can do this again :) | 09:57 |
| neith | evrardjp: I confirm, It was cool | 09:57 |
| neith | evrardjp: using xkcd's was a good idea | 09:57 |
| *** karimb has quit IRC | 09:58 | |
| evrardjp | well XKCD is good source of real life things, and many applied :p | 09:58 |
| *** karimb has joined #openstack-ansible | 10:01 | |
| *** victorhe has quit IRC | 10:03 | |
| *** vnogin has joined #openstack-ansible | 10:03 | |
| evrardjp | andymccr: do you think we should have ^gate-openstack-ansible-openstack-ansible-aio-ubuntu-xenial$ instead? | 10:09 |
| evrardjp | https://review.openstack.org/#/c/400197/5/zuul/layout.yaml line 3287 | 10:09 |
| andymccr | ahh yeah maybe | 10:10 |
| andymccr | and the one below | 10:10 |
| andymccr | that could be it. | 10:10 |
| evrardjp | I don't know why in the branch we have the ? before negating too | 10:10 |
| evrardjp | but this seems utterly complex | 10:10 |
| evrardjp | we have still kilo branches etc? | 10:11 |
| evrardjp | let me check | 10:11 |
| andymccr | evrardjp: nope | 10:11 |
| andymccr | not anymore | 10:11 |
| evrardjp | ok | 10:11 |
| andymccr | infact liberty is going away soon | 10:11 |
| andymccr | haha and yeah its complex | 10:11 |
| *** karimb has quit IRC | 10:11 | |
| evrardjp | so it makes sense to me to make it simple, and not negate the list | 10:11 |
| andymccr | let me go edit 2 15k + line files :D | 10:12 |
| evrardjp | xenial-nv$ for ^(liberty|stable/(mitaka|newton)) and xenial for branch:^(master) | 10:12 |
| andymccr | i think you're right though | 10:12 |
| andymccr | without the $ it'll not work properly | 10:12 |
| evrardjp | I suppose they gonna catch on both :p | 10:13 |
| andymccr | yeah exactly | 10:13 |
| evrardjp | Well it's just a way to make sure it behaves cleanly | 10:13 |
| evrardjp | and for the branch master, it's easy, at each big release we have to go to this file anyway, so why making everything complex :p | 10:13 |
| andymccr | there is also no real way to test it before hand | 10:14 |
| evrardjp | that's the bad part | 10:14 |
| *** electrofelix has joined #openstack-ansible | 10:14 | |
| evrardjp | but it slipped through the review | 10:14 |
| andymccr | yeah | 10:14 |
| evrardjp | do you know if there is at least a way to test the regexp? | 10:14 |
| andymccr | odyssey4me is good at picking up those kinda things, but it merged before you or him could review :) | 10:15 |
| evrardjp | I'm not sure how this was impletemented | 10:15 |
| evrardjp | ahah | 10:15 |
| andymccr | evrardjp: honestly im learning by looking at what already exists... | 10:15 |
| drifterza | thats the best way to learn | 10:15 |
| drifterza | IMO | 10:15 |
| evrardjp | agreed | 10:15 |
| andymccr | haha drifterza yeah agree | 10:18 |
| evrardjp | seems ?! is standard here | 10:19 |
| evrardjp | and we use it ourselves | 10:20 |
| *** jamielennox is now known as jamielennox|away | 10:21 | |
| andymccr | i think the idea is to specify the branches that already exist and wont change - so it'll apply to the ones that arent going to change - that way you dont have to change it next cycle when stable/ocata now exists | 10:22 |
| *** haad1 has quit IRC | 10:25 | |
| andymccr | hmm evrardjp https://review.openstack.org/#/c/391524/ not sure why the trusty jobs are running there - thoughts? the main func-ubuntu-trusty job got skipped correctly. | 10:25 |
| evrardjp | I'm testing with regexr | 10:26 |
| evrardjp | isn't that the problem when some jobs are already exist before the change in the infra? | 10:27 |
| evrardjp | jobs/reviews? | 10:27 |
| evrardjp | anyway I'm reviewing the full regexpes | 10:27 |
| evrardjp | I'll give you feedback | 10:27 |
| evrardjp | you have another patch running? | 10:28 |
| evrardjp | or should I? | 10:28 |
| andymccr | evrardjp: you can if you want, i have one ready i can mark it wip for feedback so it doesnt merge before we're happy | 10:28 |
| andymccr | but you could be right | 10:28 |
| andymccr | there arent many new patches to confirm | 10:28 |
| *** haad1 has joined #openstack-ansible | 10:29 | |
| andymccr | perhaps i should make some test ones to check that all that is missing is the $ which i think might fix it | 10:29 |
| evrardjp | you're gonna have to test after the changed merged | 10:29 |
| evrardjp | /merged/merges | 10:29 |
| evrardjp | let me review your patch | 10:29 |
| evrardjp | ahah there are other things to improve on that layout.yaml | 10:31 |
| andymccr | https://review.openstack.org/401184 <- that one | 10:31 |
| oanson | Hi, I rebased the Dragonflow patch ( https://review.openstack.org/#/c/391524/ ) on top of latest master with the updated tempest method patch ( https://review.openstack.org/#/c/398923/ ). | 10:32 |
| oanson | Do I need to add something to have test-requirements.txt installed? | 10:32 |
| oanson | Does the name of the tempest plugin have to be 'tempest-dragonflow', or is 'dragonflow' enough? | 10:33 |
| evrardjp | andymccr: reviewing it right now | 10:34 |
| andymccr | evrardjp: thanks! | 10:34 |
| evrardjp | oanson: I can't help you on this, I'm a tempest noob | 10:34 |
| andymccr | drifterza: do you know about the tempest plugin bits? | 10:34 |
| oanson | evrardjp, I see. Thanks. | 10:35 |
| drifterza | generally speaking its just the name of repo | 10:35 |
| drifterza | so IE: 'dragonflow' | 10:35 |
| evrardjp | oanson: sorry. I can try to help you later 'though | 10:36 |
| oanson | drifterza, I see in my local test environment that the 'Running Dragonflow fullstack' test fails since subunit isn't installed, although it does appear in Dragonflow's test-requirements.txt. | 10:36 |
| oanson | I understood that adding dragonflow to the tempest-plugins should solve the issue. | 10:37 |
| drifterza | wheres the pr ? | 10:37 |
| evrardjp | andymccr: I made a terrible mistake. I opened the full layout.yaml file in my browser. | 10:37 |
| oanson | https://review.openstack.org/#/c/391524/33/tests/neutron-overrides-dragonflow.yml | 10:37 |
| *** tshtilma has quit IRC | 10:37 | |
| andymccr | evrardjp: hahahahahahahahaa | 10:37 |
| andymccr | yes. | 10:37 |
| *** zhangjn has joined #openstack-ansible | 10:40 | |
| *** zhangjn has quit IRC | 10:40 | |
| *** zhangjn has joined #openstack-ansible | 10:41 | |
| *** zhangjn has quit IRC | 10:41 | |
| *** yolanda has quit IRC | 10:41 | |
| drifterza | oanson, I'm reviewing | 10:41 |
| drifterza | give me a few | 10:41 |
| oanson | Thanks | 10:41 |
| *** zhangjn has joined #openstack-ansible | 10:41 | |
| *** zhangjn has quit IRC | 10:41 | |
| *** zhangjn has joined #openstack-ansible | 10:43 | |
| *** zhangjn has quit IRC | 10:43 | |
| *** zhangjn has joined #openstack-ansible | 10:44 | |
| *** zhangjn has quit IRC | 10:44 | |
| *** zhangjn has joined #openstack-ansible | 10:44 | |
| *** zhangjn has quit IRC | 10:44 | |
| *** mmoster_ has quit IRC | 10:44 | |
| vnogin | morning guys :) | 10:45 |
| *** stuartgr has joined #openstack-ansible | 10:47 | |
| drifterza | oanson, you can add test-vars type things in the overrides | 10:48 |
| drifterza | I replied to your PR | 10:48 |
| oanson | I saw your reply | 10:49 |
| oanson | I didn't understand the test-vars point? | 10:49 |
| drifterza | so before the new method you would have vars in tests/test-vars.yml right | 10:49 |
| oanson | Let me check. I don't remember :) | 10:50 |
| drifterza | umn let me get you an example | 10:50 |
| drifterza | one sec | 10:50 |
| oanson | But I don't see it. Maybe I should have added it and didn't? | 10:50 |
| drifterza | as an example: https://github.com/openstack/openstack-ansible-os_trove/blob/master/tests/os_trove-overrides.yml#L19 | 10:51 |
| drifterza | if you require things that are not in your defaults/main.yml | 10:51 |
| drifterza | or want to override things | 10:51 |
| drifterza | I don't think you do overrides things, but its just so you know where to put them | 10:52 |
| drifterza | I hope I'm making sense | 10:53 |
| oanson | One sec. I'm trying to keep up :) | 10:53 |
| evrardjp | andymccr: I think there is also the fact you're giving the aio, I have to fully review the full list of rules :p | 10:53 |
| oanson | drifterza, I think I follow so far | 10:54 |
| andymccr | evrardjp: i think the AIO is correct though - because i want ALL trusty jobs to be ignored in master - but before master, and then i want to promote the xenial AIO to voting for master,but NOT for newton | 10:54 |
| oanson | This is in case we want to override variables/parameters/defaults in the tox/testing environment | 10:54 |
| andymccr | the other xenial jobs are all voting already | 10:54 |
| andymccr | (correctly) | 10:54 |
| *** tshtilma has joined #openstack-ansible | 10:55 | |
| andymccr | i guess the alternative is to simplify and just make xenial voting in newton too :P | 10:55 |
| drifterza | well mostly just if you want to override default variables in your tests | 10:55 |
| evrardjp | let me have a full deep check, I want to ensure we don't (with this rule) forget the impact on other openstack-ansible-os_ repos | 10:56 |
| evrardjp | and when I'll have digested all that it's gonna be done for the future anyway | 10:56 |
| oanson | drifterza, I see. | 10:57 |
| oanson | I am not sure how this allows me to install dragonflow's test-requirements.txt ? | 10:57 |
| drifterza | is test-requirements.txt specific to tempest ? | 11:00 |
| andymccr | oanson: test-requierments for dragonflow should get installed by the tempest role | 11:00 |
| drifterza | true andymccr | 11:00 |
| oanson | drifterza, dragonflow | 11:00 |
| oanson | drifterza, dragonflow's test-requirement.txt . We have libraries we only use for testing that we don;t install in production | 11:01 |
| andymccr | https://github.com/openstack/openstack-ansible-os_tempest/blob/master/tasks/tempest_install.yml#L129-L150 | 11:01 |
| oanson | andymccr, that's was my understanding | 11:01 |
| andymccr | oanson: ^ that'll check for the file and install the reqs :) or should at least, so if that isn't happening something is wrong! | 11:01 |
| oanson | The 'Run Dragonflow Fullstack' task fails since it can't find subunit | 11:02 |
| oanson | 1 sec. Let me try something | 11:02 |
| oanson | Opening a new shell from a venv breaks out of the venv? | 11:02 |
| drifterza | just type deactivate in the venv | 11:03 |
| drifterza | unless you want to keep the venv | 11:03 |
| oanson | I want to keep the venv. My test opens a new shell, since I source openrc | 11:03 |
| oanson | But it looks like that's not it | 11:05 |
| *** tshtilma has quit IRC | 11:05 | |
| chhavi | drifterza: Want to discuss on the review comments https://review.openstack.org/#/c/401054/ | 11:05 |
| drifterza | give me 10mins | 11:05 |
| oanson | Maybe the shell from ansible isn't running in the venv? Do I need to activate it manually? | 11:06 |
| *** karimb has joined #openstack-ansible | 11:09 | |
| *** tshtilma has joined #openstack-ansible | 11:10 | |
| *** yolanda has joined #openstack-ansible | 11:11 | |
| *** Jeffrey4l has joined #openstack-ansible | 11:14 | |
| openstackgerrit | Omer Anson proposed openstack/openstack-ansible-os_neutron: [WIP] Dragonflow https://review.openstack.org/391524 | 11:16 |
| chhavi | drifterza: sure | 11:23 |
| neith | guys, I have a small OS cluster now, but on the node with haproxy, I see dozens of process like: haproxy 34956 0.0 0.0 346640 79676 ? Sl 03:15 0:08 (wsgi:keystone-se -k start. Is it normal? | 11:24 |
| neith | there are not user currently using the cluster | 11:24 |
| neith | *no | 11:24 |
| evrardjp | wait | 11:24 |
| evrardjp | I'm not sure to understand what you mean neith | 11:25 |
| evrardjp | it's normal to see the child processes of your containers as root on the bare metal node | 11:26 |
| neith | evrardjp: but there are dozens | 11:26 |
| drifterza | chhavi, im back | 11:26 |
| evrardjp | neith: I'd sayt it depends on many containers you have :p | 11:27 |
| chhavi | ok, there are multiple ways to fix this issue. | 11:27 |
| chhavi | I understand all the openstack project uses the service tenant. But in that case we need the openstack project user to get assigned to the admin role with the service tenant. | 11:27 |
| openstackgerrit | Andy McCrae proposed openstack/openstack-ansible-os_swift: Test new gate rules for scenarios. https://review.openstack.org/401208 | 11:27 |
| evrardjp | I'm on one keystone container that is bored as hell, and it has already 2 keystone-se and 2 keystone-ad | 11:27 |
| neith | evrardjp: there are all the same process: (wsgi:keystone-ad -k start | 11:27 |
| evrardjp | multiply by 3 containers, you have 12 keystone | 11:28 |
| chhavi | In the case of trove, as per the documentation it uses only trove_for_trove_usage as a tenant. | 11:28 |
| neith | I have only one keystone container | 11:28 |
| evrardjp | zombie processes? | 11:29 |
| chhavi | For any project service to work, the configuration [keystone_authtoken] section should be updated with the correct user,project, tenant and password. | 11:29 |
| evrardjp | check on your container see what's inside | 11:29 |
| evrardjp | else, restart the container :p | 11:29 |
| neith | no zombie process | 11:30 |
| chhavi | drifterza: do you agree with me. | 11:30 |
| drifterza | ah okay | 11:33 |
| chhavi | 1 way to solve this is to associate a admin_trove_user with the service tenant, if we really want to use the same service tenant which we are using for all others | 11:34 |
| drifterza | but it goes against their documentation right ? | 11:34 |
| *** kylek3h has joined #openstack-ansible | 11:34 | |
| chhavi | documentation no where mentioned about this. | 11:35 |
| drifterza | I think if we can keep to how other projects address the service tenants | 11:35 |
| drifterza | then we should be better off ? | 11:36 |
| drifterza | my personal opinion | 11:36 |
| *** chhavi_ has joined #openstack-ansible | 11:38 | |
| *** kylek3h has quit IRC | 11:39 | |
| *** chhavi has quit IRC | 11:39 | |
| chhavi_ | drifterza: Regarding the service_password comment, if u are using the trove admin/regular user for service configuration, u need to use service_password, so that it can be applied in the configuration file | 11:40 |
| chhavi_ | templates/trove.conf.j2:54:nova_proxy_admin_pass = {{ trove_service_password }} | 11:41 |
| chhavi_ | these passwords are used to feed into the configuration file, and they should be in sync with the user. | 11:41 |
| drifterza | okay I'm clear on this | 11:42 |
| chhavi_ | so what you suggest, how we should fix this. | 11:42 |
| chhavi_ | 1. Either we need to create a new service user as trove and get all these stuff in. | 11:43 |
| chhavi_ | 2. Or we can use an existing users and make them use. I don't see the existing users playing any role unless we make them use for service. | 11:43 |
| chhavi_ | As per the documentation, they mentioned these users should be used "These are the accounts used by the Trove service" | 11:44 |
| drifterza | okay so lets try keep with their documentation to some degree | 11:46 |
| drifterza | I honestly thought I catered for this | 11:47 |
| drifterza | in the service creation bits | 11:47 |
| chhavi_ | When we create the service, we do not have to specify any user details, we specify the tenant, so that is why I modified the tenant so that trove tenant can be used. | 11:48 |
| *** haad1 has quit IRC | 11:48 | |
| *** victorhe has joined #openstack-ansible | 11:49 | |
| drifterza | okay then I'm happy with the pr | 11:49 |
| chhavi_ | second: once the service is created they use the service_password, in all other project since there is a single user alligned with the project name, there are no new passwords to be created. | 11:50 |
| drifterza | Thanks for clarifying that | 11:50 |
| chhavi_ | in the case of trove its different, since we added 2 new password in the user_secrets, there is no service_password associated with any usr | 11:50 |
| drifterza | yeah | 11:51 |
| chhavi_ | can we update the review | 11:51 |
| drifterza | done | 11:52 |
| chhavi_ | thanks, i was trying to further test the trove by creating few instances | 11:53 |
| *** victorhe has quit IRC | 11:53 | |
| chhavi_ | I am getting the trove flavor-list as empty, do you have any idea how to create a trove flavor, I performed glance image-create and got the image there. | 11:53 |
| *** yolanda has quit IRC | 11:55 | |
| *** yolanda has joined #openstack-ansible | 11:55 | |
| drifterza | I havent had the time to test trove fully, I would have to check | 11:56 |
| chhavi_ | ok | 11:57 |
| drifterza | thats the next tempest test actually | 11:57 |
| chhavi_ | for the tempest we would be needing the steps to run, | 11:59 |
| drifterza | its on my todo list :) | 12:00 |
| drifterza | http://git.openstack.org/cgit/openstack/trove/tree/trove/tests/tempest/tests/api/database/flavors/test_flavors.py | 12:01 |
| drifterza | im thinking its probably a meta tag on a glance image | 12:02 |
| *** openstackgerrit has quit IRC | 12:03 | |
| *** openstackgerrit has joined #openstack-ansible | 12:04 | |
| openstackgerrit | Omer Anson proposed openstack/openstack-ansible-os_neutron: [WIP] Dragonflow https://review.openstack.org/391524 | 12:05 |
| drifterza | chhavi_, http://git.openstack.org/cgit/openstack/trove/tree/devstack/plugin.sh#n258 | 12:07 |
| drifterza | like here | 12:07 |
| *** schwicht has joined #openstack-ansible | 12:10 | |
| *** leduardo has quit IRC | 12:12 | |
| *** chhavi_ has quit IRC | 12:23 | |
| *** hw_wutianwei has quit IRC | 12:25 | |
| *** johnmilton has joined #openstack-ansible | 12:25 | |
| *** johnmilton has quit IRC | 12:26 | |
| *** hw_wutianwei has joined #openstack-ansible | 12:26 | |
| *** Mudpuppy has joined #openstack-ansible | 12:26 | |
| *** chhavi_ has joined #openstack-ansible | 12:28 | |
| *** Mudpuppy has quit IRC | 12:28 | |
| *** karimb has quit IRC | 12:28 | |
| openstackgerrit | Merged openstack/openstack-ansible: Add ceilometer-api-init-delete to run-upgrade.sh https://review.openstack.org/400847 | 12:29 |
| oanson | drifterza, I think I got it - the test wasn't running from the tempest venv | 12:29 |
| oanson | How can I verify the tempest venv was installed? Is there an overrides config option? A task/notifier I need to register to? | 12:30 |
| *** gouthamr has joined #openstack-ansible | 12:32 | |
| *** schwicht has quit IRC | 12:33 | |
| *** kylek3h has joined #openstack-ansible | 12:36 | |
| neith | lxc-top gives me this though my cluster is idle, no users, no instances... : infra1_keystone_co 329.17 22.13 305.45 6.67 MB 6.17 GB | 12:39 |
| neith | how come keystone uses 300% cpu | 12:39 |
| *** victorhe has joined #openstack-ansible | 12:39 | |
| *** weshay_ has quit IRC | 12:40 | |
| *** mmoster_ has joined #openstack-ansible | 12:40 | |
| *** kylek3h has quit IRC | 12:40 | |
| drifterza | oanson, looking | 12:43 |
| *** victorhe has quit IRC | 12:43 | |
| oanson | drifterza, I think I found it | 12:43 |
| oanson | I think tempest is installed in openstack1 container, and the test runs on localhost (i.e. not in a container) | 12:44 |
| openstackgerrit | Omer Anson proposed openstack/openstack-ansible-os_neutron: [WIP] Dragonflow https://review.openstack.org/391524 | 12:44 |
| drifterza | tempest usually runs from the inventory where utility_all | 12:44 |
| drifterza | is defined | 12:44 |
| oanson | That's what I saw. I modified the fullstack test to run from there as well | 12:45 |
| drifterza | lets see what the build does | 12:45 |
| oanson | All right. Thanks for your help so far. | 12:45 |
| drifterza | sure anytime | 12:45 |
| *** hw_wutianwei has quit IRC | 12:46 | |
| *** kylek3h has joined #openstack-ansible | 12:52 | |
| *** oanson has quit IRC | 12:53 | |
| *** victorhe has joined #openstack-ansible | 13:05 | |
| drifterza | anyone here using the lxd hypervisor ? | 13:09 |
| *** victorhe has quit IRC | 13:10 | |
| *** chhavi_ has quit IRC | 13:11 | |
| mgariepy | good morning everyone | 13:16 |
| *** schwicht has joined #openstack-ansible | 13:17 | |
| drifterza | hi | 13:18 |
| mgariepy | we have one or 2 lxd server here but not integrated in openstack | 13:20 |
| drifterza | ah okay, kinda testing the nova-lxd stuff | 13:20 |
| mgariepy | ok i've never tested it, at least not yet :) | 13:21 |
| *** haad1 has joined #openstack-ansible | 13:25 | |
| *** markvoelker has joined #openstack-ansible | 13:28 | |
| *** yolanda has quit IRC | 13:29 | |
| *** victorhe has joined #openstack-ansible | 13:33 | |
| evrardjp | andymccr: do you mind removing your -W on this? https://review.openstack.org/#/c/401184 We can still refactor later, and I'd like to see the results | 13:34 |
| *** victorhe has quit IRC | 13:37 | |
| *** johnmilton has joined #openstack-ansible | 13:39 | |
| *** schwicht has quit IRC | 13:39 | |
| *** chhavi_ has joined #openstack-ansible | 13:39 | |
| *** haad1 has quit IRC | 13:41 | |
| *** jamesdenton has joined #openstack-ansible | 13:44 | |
| *** dmsimard|away is now known as dmsimard | 13:45 | |
| evrardjp | odyssey4me: andymccr: asettle: do you know the rationale behind the job ^gate-openstack-ansible-deploy-guide$ ? | 13:46 |
| *** haad1 has joined #openstack-ansible | 13:53 | |
| *** karimb has joined #openstack-ansible | 13:54 | |
| *** jamesdenton has quit IRC | 13:55 | |
| *** woodard_ has joined #openstack-ansible | 14:01 | |
| *** fguillot has joined #openstack-ansible | 14:05 | |
| *** woodard has quit IRC | 14:05 | |
| asettle | evrardjp: Sorry, was out | 14:06 |
| asettle | What's your questino? | 14:06 |
| evrardjp | np :D | 14:06 |
| evrardjp | still up there | 14:06 |
| asettle | It's to build the deploy guide in docs.o.o | 14:07 |
| evrardjp | oh ok | 14:07 |
| andymccr | done! | 14:07 |
| evrardjp | there may be a typo there | 14:07 |
| evrardjp | branch: ^(?!(liberty|stable/(mitaka/newton))).*$ | 14:07 |
| evrardjp | instead of | 14:07 |
| evrardjp | branch: ^(?!(liberty|stable/(mitaka|newton))).*$ | 14:07 |
| evrardjp | but I'm refactoring anyway | 14:07 |
| evrardjp | :p | 14:07 |
| asettle | Uhh maybe I'm going blind byt I can't see the typo? | 14:08 |
| drifterza | slash vs pipe | 14:08 |
| asettle | Oh, ha. | 14:08 |
| *** thorst has joined #openstack-ansible | 14:08 | |
| asettle | Thanks drifterza | 14:09 |
| drifterza | ;) | 14:09 |
| evrardjp | it's also possible that my copy paste was wrong, so don't spend time on it | 14:09 |
| evrardjp | yeah sorry | 14:09 |
| evrardjp | :p | 14:09 |
| evrardjp | I'm with regexpes so I'm focusing on these chars right now :p | 14:09 |
| asettle | Okay :) thanks for catching that evrardjp | 14:09 |
| asettle | andymccr is going to fix that :) | 14:10 |
| evrardjp | I will don't worry | 14:10 |
| evrardjp | or he could | 14:10 |
| evrardjp | don't mind | 14:10 |
| evrardjp | as I am trying to see if we can refactor, that's maybe owrth a little wait, there is no hurry in fixing that right now | 14:11 |
| andymccr | evrardjp: i just added it to the PR that i already had | 14:11 |
| andymccr | its an obvious fix anyway :) | 14:11 |
| evrardjp | haha separate concerns, separate patches! | 14:13 |
| evrardjp | but yes | 14:13 |
| evrardjp | fine for me | 14:13 |
| asettle | Hahaha okay good job folks | 14:13 |
| *** tshtilma has quit IRC | 14:14 | |
| *** pcaruana has quit IRC | 14:14 | |
| *** jamesdenton has joined #openstack-ansible | 14:15 | |
| neith | guys, each time a user want to upload an image through horizon it get a 78eb25f826d5.js:1210 PUT https://openstack.irt-systemx.local/api/glance/images/ 403 (FORBIDDEN) | 14:15 |
| neith | I cant see anything in the haproxy config doing this | 14:16 |
| *** karimb has quit IRC | 14:17 | |
| andymccr | neith: are you able to upload the image via the cli (e.g. it works normally, but horizon doesn't)? | 14:17 |
| neith | andymccr: yes perfectly | 14:17 |
| andymccr | hmm we have a bug up that seemslike it may be related in that case: https://bugs.launchpad.net/openstack-ansible/+bug/1639080 | 14:18 |
| openstack | Launchpad bug 1639080 in openstack-ansible "Image uploads fail in Horizon if upload mode is set to direct if endpoint set to internal." [Low,Confirmed] - Assigned to Alexandra Settle (alexandra-settle) | 14:18 |
| andymccr | there are some instructions on things you can do to change that | 14:18 |
| andymccr | you could try and see if that works? | 14:18 |
| neith | andymccr: checking | 14:19 |
| *** thorst has quit IRC | 14:19 | |
| *** yolanda has joined #openstack-ansible | 14:21 | |
| evrardjp | sorry andymccr I didn't know you had to ack the remove-incubated-oslo-code | 14:22 |
| evrardjp | we* | 14:22 |
| andymccr | evrardjp: no problem, i did that already | 14:22 |
| *** thorst has joined #openstack-ansible | 14:22 | |
| andymccr | honestly, i didnt either - i think i missed the initial email which must've come out before the summit, but I got the "we need this done by friday" email and got it done that day :P | 14:23 |
| evrardjp | well I think there was something written in the openstack-dev lists | 14:23 |
| evrardjp | about releases | 14:23 |
| evrardjp | but I didn't realise we had to ack if we have nothing to do | 14:23 |
| openstackgerrit | Donovan Francesco (drifterza) proposed openstack/openstack-ansible-galera_server: Updating MariaDB yum repo to use SSL. https://review.openstack.org/401281 | 14:24 |
| evrardjp | I thought it was just "fix it" | 14:24 |
| andymccr | yeah same. | 14:24 |
| *** jamesdenton has quit IRC | 14:24 | |
| *** victorhe has joined #openstack-ansible | 14:26 | |
| *** victorhe has quit IRC | 14:26 | |
| *** victorhe has joined #openstack-ansible | 14:26 | |
| *** thorst has quit IRC | 14:31 | |
| *** schwicht has joined #openstack-ansible | 14:36 | |
| *** kjw3 has joined #openstack-ansible | 14:38 | |
| neith | andymccr: It seems to work | 14:40 |
| neith | andymccr: but I'm not completely sure of the conf | 14:40 |
| *** victorhe has quit IRC | 14:42 | |
| *** victorhe has joined #openstack-ansible | 14:43 | |
| *** mmoster_ has quit IRC | 14:46 | |
| *** survietamine has left #openstack-ansible | 14:48 | |
| neith | andymccr: my conclusion is that OSA should set OPENSTACK_ENDPOINT_TYPE is set to publicURL. | 14:55 |
| andymccr | neith: im not sure that would work, because for other requests horizon would need internalURL (for example AUTH with keystone) | 14:56 |
| andymccr | i might have a chat to one of the horizon devs and see their thoughts on that | 14:57 |
| *** chhavi_ has quit IRC | 14:58 | |
| neith | andymccr: using legacy does not fix the isssue though | 14:58 |
| andymccr | neith: interesting | 14:59 |
| neith | the upload reach 100% then TypeError: Cannot read property 'data' of undefined | 14:59 |
| *** vnogin has quit IRC | 15:02 | |
| *** satheesh has joined #openstack-ansible | 15:02 | |
| satheesh | Hi I have installed using openstack-ansible and ceph as a backend storage. I need to get the ceph metrics in ceilometer. Does anyone have the steps. | 15:02 |
| *** oanson has joined #openstack-ansible | 15:02 | |
| *** vnogin has joined #openstack-ansible | 15:04 | |
| *** vnogin has quit IRC | 15:04 | |
| *** thorst has joined #openstack-ansible | 15:05 | |
| *** weezS has joined #openstack-ansible | 15:08 | |
| *** havocbmx-LT has joined #openstack-ansible | 15:11 | |
| *** thorst has quit IRC | 15:12 | |
| havocbmx-LT | logan: I wanted to thank you again for your help yesterday | 15:14 |
| *** victorhe has quit IRC | 15:14 | |
| *** vnogin has joined #openstack-ansible | 15:14 | |
| *** chris_hultin|AWA is now known as chris_hultin | 15:15 | |
| *** thorst has joined #openstack-ansible | 15:16 | |
| havocbmx-LT | unfortunately, guys I have a new issue that I'm just slamming my head into the wall on. I have instances able to provision and glance doing images and things are looking good. Except, for some reason I can't create or import ssh keys. I uploaded an error from the create keypair error from nova api here: http://paste.openstack.org/show/590222/ | 15:17 |
| *** karimb has joined #openstack-ansible | 15:17 | |
| *** spotz_zzz is now known as spotz | 15:19 | |
| *** thorst has quit IRC | 15:20 | |
| *** schwicht has quit IRC | 15:20 | |
| *** agrebennikov has joined #openstack-ansible | 15:21 | |
| *** victorhe has joined #openstack-ansible | 15:21 | |
| *** victorhe_ has joined #openstack-ansible | 15:22 | |
| *** klamath has joined #openstack-ansible | 15:22 | |
| *** klamath has quit IRC | 15:22 | |
| *** klamath has joined #openstack-ansible | 15:23 | |
| andymccr | havocbmx-LT: interestingly: https://bugs.launchpad.net/openstack-ansible/+bug/1576755 | 15:24 |
| openstack | Launchpad bug 1576755 in openstack-ansible trunk "Paramiko 2.0 does not work with Nova" [High,Fix released] - Assigned to Kevin Carter (kevin-carter) | 15:24 |
| evrardjp | andymccr: have a look at L79 to L121 of this https://etherpad.openstack.org/p/layout.yml | 15:24 |
| andymccr | havocbmx-LT: are you running on master? | 15:25 |
| *** victorhe has quit IRC | 15:26 | |
| andymccr | evrardjp: taking a look | 15:27 |
| evrardjp | according to what I've seen we are worrying too much about specific cases which don't apply | 15:27 |
| *** drifterza_ has joined #openstack-ansible | 15:27 | |
| *** Mudpuppy has joined #openstack-ansible | 15:29 | |
| andymccr | evrardjp: probably - or rather i think what has happened is a fear of changing things wholesale so we have kept the existing rules in place over time it has grown and become more complex! | 15:29 |
| *** jamesdenton has joined #openstack-ansible | 15:29 | |
| evrardjp | exactly | 15:29 |
| evrardjp | This is why I'd like to try this | 15:30 |
| andymccr | yeah i agree | 15:30 |
| *** drifterza has quit IRC | 15:30 | |
| andymccr | it almost seemslike that repo should be versioned too per release hahah | 15:31 |
| *** omiday has joined #openstack-ansible | 15:31 | |
| andymccr | that way all the regex goes away and i just specify the gates that need to run for that version | 15:32 |
| *** jmckind has joined #openstack-ansible | 15:32 | |
| evrardjp | andymccr: omg that would have been so easier if we could just put that in our repos as metadata | 15:32 |
| evrardjp | I thought of it when writing the regexp | 15:33 |
| evrardjp | too much energy | 15:33 |
| evrardjp | I'd rather fix the regexp :p | 15:33 |
| andymccr | haha yeah | 15:33 |
| evrardjp | I think the idea here is to skip the exceptions | 15:33 |
| andymccr | i think that would require a change to the way the infra is setup | 15:33 |
| evrardjp | and we don't have much to what I've seen | 15:34 |
| evrardjp | all the upgrade jobs | 15:34 |
| evrardjp | the security with stig7 | 15:34 |
| evrardjp | and ceilometer with gnocchi | 15:34 |
| evrardjp | but we can excent | 15:35 |
| havocbmx-LT | evrardjp: I'm running latest stable newton openstack-ansible from last week | 15:35 |
| evrardjp | extend* | 15:35 |
| evrardjp | anyway, now I think I understand how it works, it's good enough for the future | 15:35 |
| *** victorhe_ has quit IRC | 15:35 | |
| mgariepy | havocbmx-LT, 13.3.8 is mitaka, not newton. | 15:36 |
| mgariepy | newton would be 14.something | 15:36 |
| havocbmx-LT | mgariepy: hmm...let me check and make sure then | 15:37 |
| andymccr | there is a bug: https://bugs.launchpad.net/nova/+bug/1585515 about stable/mitaka nova having issues | 15:37 |
| openstack | Launchpad bug 1585515 in OpenStack Compute (nova) "Paramiko doesn't work with Nova" [Undecided,New] | 15:37 |
| *** schwicht has joined #openstack-ansible | 15:37 | |
| *** janki has quit IRC | 15:38 | |
| *** h5t4 has quit IRC | 15:38 | |
| havocbmx-LT | mgariepy: we're using the stable/newton openstack-ansible | 15:39 |
| mgariepy | havocbmx-LT, that's strange then considering http://paste.openstack.org/show/590222/ | 15:42 |
| *** berendt has quit IRC | 15:44 | |
| havocbmx-LT | mgariepy: I see what your saying... | 15:45 |
| *** chhavi_ has joined #openstack-ansible | 15:45 | |
| havocbmx-LT | mgariepy: I'll do some more investigating here then brb | 15:45 |
| havocbmx-LT | mgariepy: I show in our history of the git pull that we pulled git checkout -b stable/newton | 15:47 |
| *** schwicht has quit IRC | 15:48 | |
| neith | andymccr: seems like I get random 504 errors in horizon | 15:48 |
| *** omiday has quit IRC | 15:49 | |
| andymccr | neith: 504 is a gateway timeout, can you see in the log where the timeout occurs? | 15:49 |
| *** kristia__ has joined #openstack-ansible | 15:49 | |
| neith | andymccr: thats what i'm checking | 15:49 |
| drifterza_ | okay so I've tested qemu,kvm and lxd on 14.0.2 | 15:50 |
| *** vnogin has quit IRC | 15:50 | |
| drifterza_ | I can't get lxd to work | 15:50 |
| kristia__ | Hi I have a question. Can I do a all in one deployment and have the deployment host on the deploy target?? | 15:50 |
| drifterza_ | http://paste.openstack.org/show/590224/ | 15:50 |
| *** satheesh has quit IRC | 15:51 | |
| *** vnogin has joined #openstack-ansible | 15:52 | |
| andymccr | kristia__: im not sure i understand - you want to do an AIO, and the host you run the aio on is the target? | 15:53 |
| havocbmx-LT | mgariepy: I've down a git status on openstack-deploy which shows we're on newton or should be | 15:53 |
| havocbmx-LT | mgariepy: output of the git status http://paste.openstack.org/show/590225/ | 15:53 |
| kristia__ | andymccr: yes | 15:53 |
| andymccr | kristia__: then yip :) that's how the AIO works, you just get one host and run the deploy from there | 15:54 |
| kristia__ | ok | 15:54 |
| andymccr | drifterza_: ive not run the lxd stuff really, i think the test does a boot and ssh though - so i guess it should work. | 15:54 |
| drifterza_ | I have 3 compute nodes using the same control backend, and lxd is the only one that is not spawning an instance | 15:55 |
| kristia__ | how is the support for magnum, trove, murano, ceilometer and gpu passthough? | 15:56 |
| *** klamath has quit IRC | 15:56 | |
| *** drifterza__ has joined #openstack-ansible | 15:57 | |
| *** drifterza__ is now known as drifterza | 15:57 | |
| *** drifterza_ has quit IRC | 16:01 | |
| openstackgerrit | Nolan Brubaker proposed openstack/openstack-ansible: Add command to create consolidated AIO config https://review.openstack.org/392300 | 16:01 |
| *** schwicht has joined #openstack-ansible | 16:04 | |
| *** vnogin has quit IRC | 16:07 | |
| *** haad1 has quit IRC | 16:07 | |
| *** haad1 has joined #openstack-ansible | 16:10 | |
| neith | andymccr: I never see any 504 in the haproxy logs : | 16:10 |
| neith | andymccr: what am I missing? | 16:11 |
| *** vnogin has joined #openstack-ansible | 16:11 | |
| havocbmx-LT | mgariepy: If this is mitaki what would be the best way to upgrade this to newton given that we were on the newton branch? | 16:12 |
| *** gouthamr has quit IRC | 16:12 | |
| evrardjp | well 504 is bad so if you don't have then it's good I guess :) | 16:12 |
| neith | evrardjp: I see it in my browser but not on the logs | 16:13 |
| *** gouthamr has joined #openstack-ansible | 16:13 | |
| evrardjp | do you have proper haproxy logs? | 16:14 |
| neith | evrardjp: yep | 16:14 |
| evrardjp | what do you see on your stat sockets? | 16:14 |
| *** karimb has quit IRC | 16:14 | |
| neith | evrardjp: stat sockets? | 16:15 |
| evrardjp | jmccrory: loved your comment there: https://review.openstack.org/#/c/396219/2/tests/test.yml@28 | 16:15 |
| evrardjp | hatop -s /var/run/haproxy.stat | 16:15 |
| evrardjp | neith: ^ | 16:15 |
| evrardjp | check if your backends answer correctly | 16:15 |
| evrardjp | man I should do a haproxy and keepalived guide | 16:15 |
| evrardjp | and refactor the haproxy role | 16:16 |
| evrardjp | I should stop talking and doing it | 16:16 |
| neith | evrardjp: lol | 16:16 |
| evrardjp | that's gonna be for later, after the tests and other things I have in mind. but it's gonna be great! | 16:17 |
| evrardjp | I need to find a consul role | 16:17 |
| *** Jack_Iv has quit IRC | 16:19 | |
| neith | evrardjp: I see 2 5xx errors throught the haproxy console | 16:20 |
| evrardjp | andymccr: so we are taking the german vs brasil approach for the glance upgrade? | 16:21 |
| evrardjp | https://www.youtube.com/watch?v=RLZUKqpXYzU | 16:21 |
| *** vnogin has quit IRC | 16:21 | |
| evrardjp | so your backends don't seem to run properly? | 16:21 |
| kristia__ | Im getting this error http://paste.openstack.org/show/590231/ Please may someone help me? | 16:22 |
| evrardjp | kristia__: there is something wrong with your inventory | 16:22 |
| evrardjp | probably something misplaced in openstack_user_config | 16:23 |
| evrardjp | did you lint it? | 16:23 |
| andymccr | haha evrardjp you mean the upgrade testing? | 16:23 |
| *** vnogin has joined #openstack-ansible | 16:23 | |
| evrardjp | yes | 16:23 |
| evrardjp | we don't do a seemless upgrade, we just try to upgrade on top of the other, right? | 16:23 |
| evrardjp | well let me check at the blueprint to see if there was recommendations about how to do it | 16:23 |
| andymccr | evrardjp: well the idea is that we put some testing in place. its step 1 | 16:24 |
| evrardjp | ok | 16:24 |
| andymccr | ideally we figure out proper upgrade path after | 16:24 |
| evrardjp | it's not everything together | 16:24 |
| evrardjp | ok | 16:24 |
| evrardjp | fine for me | 16:24 |
| andymccr | but i think this catches a few issues - nova already seems to not work | 16:24 |
| *** fguillot has quit IRC | 16:24 | |
| evrardjp | well glance seem to work | 16:24 |
| evrardjp | that was what I was reviewing | 16:24 |
| *** kjw3 has quit IRC | 16:25 | |
| evrardjp | and also, I wanted to re-share this video that's been forgotten | 16:25 |
| *** fguillot has joined #openstack-ansible | 16:26 | |
| *** paulfantom has joined #openstack-ansible | 16:27 | |
| kristia__ | evrardjp: How lint the openstack_user_config?? | 16:28 |
| *** vnogin has quit IRC | 16:28 | |
| evrardjp | kristia__: http://docs.openstack.org/developer/openstack-ansible/newton/install-guide/run-playbooks.html#checking-the-integrity-of-the-configuration-files | 16:30 |
| kristia__ | ok | 16:30 |
| evrardjp | yaml lint would be the help | 16:31 |
| evrardjp | if this passes, don't hesitate to use a pastebin and we'll check | 16:31 |
| evrardjp | logan-: are you there? | 16:33 |
| *** dxiri has joined #openstack-ansible | 16:33 | |
| logan- | yep. 'grats on the talk! | 16:34 |
| kristia__ | evrardjp: please may you help me with the config scince Im new to openstack-ansible? | 16:34 |
| evrardjp | hehe thanks logan- | 16:36 |
| evrardjp | kristia__: sure | 16:36 |
| kristia__ | thanks | 16:36 |
| kristia__ | in private? | 16:36 |
| evrardjp | kristia__: nope, for yours sake | 16:36 |
| kristia__ | ok | 16:36 |
| evrardjp | I'd be offline soon, and someone can continue my work :p | 16:36 |
| *** kristia__ is now known as kristian__ | 16:36 | |
| kristian__ | ok | 16:37 |
| kristian__ | then Im going to install openstack ansible on my server | 16:37 |
| kristian__ | because that was on a vps | 16:37 |
| evrardjp | logan-: could you tell me where you have issues like the one listed here: https://review.openstack.org/#/c/399984/1 ? | 16:37 |
| evrardjp | kristian__: don't | 16:37 |
| kristian__ | why? | 16:38 |
| evrardjp | I mean don't change your workflow for our help | 16:38 |
| *** thorst has joined #openstack-ansible | 16:38 | |
| evrardjp | we can help you without connecting, we probably just need to see the config files for this kind of error | 16:38 |
| evrardjp | don't worry :) | 16:38 |
| kristian__ | yeah | 16:39 |
| *** thorst has quit IRC | 16:39 | |
| *** vnogin has joined #openstack-ansible | 16:39 | |
| kristian__ | I currently use kolla but it doesnt have stuff that openstack-ansible has | 16:39 |
| evrardjp | logan-: because I don't this in master, and I'm pretty sure that the "if 'repo_all' in group_names else None" is flawed logic, and should be either avoided or moved to group vars | 16:39 |
| evrardjp | andymccr: ^ | 16:39 |
| kristian__ | and gpu passthrough doesnt work there | 16:39 |
| evrardjp | andymccr: I meant two lines above | 16:39 |
| evrardjp | ok | 16:40 |
| evrardjp | kristian__: ^ | 16:40 |
| *** allanice001 has joined #openstack-ansible | 16:40 | |
| evrardjp | kristian__: how can we help you? | 16:40 |
| kristian__ | currently bootstrapping according to the guide | 16:41 |
| kristian__ | I will need help with the config | 16:41 |
| evrardjp | kristian__: an AIO using the quickstart, or are you following the install guide? | 16:41 |
| evrardjp | and which release are you following? | 16:41 |
| *** dxiri has quit IRC | 16:41 | |
| kristian__ | http://docs.openstack.org/developer/openstack-ansible/newton/install-guide/deploymenthost.html | 16:41 |
| *** vnogin has quit IRC | 16:41 | |
| kristian__ | currently here | 16:42 |
| *** dxiri has joined #openstack-ansible | 16:42 | |
| paulfantom | Hi, I have a problem while installing keystone via ansible on CentOS7. It comes down to yum trying to download shibboleth from ftp.gwdg.de mirror, which doesn't hold all needed packages. Do you have any workaround? | 16:42 |
| evrardjp | kristian__: ubuntuy 16.04 ? | 16:43 |
| kristian__ | yes | 16:43 |
| evrardjp | ok | 16:43 |
| evrardjp | paulfantom: that's a good question | 16:44 |
| kristian__ | rebooting server | 16:44 |
| evrardjp | I suppose the shibboleth consortium should have a proper mirror | 16:44 |
| evrardjp | let me check | 16:44 |
| *** ricardobuffa_ufs has joined #openstack-ansible | 16:44 | |
| paulfantom | they use download.opensuse.org, but it redirects to ftp.gwdg.de | 16:45 |
| *** galstrom_zzz is now known as galstrom | 16:46 | |
| evrardjp | paulfantom: yes I just checked | 16:46 |
| evrardjp | well I'm not sure he redirect to .de for me | 16:46 |
| kristian__ | evrardjp: Ok got to the config part | 16:46 |
| evrardjp | paulfantom: I think it's distributed | 16:46 |
| evrardjp | oh no you're right | 16:47 |
| evrardjp | some work some dont | 16:47 |
| paulfantom | do you have one that works? | 16:48 |
| evrardjp | paulfantom: I think we should talk to the mirror maintener of ftp.gwdg.de | 16:48 |
| evrardjp | I'm not sure we are mirroring this | 16:48 |
| evrardjp | let me check | 16:48 |
| *** schwicht has quit IRC | 16:48 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_swift: Remove join filter from optional pip module tasks https://review.openstack.org/400359 | 16:50 |
| evrardjp | nope I don't have anything right now paulfantom | 16:50 |
| evrardjp | paulfantom: could you open a bug somewhere? | 16:51 |
| kristian__ | evrardjp: I have configured what I understood. Please may you help me on the rest? | 16:51 |
| kristian__ | http://paste.openstack.org/show/590236/ | 16:51 |
| evrardjp | I think it's good if you could post a bug, for example in openstack-ansible, telling the mirror is down for the file x or y, and this way we can track this | 16:52 |
| evrardjp | paulfantom: ^ | 16:52 |
| paulfantom | ok, I'll do it. | 16:52 |
| evrardjp | if this reproduces again, we can take actions | 16:52 |
| evrardjp | in the meantime, it's a question of asking the maintainer of ftp.gwdg.de to fix that | 16:52 |
| evrardjp | and because .de is now after business hours, I don't expect anything to be done today | 16:53 |
| evrardjp | kristian__: yup, one sec | 16:53 |
| kristian__ | ok | 16:53 |
| evrardjp | paulfantom: maybe they are syncing right now and it's maybe gonna be fixed tomorrow. But anyway, that's a risk to be listed somewhere | 16:53 |
| evrardjp | kristian__: you're using an aio right? | 16:55 |
| kristian__ | yes | 16:55 |
| evrardjp | what did you do right now to get this config file | 16:55 |
| evrardjp | bootstrap-aio.sh ? | 16:55 |
| kristian__ | no | 16:55 |
| *** drifterza has quit IRC | 16:55 | |
| evrardjp | I'm all ears :) | 16:56 |
| kristian__ | from /etc/openstacl-ansible/openstack_user_config.yml.aio | 16:56 |
| *** vnogin has joined #openstack-ansible | 16:56 | |
| *** galstrom is now known as galstrom_zzz | 16:56 | |
| kristian__ | *openstack_deploy | 16:56 |
| logan- | evrardjp: sorry-had an interruption but I use those vars to heavily limit what gets installed on my computes | 16:57 |
| evrardjp | logan-: I understand why, I'd like to see where and think of a better solution if possible | 16:59 |
| evrardjp | kristian__: there is an easy way to have an aio | 17:00 |
| evrardjp | there is a specific doc for that | 17:00 |
| evrardjp | are you interested? | 17:00 |
| kristian__ | yes | 17:00 |
| evrardjp | kristian__: here: http://docs.openstack.org/developer/openstack-ansible/newton/developer-docs/quickstart-aio.html | 17:00 |
| logan- | understood -- I could post a more complete list of my overrides for newton once they're done and we could look at ways to segment where/what we install. i think there's definitely room to optimize especially w/r/t things that only need to be installed in the repo containers | 17:01 |
| evrardjp | it's gonna help you I think | 17:01 |
| evrardjp | logan-: I think that now that we have group_vars it's time to properly use this :p | 17:01 |
| logan- | group vars don't really work for this with how we have package installs set up in the roles | 17:02 |
| evrardjp | but anyway the condition looked really ugly because it was linking to inventory in the role | 17:02 |
| logan- | role vars > group vars | 17:02 |
| evrardjp | which I'd like to avoid as much as possible for portability | 17:02 |
| evrardjp | logan-: that's true | 17:02 |
| evrardjp | but if it's something that could be considered as conditional it shouldn't be set as a static var | 17:02 |
| evrardjp | imo | 17:02 |
| openstackgerrit | Merged openstack/openstack-ansible-os_glance: Glance upgrade testing. https://review.openstack.org/396219 | 17:03 |
| kristian__ | ok going to try not destroy my server :D | 17:03 |
| kristian__ | evrardjp: what about lxd address and ranges in /etc/openstack_deploy/user_variables.yml | 17:03 |
| evrardjp | we may need to break up the var into smaller vars and combine them in the tasks | 17:03 |
| logan- | yep | 17:04 |
| evrardjp | kristian__: when following the http://docs.openstack.org/developer/openstack-ansible/newton/developer-docs/quickstart-aio.html , the bootstrap-aio script will handle the creation of that | 17:04 |
| evrardjp | if you're not happy with it, you can still customize it before running tbhe run-playbooks | 17:05 |
| kristian__ | ok | 17:05 |
| kristian__ | running | 17:05 |
| *** jmckind_ has joined #openstack-ansible | 17:05 | |
| *** jmckind has quit IRC | 17:06 | |
| *** thorst has joined #openstack-ansible | 17:06 | |
| evrardjp | kristian__: if you've done things on /etc/openstack_deploy/, I cannot ensure everything will be running fine, this guide was used as "here is how to quick start from nothing" | 17:07 |
| evrardjp | which is what you are looking for in testing right? | 17:07 |
| evrardjp | and then you can go a little further, step by step | 17:07 |
| *** thorst_ has joined #openstack-ansible | 17:07 | |
| kristian__ | we will see if it will deploy | 17:07 |
| kristian__ | it should | 17:07 |
| *** haad1 has quit IRC | 17:07 | |
| kristian__ | I didnt do anything that would break the install | 17:08 |
| kristian__ | currently on lxc_hosts creating images | 17:09 |
| dxiri | hey guys, I have one question, can I have my controller running on a different subnet than my compute hosts? | 17:09 |
| *** sdake_ is now known as sdake | 17:09 | |
| dxiri | are there any special considerations? | 17:09 |
| *** thorst has quit IRC | 17:11 | |
| *** thorst_ has quit IRC | 17:11 | |
| evrardjp | dxiri: you're doomed! | 17:14 |
| evrardjp | no, I'm joking | 17:14 |
| dxiri | evrardjp: lol :P | 17:14 |
| evrardjp | but yes there is I guess things to do | 17:14 |
| evrardjp | dxiri: but there is connectivity between the two? | 17:14 |
| dxiri | osa really got traction over here, so now we will have a sort of deployment cloud with big boxes, so that people can spin up there controllers there using VMs, and people will later add there own compute nodes | 17:15 |
| dxiri | evrardjp: yes there is connectivity between the 2 | 17:15 |
| evrardjp | so as long as you can connect to your clusters it should be fine | 17:16 |
| evrardjp | for example your compute node can connect on rabbit etc... | 17:16 |
| evrardjp | it's not an often used case | 17:16 |
| evrardjp | but I don't see a problem that could be caused by routing the mgmt | 17:16 |
| dxiri | these are all lab boxes so I get to experiment :) | 17:17 |
| evrardjp | dxiri: will the LB be reachable from everywhere? | 17:17 |
| dxiri | evrardjp: yes | 17:17 |
| dxiri | provided people have their VPN up | 17:17 |
| evrardjp | so basically it's the same network :p | 17:17 |
| evrardjp | haha! | 17:18 |
| evrardjp | now I understand better what you're doing | 17:18 |
| evrardjp | you want to vpn the control plane from the compute to the controller networks | 17:18 |
| evrardjp | right? | 17:18 |
| evrardjp | routed vpn | 17:18 |
| dxiri | no, let me put some context | 17:18 |
| dxiri | lets say I have a controller sitting on a VM inside a 172.1.1.0/24 network | 17:19 |
| evrardjp | ok | 17:19 |
| dxiri | and I have a couple of compute hosts sitting on 172.2.2.0/24 | 17:19 |
| evrardjp | only one nic right | 17:19 |
| dxiri | L3 connectivity is all good, I can ping both ways | 17:19 |
| dxiri | only 1 nic correct | 17:19 |
| evrardjp | inside your vm, you bridge your br-mgmt to 172.1.1.0/24 ? | 17:20 |
| dxiri | yes | 17:21 |
| evrardjp | that sentence doesn't make sense network wise but you got me | 17:21 |
| dxiri | yeah I got your point | 17:21 |
| evrardjp | and your external lb vip ip is on the same subnet then, but with a different ip | 17:21 |
| evrardjp | than your internal lb vip | 17:22 |
| dxiri | on 172.1.1.0/24 yes | 17:22 |
| dxiri | for the external lb vip | 17:22 |
| evrardjp | that's weird, man! :p | 17:23 |
| evrardjp | but it should work just fine | 17:23 |
| evrardjp | you just have bridged the lxc br-mgmt broadcast domain to your 172.1.1.0/24 | 17:23 |
| evrardjp | I hope the l2 of this network is not too heavily crowded, for your own sake :p | 17:24 |
| evrardjp | but technically it has no impact | 17:24 |
| evrardjp | on your deployment I mean | 17:24 |
| evrardjp | oh wait small question | 17:25 |
| evrardjp | behind your lb | 17:25 |
| evrardjp | on your vm br-mgmt | 17:25 |
| evrardjp | which kind of IP you have? | 17:25 |
| dxiri | no HA for now, if that is your question | 17:25 |
| dxiri | just a single controller and everything would go there | 17:26 |
| evrardjp | dxiri: no | 17:26 |
| evrardjp | for example what's the ip of your galera containers | 17:26 |
| *** jmckind has joined #openstack-ansible | 17:26 | |
| evrardjp | or of your keystone containers, or rabbit containers | 17:26 |
| havocbmx-LT | mgariepy: thankfully, we're no in full prod yet so we are rerolling again after blowing it all up and verifying that it's on newton. You were right for some reason it was rolling mitaki even though according to the logs and checkout we were on newton after we repulled everything it's showing the right stuff now so fingers crossed | 17:26 |
| dxiri | ah....I was thinking to use the same as the examples on the appendix for osa | 17:27 |
| evrardjp | well that's gonna be involving routing then | 17:28 |
| evrardjp | :p | 17:28 |
| evrardjp | both ways remember ! | 17:28 |
| *** jmckind_ has quit IRC | 17:28 | |
| *** jamesde__ has joined #openstack-ansible | 17:28 | |
| *** javeriak has joined #openstack-ansible | 17:28 | |
| *** allanice001 has quit IRC | 17:29 | |
| evrardjp | the easier, IMO would be to bridge your vm interface in 172.1.1.0/24 to your br-mgmt, and use this same range for br-mgmt inside openstack-ansible | 17:29 |
| evrardjp | this way, you all work internally | 17:29 |
| evrardjp | (except if you want to isolate for security reasons, but then you have to figure out ways to bring the compute nodes into this network | 17:29 |
| dxiri | you mean use 172.1.1.0/24for the container cidr on the config? | 17:30 |
| *** jamesdenton has quit IRC | 17:30 | |
| evrardjp | let me verify the name but yes | 17:30 |
| *** paulfantom has quit IRC | 17:31 | |
| evrardjp | yup | 17:31 |
| evrardjp | don't forget to proper list what was used in used_ips | 17:31 |
| evrardjp | if you cannot do that you have no choice, you're gonna have fun with routes! | 17:32 |
| evrardjp | static routes of course, because irp are too fun | 17:32 |
| *** javeriak has quit IRC | 17:32 | |
| kristian__ | evrardjp: where can I change public ip of openstack? | 17:33 |
| evrardjp | kristian__: for your aio? | 17:33 |
| kristian__ | yes | 17:33 |
| evrardjp | did you deploy it already? | 17:33 |
| kristian__ | its deploying :/ | 17:34 |
| evrardjp | if you're deploying it's already too late | 17:34 |
| evrardjp | :p | 17:34 |
| evrardjp | what step are you? | 17:34 |
| kristian__ | os_cinder | 17:34 |
| evrardjp | too late then | 17:34 |
| evrardjp | :p | 17:34 |
| kristian__ | then how to tear it down?? | 17:34 |
| evrardjp | you just have to delete the containers with a playbook it's quite easy | 17:34 |
| evrardjp | but before doing that | 17:35 |
| evrardjp | your external lb vip is in /etc/openstack_deploy/openstack_user_config.yml | 17:35 |
| evrardjp | external_lb_vip_address | 17:35 |
| kristian__ | ok | 17:35 |
| kristian__ | what about cidr_networks and used ips? | 17:36 |
| kristian__ | also those networks? | 17:36 |
| dxiri | evrardjp: cool thanks :) | 17:38 |
| dxiri | evrardjp: will have to look if I have enough space on the subnet to do the containers on the same subnet | 17:39 |
| *** jlockwood has joined #openstack-ansible | 17:39 | |
| ricardobuffa_ufs | Hi everyone. We were having this error before but our configuration was not exactly the same of the test environment, but now we changed to get the configuration as close of the test environment as possible. | 17:39 |
| ricardobuffa_ufs | The slightest difference now it's just we got 2 infras and 2 compute nodes. Before we were defining the storage in the infra machines, but now we are using a single machine for storage. http://pastebin.com/Uv7dDrKQ | 17:39 |
| openstackgerrit | Jimmy McCrory proposed openstack/openstack-ansible-galera_server: [WIP] Install MariaDB 10.1 server https://review.openstack.org/401123 | 17:43 |
| *** victorhe has joined #openstack-ansible | 17:45 | |
| kristian__ | enrardjp: how can I configure networks when I want enp2s0 for neutron? | 17:46 |
| *** javeriak has joined #openstack-ansible | 17:47 | |
| *** javeriak has quit IRC | 17:48 | |
| *** vnogin has quit IRC | 17:50 | |
| *** javeriak has joined #openstack-ansible | 17:50 | |
| *** vnogin has joined #openstack-ansible | 17:50 | |
| jmccrory | ricardobuffa_ufs : the infra hosts are still in the storage_hosts group in your inventory, you'll need to fix that. there's a script to help there: http://docs.openstack.org/developer/openstack-ansible/newton/developer-docs/inventory.html | 17:51 |
| *** javeriak has quit IRC | 17:52 | |
| *** oneswig has quit IRC | 17:53 | |
| *** jamesde__ has quit IRC | 17:54 | |
| *** sdake has quit IRC | 17:54 | |
| *** kristian__ has quit IRC | 17:55 | |
| *** vnogin has quit IRC | 17:55 | |
| jmccrory | ricardobuffa_ufs : actually, i'm not sure if that will allow you to remove just the hosts from just that group. you might need to edit the json file directly | 17:55 |
| *** javeriak has joined #openstack-ansible | 17:57 | |
| ricardobuffa_ufs | jmccrory : Can we remove /etc/openstack_deploy/openstack_inventory.json and run everything from the beginning? | 17:58 |
| jmccrory | yep, that would be even easier. make sure to destroy any containers first: lxc-containers-destroy.yml playbook | 18:00 |
| jmccrory | ricardobuffa_ufs : this is still a test environment, right? | 18:00 |
| ricardobuffa_ufs | yes, still a test at the start | 18:01 |
| *** javeriak has quit IRC | 18:02 | |
| ricardobuffa_ufs | Thanks a lot, jmccrory, we're going to try it | 18:08 |
| *** drifterza has joined #openstack-ansible | 18:09 | |
| *** vnogin has joined #openstack-ansible | 18:10 | |
| evrardjp | too bad kristian_ left | 18:11 |
| evrardjp | dxiri: it is possible to make it work routed if you can't find room, but I thought it would simplify things to have everything set up on one net | 18:12 |
| drifterza | are there known issues with nova-lxd and vxlan ? | 18:14 |
| cloudnull | drifterza: I've not seen any. | 18:14 |
| drifterza | bizarre | 18:14 |
| cloudnull | though I've not really burned in on it | 18:14 |
| evrardjp | hey cloudnull and jmccrory! | 18:14 |
| drifterza | and nova-lxd and booting from ceph for ephemeral ? | 18:14 |
| drifterza | that probably doesnt work | 18:15 |
| jmccrory | hi evrardjp | 18:15 |
| cloudnull | ephemeral storage in lxd needs to be local on the box using zfs or btrfs | 18:15 |
| cloudnull | otherwise it'll through an error saying no valid host found, or something similart | 18:15 |
| cloudnull | **similar | 18:15 |
| drifterza | so if we have | 18:16 |
| drifterza | nova_libvirt_images_rbd_pool defined | 18:16 |
| drifterza | and lxd as nova_virt_type | 18:16 |
| drifterza | it will error out | 18:16 |
| drifterza | ? | 18:16 |
| drifterza | I think I'm hitting this | 18:16 |
| cloudnull | drifterza: https://github.com/openstack/nova-lxd/blob/9bf9db7e52a37ab7761ea63daacd3a41725d98b8/nova/virt/lxd/driver.py#L1194-L1195 | 18:17 |
| drifterza | I dont get that error | 18:17 |
| drifterza | I get http://paste.openstack.org/show/590224/ | 18:17 |
| cloudnull | hey evrardjp | 18:18 |
| openstackgerrit | Chhavi Agarwal proposed openstack/openstack-ansible-os_trove: Trove service user configuration should be aligned https://review.openstack.org/401054 | 18:18 |
| cloudnull | drifterza: odd. | 18:19 |
| cloudnull | do you have ephemeral storage in the flavor ? | 18:19 |
| drifterza | cloudnull, I'm setting nova_virt_type on individual compute nodes inventory | 18:19 |
| *** Jack_Iv has joined #openstack-ansible | 18:19 | |
| cloudnull | also can you use the lxd api locallaly? | 18:19 |
| drifterza | so I have kvm for some and lxd for others | 18:19 |
| cloudnull | ah cool | 18:20 |
| chhavi_ | evrajdp: As per your review comments I have updated the commit message, | 18:20 |
| drifterza | and the lxd ones are the only ones that screw out | 18:20 |
| drifterza | same control plane | 18:20 |
| drifterza | etc.. | 18:20 |
| evrardjp | chhavi_: ok I'll review | 18:20 |
| cloudnull | you will need to set the image meta-data to the correct hypervisor-type | 18:20 |
| chhavi_ | In the trove service setup, we had already commented that we are following the manual_install document. | 18:20 |
| cloudnull | and you have to do that via the cli | 18:20 |
| evrardjp | drifterza: cloudnull would you think that this is worth a documentation in nova? How to run multiple things in the same cluster? | 18:20 |
| cloudnull | horizon does not have an option to define lxd | 18:20 |
| drifterza | evrardjp, definately | 18:20 |
| cloudnull | evrardjp: boom! https://cloudnull.io/2016/11/multi-hypervisor-openstack-cloud/ | 18:21 |
| drifterza | cloudnull, I upload the raw xenial image to glance | 18:21 |
| drifterza | kvm consumes no issues | 18:21 |
| cloudnull | hahaha | 18:21 |
| drifterza | so I need to set meta for lxd specifically? | 18:21 |
| drifterza | so two images exactly the same ? | 18:21 |
| drifterza | but update the ones meta to reflect hypervisor type lxd | 18:22 |
| cloudnull | drifterza: set the hypervisor-type on the image to lxd using the openstack cli | 18:22 |
| cloudnull | openstack image set "$i" --property hypervisor_type=lxd | 18:22 |
| drifterza | God if its that I will freak | 18:23 |
| cloudnull | evrardjp: but yes. more formal docs, would be great | 18:23 |
| cloudnull | something better than my mindless ramblings | 18:23 |
| drifterza | interesting would setting hypervisor type effect kvm deployments | 18:23 |
| cloudnull | yes. | 18:24 |
| drifterza | oh suck | 18:24 |
| cloudnull | i'd bet that you have more kvm than lxd | 18:24 |
| drifterza | so we need to cart around two images for everything | 18:24 |
| evrardjp | cloudnull: I like your ramblings! not the mindless ones 'though | 18:24 |
| cloudnull | so kvm works more often because luck | 18:24 |
| drifterza | LOL | 18:24 |
| drifterza | I wonder if using LXD in prod is a good idea | 18:25 |
| cloudnull | drifterza: https://cloudnull.io/2016/11/multi-hypervisor-openstack-cloud/#isolatingyourexistingimages | 18:25 |
| evrardjp | ask that to docker guys | 18:25 |
| evrardjp | they will tell you it's production ready | 18:25 |
| cloudnull | you can pull the list of our current images. and set them all to qemu | 18:25 |
| drifterza | the crap part is you cant use iscsi paths or ceph ephemeral | 18:25 |
| evrardjp | "I don't see why not" | 18:25 |
| evrardjp | this works with lxd too | 18:25 |
| cloudnull | then redefine your one image to be lxd | 18:25 |
| evrardjp | cloudnull: set them all to qemu? | 18:26 |
| drifterza | I'm not keen to hack the host machine to do bind mounts for iscsi paths | 18:26 |
| drifterza | that sucks to me | 18:26 |
| cloudnull | qemu / kvm same thing | 18:26 |
| cloudnull | nova sees kvm hypervisors as qemu | 18:27 |
| evrardjp | oh yes I misread what you meant | 18:27 |
| evrardjp | you didn't mean to use qemu utils to move them from a format to another, you just wanted to re-upload them by forcing to specify the hypervisor_type | 18:27 |
| cloudnull | no. | 18:27 |
| cloudnull | drifterza: once the metadata is all in place, everything will work and nobdody will be the wiseer. | 18:28 |
| evrardjp | to qemu | 18:28 |
| evrardjp | cloudnull: isn't it a good practice anyway to specify hypervisor_type when you have more than one hypervisor? | 18:29 |
| cloudnull | yes. | 18:29 |
| cloudnull | in newton if you wanted to run ironic and kvm in the same region you'd need to define the hypervisor type | 18:29 |
| evrardjp | I don't like these y/n games, I am leaving for today! | 18:29 |
| cloudnull | eles things would get very confued. | 18:30 |
| drifterza | cloudnull, I'm assuming you guys havent used lxd with shared-storage, IE iscsi | 18:30 |
| *** vnogin has quit IRC | 18:30 | |
| cloudnull | drifterza: i've not. | 18:30 |
| cloudnull | evrardjp: :( | 18:30 |
| cloudnull | come back | 18:30 |
| cloudnull | :) | 18:30 |
| irtermite | ask a y/n question, get a y/n answer ;) | 18:30 |
| evrardjp | I've discovered today that rocket league is installed on the ps4 at the office | 18:31 |
| drifterza | I think as a default one should always specify hypervisor_type | 18:31 |
| *** vnogin has joined #openstack-ansible | 18:31 | |
| *** chhavi_ has quit IRC | 18:31 | |
| drifterza | even if its only a qemu env | 18:31 |
| cloudnull | oh. rocket leage is fun | 18:31 |
| evrardjp | irtermite: I've got the answer I deserve! but I like it that way :p | 18:31 |
| cloudnull | drifterza: ++ i tend to agree | 18:31 |
| irtermite | hehe | 18:31 |
| drifterza | cloudnull, Ill try the meta update | 18:31 |
| drifterza | hopefully it will work then | 18:31 |
| drifterza | this is 14.0.2 fyi | 18:31 |
| cloudnull | nice! | 18:32 |
| evrardjp | cloudnull: well it was my second session of rocket league ever, and I won one of the two games we had! I'm eager to buy a pc or a console now | 18:32 |
| cloudnull | dxiri: did you get the dhcp things working from the other day? | 18:32 |
| cloudnull | evrardjp: PC FTMFW | 18:32 |
| drifterza | evrardjp, did you look at https://review.openstack.org/#/c/401054/ | 18:32 |
| cloudnull | :) | 18:32 |
| cloudnull | neith: hows the POC going? | 18:33 |
| *** electrofelix has quit IRC | 18:33 | |
| evrardjp | drifterza: not yet, but I just promised I'll be doing it | 18:34 |
| dxiri | cloudnull: not yet, still at the concept stage | 18:34 |
| evrardjp | oh man I never really used trove, but I already start to hate it :p | 18:34 |
| dxiri | same thing I asked the separate controller questions today | 18:35 |
| drifterza | evrardjp, its a pig dude | 18:35 |
| *** vnogin has quit IRC | 18:35 | |
| evrardjp | hehe | 18:35 |
| *** markvoelker has quit IRC | 18:35 | |
| evrardjp | I don't know if I should love or hate | 18:35 |
| evrardjp | because bacon obviously | 18:35 |
| cloudnull | dxiri: seperate controller? | 18:35 |
| drifterza | well infra seems to use it | 18:36 |
| * cloudnull may be missing scroll back | 18:36 | |
| drifterza | so it cant be all that bad | 18:36 |
| dxiri | cloudnull: scroll up and see my chat with evrardjp | 18:36 |
| drifterza | I also hate that we cant use service/trove | 18:36 |
| drifterza | seems so silly | 18:36 |
| evrardjp | drifterza: yes, or use config template for this seems overkill, it should be done correctly from the start! | 18:36 |
| evrardjp | cloudnull: dxiri is willing to have control nodes on a separate subnet than compute nodes | 18:37 |
| evrardjp | I told him it was not hard if they have all proper ip connectivity | 18:37 |
| evrardjp | that's the summary | 18:37 |
| drifterza | evrardjp, sadly I think we are stuck between a rock and a hard place with this | 18:37 |
| drifterza | which sucks | 18:37 |
| drifterza | unless you have a clever way to manipulate this | 18:37 |
| drifterza | I'm sure you do | 18:38 |
| drifterza | :) | 18:38 |
| evrardjp | well I didn't want to go that place | 18:38 |
| drifterza | comon you love this sorta of hackery stuff | 18:38 |
| evrardjp | I'm challenged enough | 18:38 |
| evrardjp | :p | 18:38 |
| evrardjp | I do | 18:38 |
| evrardjp | darn it! | 18:38 |
| cloudnull | ++ dxiri yes. you can do it :) | 18:38 |
| cloudnull | so long as there's a route, anything will work. | 18:39 |
| *** jlockwood has quit IRC | 18:39 | |
| evrardjp | that's what I said | 18:39 |
| evrardjp | almost | 18:39 |
| evrardjp | but that's what I meant! | 18:39 |
| drifterza | evrardjp, if you keen to hack this one, it would be awesome. If not we can live with this hack | 18:39 |
| cloudnull | we're using 2 regions in the osic now, both on different subnets and a miserable mix of routes allowing interconnectivity. | 18:40 |
| evrardjp | let's live with the hack until someone complaints | 18:40 |
| evrardjp | but man the commit message is ugly | 18:40 |
| dxiri | cloudnull: evrardjp https://www.youtube.com/watch?v=qztuEucrNBc | 18:40 |
| drifterza | yeah | 18:40 |
| drifterza | and a release note should be part of this PR | 18:40 |
| cloudnull | exactly. :) | 18:40 |
| *** jlockwood has joined #openstack-ansible | 18:41 | |
| evrardjp | I'm fixing this it hurts me | 18:41 |
| drifterza | yay | 18:42 |
| drifterza | ++ for evrardjp | 18:42 |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-os_trove: Trove service user configuration should be aligned https://review.openstack.org/401054 | 18:42 |
| evrardjp | chhavi_ I've updated your commit message | 18:44 |
| *** karimb has joined #openstack-ansible | 18:44 | |
| evrardjp | ^ useless message for posterity | 18:44 |
| cloudnull | ok back to my ironic pit of dispare. | 18:47 |
| evrardjp | hehe | 18:47 |
| evrardjp | enjoy | 18:47 |
| cloudnull | oh you know it | 18:47 |
| evrardjp | :p | 18:47 |
| evrardjp | cloudnull: buying a pc for gaming is far more expensive than buying a console, for what I play at least :p | 18:48 |
| *** kristian__ has joined #openstack-ansible | 18:48 | |
| evrardjp | and I have kilometers away from me | 18:48 |
| cloudnull | get a server then play rocket league on it :) | 18:48 |
| evrardjp | welcome back kristian__! | 18:49 |
| cloudnull | which can also be used for dev | 18:49 |
| evrardjp | cloudnull: that doesn't need a proper gpu or you meant a server with a proper gpu? | 18:49 |
| evrardjp | (it's not a y/n question!) | 18:49 |
| cloudnull | https://cloudnull.io/2016/11/home-server-episode-1/ | 18:49 |
| evrardjp | I expected a True | 18:49 |
| evrardjp | oh man I am so sad for you | 18:50 |
| evrardjp | I tried this it was an awful experience | 18:50 |
| cloudnull | 16 core, 150gb ram, many ssds, r9 gpu. | 18:50 |
| cloudnull | works well so far | 18:50 |
| evrardjp | you passthrough your gpu to a vm right? | 18:50 |
| cloudnull | yes. | 18:50 |
| cloudnull | that's what I'm doing now. | 18:51 |
| cloudnull | iommu + sriov | 18:51 |
| drifterza | oh man I'm so over gaming | 18:51 |
| cloudnull | w/ gpu passthrough | 18:51 |
| evrardjp | oh man this shit broke everytime I had to a pacman -Syu | 18:51 |
| cloudnull | no pacman. :p | 18:51 |
| drifterza | yaourt -Syu | 18:51 |
| drifterza | for the win | 18:51 |
| evrardjp | yaourt indeed | 18:51 |
| cloudnull | my box is an osa deployment | 18:51 |
| evrardjp | but I didn't know if ppl knew :p | 18:51 |
| evrardjp | yaourt is french btw | 18:52 |
| drifterza | yeah? | 18:52 |
| evrardjp | not only the name | 18:52 |
| evrardjp | yeah | 18:52 |
| drifterza | nice | 18:52 |
| evrardjp | not that it's a good thing or a bad thing | 18:52 |
| drifterza | I didnt know that | 18:52 |
| evrardjp | because I am belgian you know :p | 18:52 |
| *** ricardobuffa_ufs has quit IRC | 18:52 | |
| evrardjp | but that's too much possible sources of trolling for one day | 18:53 |
| *** kristian__ has quit IRC | 18:53 | |
| evrardjp | cloudnull: with lxd? | 18:53 |
| cloudnull | w/ kvm on the main box, lxd on one of my brix. | 18:53 |
| evrardjp | I mean it's a deploy with osa and libvirt/kvm + lxd ? I am pretty sure you're good on that | 18:53 |
| evrardjp | :p | 18:53 |
| *** victorhe has quit IRC | 18:53 | |
| cloudnull | I had ironic doing things but have decided i no longer wanted to do that. | 18:54 |
| evrardjp | cloudnull: I am pretty sure you'll be leading the upgrade work now | 18:54 |
| cloudnull | If i can get some time to do things it'd be great. | 18:54 |
| drifterza | replace ironic with maas dude | 18:54 |
| drifterza | and be done with life | 18:54 |
| *** Jack_Iv has quit IRC | 18:54 | |
| evrardjp | drifterza: cloudnull is already used to cobbler | 18:55 |
| evrardjp | no need for another shiny thing | 18:55 |
| evrardjp | it has a cli, and can be interfaced | 18:55 |
| cloudnull | drifterza: ++ | 18:55 |
| *** Jack_Iv has joined #openstack-ansible | 18:55 | |
| cloudnull | maas is great. but i do like my cobbler things | 18:55 |
| evrardjp | is maas good? | 18:55 |
| drifterza | jeeze its the shit dude | 18:55 |
| *** victorhe has joined #openstack-ansible | 18:55 | |
| cloudnull | the latest maas release is fantaatic | 18:56 |
| drifterza | is it out of beta now | 18:56 |
| drifterza | maas 2.0 | 18:56 |
| evrardjp | cloudnull: cobbler sucks for the debian partitioning | 18:56 |
| drifterza | I'm still using the bzr one | 18:56 |
| evrardjp | drifterza: I am not sure if you're sarcastic or not | 18:56 |
| evrardjp | that's good | 18:56 |
| drifterza | evrardjp, I got maas booting windows 2012r2 on metal | 18:56 |
| logan- | might have to try maas. im on cobbler right now too | 18:57 |
| drifterza | cobbler sucks compared to maas | 18:57 |
| drifterza | and maas can run inside lxd | 18:57 |
| evrardjp | I wanted to try maas long ago | 18:58 |
| evrardjp | but it sucked at that time | 18:58 |
| drifterza | look it was shit a awhile ago | 18:58 |
| drifterza | i would recommend a revisit if you have the time | 18:58 |
| cloudnull | evrardjp: partitioning is hard w/ preseeds but i have that down pat | 18:58 |
| mgariepy | evrardjp, presseed suck for partitionning | 18:59 |
| logan- | kickstart partitioning makes so much more sense | 18:59 |
| drifterza | cloudnull, with maas 2.0, partitioning is easy with the UI | 18:59 |
| evrardjp | cloudnull: I remember I had to use %, minimum space, another partition with %, and it was a nightmare | 18:59 |
| *** Jack_Iv has quit IRC | 18:59 | |
| evrardjp | mgariepy: I think we all agree | 18:59 |
| evrardjp | mgariepy: that's why you are on centos right | 19:00 |
| evrardjp | anaconda is the deal | 19:00 |
| evrardjp | works perfectly with ks | 19:00 |
| mgariepy | lol i'm not on centos yet. | 19:00 |
| evrardjp | files | 19:00 |
| cloudnull | drifterza: ++ maas is nice | 19:00 |
| drifterza | ah shit with maas they have lvm support now part of the I | 19:00 |
| evrardjp | ok I know what I'll test when I'll be in Belgium | 19:00 |
| mgariepy | and i only do centos stuff as a side effect of political issue ;) | 19:00 |
| drifterza | so centos/xenial etc.. sorted | 19:00 |
| drifterza | pardon my french but screw centos | 19:01 |
| mgariepy | drifterza, it's not french ;p haha | 19:01 |
| evrardjp | haha drifterza | 19:01 |
| drifterza | I thought evrardjp would like that meem | 19:02 |
| evrardjp | :D | 19:02 |
| evrardjp | mgariepy: do you have the same thing with french ppl or is it just with americans? | 19:02 |
| drifterza | not all french and not all americans | 19:03 |
| drifterza | some are very nice | 19:03 |
| mgariepy | evrardjp, same thing as what ? | 19:04 |
| jrosser | cloudnull: we started looking at multinode-aio / 16.04 | 19:05 |
| jrosser | cobbler is a huge pita in 16.04 | 19:05 |
| cloudnull | cool | 19:05 |
| cloudnull | buh. | 19:05 |
| drifterza | its gonna be cool if we can get powerkvm gates | 19:06 |
| drifterza | always wondered if thats better than x86 | 19:06 |
| cloudnull | jrosser: is it a newer version | 19:06 |
| cloudnull | and all the things are broken? | 19:06 |
| cloudnull | or is it something else? | 19:06 |
| jrosser | cobbler versions is an area of large wtf | 19:07 |
| jrosser | and there isnt a xenial package | 19:07 |
| jrosser | so its conpile / diy / fix mess | 19:07 |
| cloudnull | also i know raddaoui was working on some changes to the build systems . | 19:07 |
| jrosser | i have a playbook that does it | 19:07 |
| logan- | jrosser: i pretty much gave up on running cobbler in ubuntu | 19:07 |
| cloudnull | maybe he has some bits to make it better? | 19:07 |
| cloudnull | idk | 19:07 |
| cloudnull | jrosser: that kinda sux | 19:08 |
| evrardjp | ok I am off for today, I received a game and I didn't play it yet! | 19:08 |
| drifterza | again maas has an api and it rox better than cobbler | 19:08 |
| evrardjp | (for a console :p ) | 19:08 |
| drifterza | later evrardjp have a good one | 19:08 |
| jrosser | yeah it does, beware the cobbler source tgz too | 19:08 |
| evrardjp | not trolling i promise | 19:08 |
| jrosser | they dont match the github branches | 19:08 |
| logan- | haha | 19:09 |
| drifterza | ah shit | 19:10 |
| *** stuartgr has quit IRC | 19:10 | |
| drifterza | andymccr, you there | 19:10 |
| evrardjp | drifterza: it's probably late for him :p | 19:12 |
| drifterza | you still here | 19:13 |
| drifterza | go home dude | 19:13 |
| evrardjp | yes | 19:13 |
| evrardjp | I'm off! | 19:13 |
| evrardjp | you too :p | 19:13 |
| drifterza | lol I am home | 19:13 |
| drifterza | I'm a work-a-holic | 19:13 |
| *** kristian__ has joined #openstack-ansible | 19:14 | |
| *** Jack_Iv has joined #openstack-ansible | 19:15 | |
| *** Mudpuppy_ has joined #openstack-ansible | 19:15 | |
| *** jwitko has quit IRC | 19:15 | |
| *** kristian__ has quit IRC | 19:16 | |
| *** kristian__ has joined #openstack-ansible | 19:16 | |
| dxiri | mmm, any clues as to why one of my computes is doing this? | 19:18 |
| dxiri | 2016-11-23 13:15:30.448 4094 ERROR oslo_service.periodic_task IOError: [Errno 13] Permission denied: '/var/lib/nova/instances/locks/nova-storage-registry-lock' | 19:18 |
| dxiri | can't launch any instances on that host | 19:18 |
| *** Mudpuppy has quit IRC | 19:19 | |
| dxiri | and I just rebooted the server itself, same problem | 19:19 |
| openstackgerrit | Jimmy McCrory proposed openstack/openstack-ansible-galera_server: [WIP] Install MariaDB 10.1 server https://review.openstack.org/401123 | 19:19 |
| andymccr | drifterza: whats up? | 19:19 |
| *** asettle has quit IRC | 19:19 | |
| drifterza | see : https://bugs.launchpad.net/openstack-ansible/+bug/1636017 | 19:19 |
| openstack | Launchpad bug 1636017 in openstack-ansible "no logging for nova* containers unless container is restarted" [Medium,Confirmed] - Assigned to Donovan Francesco (donovan-francesco) | 19:19 |
| *** Mudpuppy_ has quit IRC | 19:19 | |
| drifterza | suprised this has not been caught before | 19:20 |
| andymccr | yeah me too... | 19:22 |
| drifterza | But yeah thats definately happening | 19:22 |
| drifterza | just did a fresh install | 19:22 |
| drifterza | need to look at the rsyslog playbook | 19:23 |
| drifterza | we need to ensure the service is listening on port 514 | 19:23 |
| kristian__ | Hi, Please may someone tell me in what files this ip is located? | 19:24 |
| kristian__ | 172.29.236.100 | 19:24 |
| *** jamesdenton has joined #openstack-ansible | 19:25 | |
| logan- | drifterza: if you run the rsyslog_server role tests does it also break | 19:25 |
| logan- | wonder if its the role breaking or something with the integrated repo configs | 19:26 |
| drifterza | the config files and everything is correct | 19:26 |
| drifterza | but the service isnt being started | 19:26 |
| logan- | ahh | 19:26 |
| logan- | xenial? | 19:26 |
| drifterza | yip | 19:26 |
| drifterza | check here: http://paste.openstack.org/show/590247/ | 19:27 |
| logan- | anything useful in systemctl status for the syslog unit | 19:27 |
| drifterza | full stack installed | 19:27 |
| logan- | https://github.com/openstack/openstack-ansible-rsyslog_server/blob/master/tasks/rsyslog_server_post_install.yml | 19:27 |
| drifterza | http://paste.openstack.org/show/590248/ | 19:27 |
| drifterza | systemd says its running but its not | 19:28 |
| logan- | :/ | 19:28 |
| drifterza | or maybe it needs to be restarted since the config files changed | 19:28 |
| drifterza | so somethings not right with our post plays | 19:28 |
| logan- | let me see if i can repro it real quick without a full integrated build, just using the role tests | 19:29 |
| drifterza | I'm using 14.0.2 | 19:30 |
| drifterza | not sure if it happens on master | 19:30 |
| logan- | ok | 19:30 |
| drifterza | but would be good if you could check too | 19:30 |
| *** jamesdenton has quit IRC | 19:31 | |
| *** schwicht has joined #openstack-ansible | 19:32 | |
| andymccr | drifterza: im guessing nobody uses the rsyslog server stuff - it looks like we havent set up the log shipping properly - or maybe we need to run the rsyslog client role | 19:34 |
| drifterza | Yeah, ill work on a PR to fix this. | 19:35 |
| logan- | confirmed by just running the role test | 19:35 |
| drifterza | doing the same thing logan- ? | 19:35 |
| logan- | yep | 19:35 |
| drifterza | darn | 19:35 |
| logan- | rsyslog is indeed running | 19:35 |
| drifterza | but not listening right | 19:35 |
| logan- | correct | 19:35 |
| andymccr | kristian__: do you have more info on what ip that is? in general all IPs live inside the inventory /etc/openstack_deploy/openstack_inventory.json but that file is generated by taking the definitions in /etc/openstack_deploy/openstack_user_config.yml | 19:35 |
| drifterza | ah sigh | 19:35 |
| drifterza | thanks for confirming | 19:35 |
| drifterza | is this on 14.0.2 or master aswell ? | 19:36 |
| logan- | that was on 14.0.2 | 19:36 |
| drifterza | any chance you can confirm on master ? | 19:36 |
| logan- | yea | 19:36 |
| logan- | one moment | 19:36 |
| drifterza | thanks dude | 19:36 |
| kristian__ | andymccr: so I want to have a aio deployment, eno1 ip is 192.168.0.133 and enp2s0 ip is 192.168.0.121. I would like to have neutron on the enp2s0 interface. Also idk but my floating ips are 192.168.0.200-192.168.0.250 | 19:38 |
| *** victorhe has quit IRC | 19:39 | |
| *** victorhe_ has joined #openstack-ansible | 19:39 | |
| *** Jack_Iv has quit IRC | 19:40 | |
| *** victorhe has joined #openstack-ansible | 19:40 | |
| logan- | occurs on master as well drifterza | 19:42 |
| drifterza | ah crap | 19:42 |
| drifterza | could you update the bug with that info aswell please ? | 19:42 |
| *** victorhe_ has quit IRC | 19:43 | |
| drifterza | cloudnull, even with hypervisor_type set it fails with : LXDAPIException: not found | 19:43 |
| drifterza | cloudnull, do you have a nova.conf I can look at that works ? | 19:44 |
| andymccr | kristian__: ahh i see. it's the AIO - so the AIO is just designed to configure everything for you and setup the interfaces/links etc as required. Since it's an AIO you don't necessarily have to use the physical interfaces etc. The address ranges etc are all pre-configured | 19:44 |
| kristian__ | and when I want to access it from outside including the api? | 19:45 |
| kristian__ | and also access the vms | 19:45 |
| *** jamesdenton has joined #openstack-ansible | 19:46 | |
| drifterza | cloudnull, this is my nova.conf on the lxd node : http://paste.openstack.org/show/590249/ | 19:49 |
| drifterza | see anything I'm missing ? | 19:50 |
| andymccr | kristian__: in general you wouldn't want your API and tenant network on the same network, although you could do that. but if you want, what you could do is (on a clean build) run the bootstrap-aio.sh script, and then adjust the /etc/openstack_deploy/openstack_user_config.yml to reflect the appropriate network range, as well as specify used_ips of .121, .133 and .200-.250 | 19:50 |
| andymccr | that's probably closer to what you're looking at doing | 19:50 |
| kristian__ | andymccr: so just run bootstrap-aio, then change ips in openstack_user_config and deploy | 19:51 |
| *** phalmos has joined #openstack-ansible | 19:52 | |
| dxiri | guys, any guidance on how to set different settings per compute host? for example, what if I want to have different cpu_allocation_ratio values for 2 different compute hosts | 19:53 |
| *** johnmilton has quit IRC | 19:53 | |
| andymccr | kristian__: yeah that is probably the easiest way to do it if you're running an AIO | 19:53 |
| kristian__ | ok | 19:54 |
| andymccr | it'll setup the base of what you need, and then you can adjust the ips in that user_config file | 19:54 |
| andymccr | and then run everything! | 19:54 |
| kristian__ | andymccr: so change only external_lb_vip_address and it should work? | 19:56 |
| kristian__ | plus provider networks | 19:57 |
| drifterza | cloudnull, you gone ? | 19:57 |
| andymccr | kristian__: so change your external_lb_vip_address (and internal_lb_vip_address), as well as the aio1 host ip address references lower down, and the cidr_networks for container and tunnel, you may need to use 2 different subnets for the same network (the 192.168.0.0/24) and then add the ips .131, .122, .200-.250 into the used_ips: section. | 19:58 |
| andymccr | basically any ip references :) except storage which should be fine | 19:58 |
| andymccr | and actually if you have real floating ips | 19:58 |
| andymccr | you wont need to change the tunnel address either | 19:58 |
| andymccr | so just the container network | 19:59 |
| andymccr | but you will need to configure a provider network (probably of type flat) that goes on the linuxbridge_agent containers (maybe just adjust one of the ones there br-vlan for example) | 20:00 |
| drifterza | thanks logan- | 20:00 |
| logan- | ty for confirming the bug :P | 20:01 |
| *** phalmos has quit IRC | 20:02 | |
| drifterza | no problem | 20:03 |
| drifterza | Ill start working on the PR to fix this | 20:03 |
| drifterza | sad part of this is every service that runs needs a restart for logs to start spooling remotely | 20:03 |
| drifterza | I guess thats why rsyslog should be done properly before the rest of the plays | 20:03 |
| drifterza | still have to work out this weirdness with nova-lxd | 20:06 |
| drifterza | sigh | 20:06 |
| drifterza | But I need sleep now ;) | 20:06 |
| andymccr | ok im out for today too :) seeya all. | 20:07 |
| drifterza | ciao guys | 20:07 |
| spotz | seeya drifterza | 20:07 |
| spotz | seeya andymccr | 20:07 |
| *** cmart has quit IRC | 20:08 | |
| *** jamesdenton has quit IRC | 20:09 | |
| *** oanson has quit IRC | 20:09 | |
| *** jmckind_ has joined #openstack-ansible | 20:09 | |
| *** havocbmx-LT has quit IRC | 20:10 | |
| openstackgerrit | Pedro Perez proposed openstack/openstack-ansible-os_cloudkitty: Clean up unused vars https://review.openstack.org/401405 | 20:11 |
| *** jmckind has quit IRC | 20:11 | |
| *** drifterza has quit IRC | 20:11 | |
| *** hybridpollo has joined #openstack-ansible | 20:13 | |
| *** jamesdenton has joined #openstack-ansible | 20:13 | |
| *** Jack_Iv has joined #openstack-ansible | 20:16 | |
| *** jperry has joined #openstack-ansible | 20:19 | |
| *** hybridpollo has quit IRC | 20:19 | |
| kristian__ | andymccr: I got this error | 20:21 |
| kristian__ | FAILED - RETRYING: TASK: lxc_container_create : Wait for ssh to be available (3 retries left). | 20:21 |
| *** allanice001 has joined #openstack-ansible | 20:21 | |
| *** Jack_Iv has quit IRC | 20:22 | |
| *** Jack_Iv has joined #openstack-ansible | 20:22 | |
| *** jmckind_ has quit IRC | 20:22 | |
| *** cmart has joined #openstack-ansible | 20:26 | |
| *** Jack_Iv has quit IRC | 20:27 | |
| *** hybridpollo has joined #openstack-ansible | 20:32 | |
| kristian__ | andymccr: do you know what can be causing it? | 20:34 |
| *** v1k0d3n has quit IRC | 20:46 | |
| *** dxiri has quit IRC | 20:49 | |
| *** dxiri has joined #openstack-ansible | 20:50 | |
| *** gouthamr has quit IRC | 20:50 | |
| *** sdake has joined #openstack-ansible | 20:54 | |
| *** cmart has quit IRC | 20:58 | |
| *** schwicht has quit IRC | 21:04 | |
| -openstackstatus- NOTICE: Due to a problem with the cinder volume backing the log server, jobs are failing with POST_FAILURE. Please avoid issuing 'recheck' commands until the issue is resolved. | 21:05 | |
| *** ChanServ changes topic to "Due to a problem with the cinder volume backing the log server, jobs are failing with POST_FAILURE. Please avoid issuing 'recheck' commands until the issue is resolved." | 21:06 | |
| *** spotz is now known as spotz_zzz | 21:10 | |
| *** jamesdenton has quit IRC | 21:18 | |
| *** Mudpuppy has joined #openstack-ansible | 21:19 | |
| *** phalmos has joined #openstack-ansible | 21:19 | |
| *** karimb has quit IRC | 21:20 | |
| *** karimb has joined #openstack-ansible | 21:20 | |
| *** jamielennox|away is now known as jamielennox | 21:22 | |
| kristian__ | Hi I got this error while deploying http://paste.openstack.org/show/590256/ . Please may someone help me? | 21:23 |
| *** fguillot has quit IRC | 21:24 | |
| *** schwicht has joined #openstack-ansible | 21:26 | |
| *** cmart has joined #openstack-ansible | 21:28 | |
| cloudnull | kristian__: looks like br-mgmt does not exist | 21:28 |
| cloudnull | on the node | 21:28 |
| *** victorhe has quit IRC | 21:29 | |
| cloudnull | drifterza my bad. | 21:29 |
| cloudnull | drifterza hit me up later if lxd still wont go | 21:29 |
| *** allanice001 has quit IRC | 21:30 | |
| kristian__ | cloudnull: http://paste.openstack.org/show/590258/ | 21:30 |
| cloudnull | osa wont create the base network on the physical host for you | 21:31 |
| *** schwicht has quit IRC | 21:31 | |
| cloudnull | the user config is simply telling the system how to wire things up | 21:31 |
| kristian__ | oh so how to do that? | 21:32 |
| cloudnull | that all really depends on your arch | 21:32 |
| cloudnull | we do have some examples thoughj | 21:32 |
| cloudnull | https://github.com/openstack/openstack-ansible/blob/master/etc/network/interfaces.d/openstack_interface.cfg.prod.example | 21:32 |
| kristian__ | its ubuntu 16.04 | 21:32 |
| cloudnull | you dont need all the bonds and such, or even the vlans for that matter. | 21:33 |
| cloudnull | however you will need the bridges. | 21:33 |
| cloudnull | this covers the bridges needed http://docs.openstack.org/developer/openstack-ansible/newton/install-guide/overview-network-arch.html | 21:34 |
| cloudnull | you may want to review http://docs.openstack.org/developer/openstack-ansible/newton/install-guide/targethosts.html | 21:34 |
| cloudnull | more specifically the network diagram here http://docs.openstack.org/developer/openstack-ansible/newton/install-guide/app-config-prod.html#production-environment-config | 21:35 |
| *** Jeffrey4l has quit IRC | 21:35 | |
| cloudnull | kristian__: how many nodes are you deploying? | 21:35 |
| kristian__ | aio | 21:35 |
| *** Jeffrey4l has joined #openstack-ansible | 21:35 | |
| cloudnull | oh. if you just run ./scripts/gate-check-commit.sh | 21:35 |
| cloudnull | it'll do all that for you | 21:35 |
| kristian__ | ok | 21:36 |
| *** f13o has joined #openstack-ansible | 21:36 | |
| cloudnull | if you're looking to learn how it all is put together the quick-start guide can be useful http://docs.openstack.org/developer/openstack-ansible/newton/developer-docs/quickstart-aio.html | 21:37 |
| cloudnull | however if you just need an AIO up and running the script makes it all go. | 21:37 |
| *** schwicht has joined #openstack-ansible | 21:38 | |
| openstackgerrit | Flávio Ramalho proposed openstack/openstack-ansible-os_ironic: Adds support for the ironic-oneview drivers. https://review.openstack.org/400241 | 21:38 |
| cloudnull | and is gerenally what I use to kick a new test node | 21:38 |
| openstackgerrit | Pedro Perez proposed openstack/openstack-ansible: Add Cloudkitty to integrated repo. https://review.openstack.org/397313 | 21:38 |
| *** schwicht has quit IRC | 21:39 | |
| *** jperry has quit IRC | 21:40 | |
| *** schwicht has joined #openstack-ansible | 21:40 | |
| *** weezS has quit IRC | 21:41 | |
| *** victorhe has joined #openstack-ansible | 21:43 | |
| *** cmart has quit IRC | 21:43 | |
| *** thorst_ has joined #openstack-ansible | 21:47 | |
| *** yolanda has quit IRC | 21:50 | |
| *** vnogin has joined #openstack-ansible | 21:51 | |
| *** vnogin has quit IRC | 21:51 | |
| *** vnogin has joined #openstack-ansible | 21:51 | |
| *** schwicht has quit IRC | 21:55 | |
| *** vnogin has quit IRC | 21:59 | |
| kristian__ | cloudnull: Im getting no more hosts left error | 22:01 |
| cloudnull | that's an error coming back from an ansible run . | 22:02 |
| cloudnull | which means something failed. | 22:02 |
| kristian__ | sorry this | 22:02 |
| cloudnull | you will have to go through the scrollback to figure that out | 22:02 |
| kristian__ | http://paste.openstack.org/show/590261/ | 22:02 |
| *** oneswig has joined #openstack-ansible | 22:02 | |
| cloudnull | something failed in the run | 22:03 |
| kristian__ | this is the first error before the NORICE | 22:03 |
| kristian__ | *NOTICE | 22:03 |
| cloudnull | there should be a bash or ansible trace someone above that. | 22:03 |
| kristian__ | ok | 22:04 |
| kristian__ | going to look for it | 22:04 |
| kristian__ | why it is using that bad ip?? | 22:06 |
| kristian__ | http://paste.openstack.org/show/590263/ | 22:06 |
| kristian__ | it looks like it got rewritten :( | 22:06 |
| cloudnull | the AIO script sets up an env with everything known for it. | 22:07 |
| *** vnogin has joined #openstack-ansible | 22:07 | |
| cloudnull | its using the 100 address because its internal only and not plugged into the outside world | 22:07 |
| cloudnull | you can access the aio externally through the default nic | 22:08 |
| cloudnull | haproxy will provide for load balancing | 22:08 |
| kristian__ | yes but it looks like it had a communication problem | 22:09 |
| *** cmart has joined #openstack-ansible | 22:10 | |
| *** v1k0d3n has joined #openstack-ansible | 22:10 | |
| *** oneswig has quit IRC | 22:17 | |
| *** thorst_ has quit IRC | 22:20 | |
| cloudnull | ok. im off, take care all. | 22:21 |
| kristian__ | ok bye | 22:21 |
| *** f13o has quit IRC | 22:27 | |
| *** kristian__ has quit IRC | 22:29 | |
| openstackgerrit | Steve Lewis (stevelle) proposed openstack/openstack-ansible: Inventory Generation rename vars https://review.openstack.org/401460 | 22:32 |
| *** chris_hultin is now known as chris_hultin|AWA | 22:37 | |
| *** agrebennikov has quit IRC | 22:40 | |
| *** ChanServ changes topic to "Launchpad: https://launchpad.net/openstack-ansible || Weekly Meetings: https://wiki.openstack.org/wiki/Meetings/openstack-ansible || Review Dashboard: https://goo.gl/tTmdgs" | 22:50 | |
| -openstackstatus- NOTICE: The affected filesystems on the log server are repaired. Please leave 'recheck' comments on any changes which failed with POST_FAILURE. | 22:50 | |
| *** vnogin has quit IRC | 22:53 | |
| mrhillsman | any ideas why a fresh install of stable/mitaka is returning - https://gist.githubusercontent.com/mrhillsman/20b905ffbfd701b29d90aa3f75c1f6e0/raw/45eefd95d80741f5ef6702e475a0b18fc50e3c23/mitaka-errors.sh | 22:54 |
| *** victorhe has quit IRC | 22:57 | |
| logan- | did you have newton bootstrapped on there at some point? | 22:57 |
| logan- | looks like it is pulling in requirements from rally. does /etc/ansible/roles/os_rally exist? | 22:58 |
| *** vnogin has joined #openstack-ansible | 22:59 | |
| mrhillsman | i deleted it | 23:03 |
| *** Rodrigo_BR has joined #openstack-ansible | 23:03 | |
| mrhillsman | i believe newton was possibly there then completely removed | 23:03 |
| mrhillsman | let me try | 23:03 |
| mrhillsman | that | 23:03 |
| logan- | you might want to wipe all of the roles in /etc/ansible/roles | 23:04 |
| logan- | and run bootstrap-ansible again | 23:04 |
| mrhillsman | thx | 23:04 |
| logan- | yep np | 23:04 |
| Rodrigo_BR | Hi | 23:06 |
| Rodrigo_BR | I am receive intermittent errors during launch of new instances, 2016-11-23 22:43:56,350 - url_helper.py[WARNING]: Calling 'http://169.254.169.254/2009-04-04/meta-data/instance-id' failed [119/120s]: unexpected error ['Timeout' object has no attribute 'response']. sometimes work sometimes fail | 23:06 |
| Rodrigo_BR | I checked the nova-metadata logs e don't see any erros | 23:08 |
| logan- | Rodrigo_BR: sounds like the requests might not even be making it to the metadata server | 23:11 |
| stevelle | more eyeballs desired: https://review.openstack.org/#/q/topic:inventory-refactor+status:open | 23:24 |
| *** dxiri has quit IRC | 23:25 | |
| *** dxiri has joined #openstack-ansible | 23:26 | |
| *** dxiri has quit IRC | 23:33 | |
| *** jlockwood has quit IRC | 23:34 | |
| *** sdake has quit IRC | 23:37 | |
| *** sdake has joined #openstack-ansible | 23:40 | |
| *** sdake has quit IRC | 23:40 | |
| *** sdake has joined #openstack-ansible | 23:40 | |
| *** jamesdenton has joined #openstack-ansible | 23:50 | |
| *** jamesdenton has quit IRC | 23:52 | |
| *** omiday has joined #openstack-ansible | 23:55 | |
| *** omiday has quit IRC | 23:55 | |
| *** omiday has joined #openstack-ansible | 23:55 | |
| openstackgerrit | Jimmy McCrory proposed openstack/openstack-ansible-galera_server: [WIP] Install MariaDB 10.1 server https://review.openstack.org/401123 | 23:55 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!