| *** sshnaidm|rover is now known as sshnaidm|afk | 00:56 | |
| dmsimard | weshay, tristanC: btw https://review.openstack.org/#/c/580256/ | 03:05 |
|---|---|---|
| ara-slack | <dmsimard> @logan.attwood I've been thinking about this a bit, I think it's still a sane default to use compression but I want to support your use case too. I'll figure something out. | 03:07 |
| tristanC | dmsimard: shouldn't ara module be part of ansible instead? | 03:08 |
| dmsimard | tristanC: we could have the same discussion for the callback | 03:09 |
| tristanC | dmsimard: i mean the callback are setup in ansible.cfg, it's site specifics | 03:09 |
| tristanC | dmsimard: but using module in playbook means that the playbook won't run without the right ansible.cfg... | 03:10 |
| dmsimard | tristanC: I'm not sure I follow | 03:10 |
| dmsimard | tristanC: action plugins and callback plugins are not different in scope, they're both executed locally on the control node (executor) | 03:10 |
| tristanC | dmsimard: in the context of zuul, if you add ara module to job's playbook, then doesn't it require user to install ara in order to re-use the job's playbook? | 03:11 |
| dmsimard | tristanC: oh I see what you mean | 03:12 |
| tristanC | but yeah, ideally ara should be part of ansible ;-) | 03:13 |
| dmsimard | tristanC: this only allows ara_record/ara_read to be used by the executor -- it only becomes "mandatory" if people start using ara_record/ara_read in zuul-jobs or something like that | 03:13 |
| tristanC | dmsimard: executor only execute playbooks from zuul jobs | 03:15 |
| tristanC | seems like this will make re-usable zuul-jobs outside of the zuul context even more difficult if you need to also install and setup ara | 03:16 |
| tristanC | maybe those ara module could be contributed to ansible, and be noop if the callback isn't installed? | 03:17 |
| dmsimard | tristanC: it only becomes a hard dependency if people start contributing ara_record tasks to zuul-jobs which we have lived without so far just fine | 03:17 |
| dmsimard | but it also makes it available for end users if they want to use it in their project jobs | 03:17 |
| dmsimard | I can see both ways but I'm biased :p | 03:18 |
| dmsimard | do you know of people that deploy zuul without ara ? just curious | 03:18 |
| tristanC | no, but i know people using ansible without ara, and those will have a hard-time re-use zuul-jobs playbook relying on ara module | 03:19 |
| dmsimard | why ? this is something provided by zuul | 03:20 |
| tristanC | i mean, the (yet to be defined) zuul-runner tool to reproduce zuul-jobs locally could also setup ara, but that's an extra step to an already tricky process | 03:20 |
| dmsimard | zuul-jobs doesn't depend on ara, my patch doesn't make it depend on ara either | 03:20 |
| dmsimard | oh, okay, that's the part that I was missing | 03:21 |
| dmsimard | tristanC: well that's solved by adding ara to zuul's req.txt? /shrug | 03:21 |
| dmsimard | wasn't my expectation to push for a tighter integration with that patch | 03:22 |
| dmsimard | but I understand the concern | 03:22 |
| dmsimard | I came across the "issue" by trying to use ara_record in a zuul job and realized it wasn't available, that's all :) | 03:22 |
| dmsimard | for me it was just a free feature we weren't offering users, nothing more | 03:23 |
| tristanC | this wouldn't be a concern if ara_record was part of ansible | 03:23 |
| tristanC | well that "free freature" also means tighter integration with ara | 03:23 |
| dmsimard | I haven't considered what should be upstreamed in Ansible since we initially discussed it at the project inception | 03:24 |
| dmsimard | we decided against upstreaming things back then, mostly due to preference in the openstack contribution and testing model | 03:25 |
| dmsimard | It's probably a good idea to revisit for 1.0 | 03:25 |
| tristanC | dmsimard: i would vote for ara being part of ansible default callbacks/modules, it makes play analysis so much easier ;-) | 03:26 |
| dmsimard | tristanC: it didn't go over too well with molecule trying to do it recently | 03:27 |
| dmsimard | https://github.com/ansible/ansible/issues/38200 | 03:27 |
| dmsimard | I think the general idea being that it's okay that Ansible has great "external" unsupported unofficial projects, they are linking to them somewhere | 03:28 |
| dmsimard | https://docs.ansible.com/ansible/latest/community/other_tools_and_programs.html | 03:29 |
| openstackgerrit | David Moreau Simard proposed openstack/ara master: Prevent exception when LOG_FILE is empty https://review.openstack.org/574669 | 03:32 |
| openstackgerrit | David Moreau Simard proposed openstack/ansible-role-ara master: WIP: Zuul jobs experiments https://review.openstack.org/557933 | 03:33 |
| openstackgerrit | David Moreau Simard proposed openstack/ansible-role-ara master: WIP: Zuul jobs experiments https://review.openstack.org/557933 | 05:23 |
| openstackgerrit | Merged openstack/ara master: Prevent exception when LOG_FILE is empty https://review.openstack.org/574669 | 05:34 |
| *** sshnaidm|afk has quit IRC | 07:14 | |
| *** sshnaidm|afk has joined #ara | 07:21 | |
| *** jparrill has joined #ara | 07:23 | |
| *** sshnaidm|afk is now known as sshnaidm|rover | 08:16 | |
| *** quiquell has joined #ara | 08:45 | |
| quiquell | Hello | 08:45 |
| quiquell | I am getting a 'Flask' object has no attribute '_cache' | 08:45 |
| quiquell | And then a POST timeout at a job I have configured with ara | 08:46 |
| quiquell | http://logs.openstack.org/90/580090/7/check/tripleo-ci-centos-7-undercloud-containers/9dc84d3/job-output.txt.gz#_2018-07-04_18_10_01_532311 | 08:46 |
| quiquell | join #tripleo-ci | 09:39 |
| *** sshnaidm|rover is now known as sshnaidm|rov|lnc | 11:33 | |
| *** sshnaidm|rov|lnc is now known as sshnaidm|rover | 12:15 | |
| *** quiquell is now known as quiquell|lunch | 12:27 | |
| *** quiquell|lunch is now known as quiquell | 12:52 | |
| dmsimard | quiquell: hey, just woke up. I'll take a look asap. | 13:18 |
| *** bcoca has joined #ara | 13:47 | |
| *** bcoca has joined #ara | 13:47 | |
| *** quiquell is now known as quiquell|mtg | 13:53 | |
| *** jparrill has quit IRC | 13:55 | |
| ara-slack | <logan.attwood> @dmsimard fwiw, postgres does compression w/ the TOAST table | 13:58 |
| *** themurph has joined #ara | 14:02 | |
| *** tbielawa has joined #ara | 14:30 | |
| *** tbielawa is now known as tbielawa|mtg | 14:38 | |
| *** spiette has joined #ara | 14:42 | |
| *** jparrill has joined #ara | 14:53 | |
| *** spiette has quit IRC | 14:53 | |
| *** spiette has joined #ara | 14:58 | |
| *** quiquell|mtg is now known as quiquell|off | 15:26 | |
| *** tbielawa|mtg is now known as tbielawa | 15:26 | |
| *** tbielawa is now known as tbielawa|brb | 15:33 | |
| *** tbielawa|brb is now known as tbielawa | 15:44 | |
| dmsimard | quiquell|off: I see what's happening but I have no idea how you are running into that. | 16:45 |
| *** tbielawa is now known as tbielawa|mtg | 17:04 | |
| *** tbielawa|mtg is now known as tbielawa | 17:12 | |
| *** sshnaidm|rover has quit IRC | 17:43 | |
| *** sshnaidm|rover has joined #ara | 17:52 | |
| *** tbielawa is now known as tbielawa|mtg | 17:55 | |
| *** resmo has joined #ara | 18:01 | |
| dmsimard | quiquell|off: so it looks like you have a report here: http://logs.openstack.org/90/580090/9/check/tripleo-ci-centos-7-undercloud-containers/33d2ac2/logs/undercloud/home/zuul/ara-report/ | 18:16 |
| *** tbielawa|mtg is now known as tbielawa|lunch | 18:19 | |
| openstackgerrit | David Moreau Simard proposed openstack/ara master: Instanciate the application inside the callback https://review.openstack.org/580465 | 18:30 |
| openstackgerrit | David Moreau Simard proposed openstack/ara master: Instanciate the application inside the callback https://review.openstack.org/580465 | 18:45 |
| *** sshnaidm|rover is now known as sshnaidm|off | 18:46 | |
| *** tbielawa|lunch has quit IRC | 19:10 | |
| *** tbielawa has joined #ara | 19:10 | |
| *** resmo has quit IRC | 20:04 | |
| *** themurph has quit IRC | 20:35 | |
| *** tbielawa is now known as tbielawa|g0n3 | 20:47 | |
| *** tbielawa|g0n3 has quit IRC | 20:47 | |
| *** resmo has joined #ara | 21:22 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!