| *** harlowja has joined #ara | 00:26 | |
| *** harlowja has quit IRC | 00:27 | |
| weshay | tristanC, hrm.. that's interesting, I will be in touch about that :) | 01:38 |
|---|---|---|
| tristanC | weshay: well, instead of doing manual "ara generate html", just copy the ansible.sqlite in a ara-report folder in the logs artifacts | 01:43 |
| tristanC | e.g. ara/undercloud/ara-report/ansible.sqlite and ara/overcloud/ara-report/ansible.sqlite | 01:43 |
| tristanC | then the middleware will take care of rendering the webinterface when you access any ara-report folder containing that ansible.sqlite file | 01:44 |
| *** evrardjp has quit IRC | 03:38 | |
| *** evrardjp has joined #ara | 03:39 | |
| *** bcoca has quit IRC | 03:47 | |
| *** evrardjp has quit IRC | 03:50 | |
| *** evrardjp has joined #ara | 03:51 | |
| *** jparrill has quit IRC | 07:37 | |
| *** jparrill has joined #ara | 07:39 | |
| *** guilhermesp has left #ara | 13:59 | |
| *** sshnaidm|rover is now known as sshnaidm|afk | 14:11 | |
| *** sshnaidm|afk is now known as sshnaidm|rover | 15:38 | |
| openstackgerrit | David Moreau Simard proposed openstack/ansible-role-ara master: WIP: Zuul jobs experiments https://review.openstack.org/557933 | 16:06 |
| ara-slack | <logan.attwood> i'm down to try it out, what would the pip command look like? | 17:46 |
| ara-slack | <dmsimard> @logan.attwood everything that was required should have merged so "pip install git+https://github.com/openstack/ara" | 18:27 |
| ara-slack | <dmsimard> That should hopefully make it so your machine doesn't run out of ram when browsing the reports in the web interface | 18:43 |
| ara-slack | <dmsimard> (pretty convenient) | 18:43 |
| ara-slack | <logan.attwood> would this make it possible to use the sqlalchemy `JSON` type instead of `TEXT` where json support exists? http://docs.sqlalchemy.org/en/latest/core/custom_types.html#sqlalchemy.types.TypeDecorator.with_variant | 19:40 |
| ara-slack | <dmsimard> @logan.attwood for postgre or something like that ? | 20:04 |
| ara-slack | <logan.attwood> apparently newer versions of mysql support it as well | 20:04 |
| ara-slack | <dmsimard> I would tend to say yes, but actually the new backend (and API) based on django/django-rest-framework is actually almost ready for early adoption so I'm trying to keep the current branch maintenance to a minimum | 20:05 |
| ara-slack | <dmsimard> I'm not sure how the django model represents those (it's not sqlalchemy) so I'd rather double check that we're doing the right thing with django instead | 20:06 |
| ara-slack | <logan.attwood> my main interest is in hooking up metabase to it, tbh | 20:06 |
| ara-slack | <logan.attwood> and pushing filtering logic down to the database in json fields | 20:06 |
| ara-slack | <logan.attwood> oh jeeze | 20:07 |
| ara-slack | <logan.attwood> probably could do that to get native uuid support | 20:07 |
| ara-slack | <dmsimard> not familiar with metabase but the API has built-in support for json obviously :P | 20:08 |
| ara-slack | <dmsimard> there's an integration test that I've been showing as a demo of the new backend/API if you'd like to have a look | 20:08 |
| ara-slack | <dmsimard> metabase looks pretty cool | 20:09 |
| ara-slack | <logan.attwood> this is a bit of a bummer :S https://github.com/openstack/ara/blob/master/ara/models.py#L382 | 20:20 |
| ara-slack | <dmsimard> @logan.attwood so 1.0 has something similar but I'm open to suggestions, the general idea being that text (or serialized json) compresses VERY well and keeps the database size in check | 20:43 |
| ara-slack | <dmsimard> keeping things uncompressed makes things perhaps easier to poke at but at the cost of storage | 20:45 |
| ara-slack | <dmsimard> the compression ratio (for the amount of compute it requires) is pretty crazy. | 20:45 |
| *** jparrill has quit IRC | 20:59 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!