| *** openstackgerrit has joined #ara | 02:11 | |
| openstackgerrit | Merged recordsansible/ara master: Run actual integration tests for the ara-api role jobs https://review.opendev.org/666894 | 02:11 |
|---|---|---|
| openstackgerrit | David Moreau Simard proposed recordsansible/ara master: Add distributed_sqlite db backend and WSGI app https://review.opendev.org/665323 | 02:12 |
| *** spiette has quit IRC | 10:11 | |
| *** gvincent has joined #ara | 12:05 | |
| *** gvincent has quit IRC | 12:30 | |
| *** metabsd has joined #ara | 12:32 | |
| *** spiette has joined #ara | 14:23 | |
| *** bwatson has joined #ara | 14:35 | |
| bwatson | trying to: pip install ara[server] and running into this: ERROR: Could not find a version that satisfies the requirement Django>=2.1.5; extra == "server" (from ara[server]) | 14:36 |
| bwatson | is this a python2/3 thing (currently using 2.7)? | 14:36 |
| openstackgerrit | David Moreau Simard proposed recordsansible/ara master: Add distributed_sqlite db backend and WSGI app https://review.opendev.org/665323 | 14:36 |
| dmsimard | bwatson: >1.0 requires py3 right now. | 14:37 |
| bwatson | hmmm, OK. I really need to get off my butt and make the jump | 14:38 |
| dmsimard | The API server components absolutely require py3 -- the Ansible plugins might work with py2 but they're untested | 14:38 |
| dmsimard | py2 is EOL next year and since 1.0 is basically rewritten from scratch, it made more sense to be forward looking and future proof | 14:40 |
| bwatson | totally understandable | 14:40 |
| dmsimard | bwatson: however, the new API server and web interface are excellent candidates for containers so you could technically run the py3 api server on a py2 host by running it inside a container | 14:44 |
| dmsimard | ara-web is stateless and expects an API endpoint while all the state for the API server is stored in a database (sqlite/mysql/postgresql) | 14:45 |
| dmsimard | there can definitely be a py2 compatible callback to send results to a remote API server but it hasn't been written yet | 14:45 |
| bwatson | probably a better idea to put on my big boy pants and get with py3 | 15:09 |
| *** vincent- has joined #ara | 15:21 | |
| *** vincent- has quit IRC | 15:22 | |
| *** vincent- has joined #ara | 15:22 | |
| Marbug | Hi, I'm thinking on using the ara_record modules in one of the python modules I wrote for ansible. Is there a way to import the ara_record part to be able to push information directly from the module? (I'm not a python expert btw :)) | 15:34 |
| Marbug | I suppose I can also check which version of ara is used, and in case of a difference in version between ara-0.16.1 and ara-1.0 I could do a switch in my module to use the correct version | 15:40 |
| *** sshnaidm is now known as sshnaidm|afk | 15:40 | |
| dmsimard | Marbug: I'm not sure I understand the question, can you tell me more about the use case ? | 15:43 |
| Marbug | Well dmsimard I wrote a module that will gather some facts of the environment, and in a 2nd task I use ara_record to push a few records with the information I gathered in the previous custom module I wrote. Now I'm wondering if I can't just call ara commands in python to be able to push those records directly | 15:45 |
| dmsimard | ara_record is more or less a wrapper that recovers the current playbook id and then works with the record API endpoint to post data | 15:46 |
| dmsimard | You could technically write your own module that uses the API | 15:47 |
| Marbug | so I can actually reuse some code from the ara_record module, and integrate that into my own right ? | 15:47 |
| Marbug | perfect :) | 15:47 |
| dmsimard | Well, that's part of why the new API exists | 15:47 |
| dmsimard | Although if there's something missing from ara_record, I'd rather we improve it | 15:48 |
| Marbug | will need to get my ara-0.16 back to work first, as for some strange reason ara_record is throwing me "ARA is required to run this module" while it's installed through pip, so will need to fix that first | 15:48 |
| Marbug | it's not that it's missing, butI want to still give support between the versions | 15:48 |
| dmsimard | The way it works in 0.x and 1.x is different | 15:50 |
| Marbug | and since environmlents might be different, we can't just add the plugin/library location of the system to the config, as the differ between systems, so for ara-0.16 we added the action and call back plugins to the main repo of ansible (which requires ara to be installed) | 15:50 |
| Marbug | as I assumed, and that was why I do want to write those pieces into the module I wrote to gather the facts | 15:50 |
| dmsimard | In 0.x the id of the playbook is cached by the callback while in 1.x, we retrieve the playbook id by looking up the play and then finding the parent playbook | 15:50 |
| Marbug | so for people using 0.16, it'll still push the info to 0.16, and for people or systems using the newer version, will jsut use the newer version | 15:51 |
| Marbug | so, I would think, it might be technically possible ? | 15:51 |
| Marbug | to combine it into a module supporting older and newer versions for our use case? | 15:52 |
| dmsimard | Anything is possible, just a matter of spending time on it :) | 15:52 |
| Marbug | true :) | 15:53 |
| dmsimard | If you have questions along the way I can help | 15:54 |
| Marbug | so let me get my old version back to work, getting the gathering module working for python 2+3 at the same time, and then I can continue on getting ara-1.0 to work :) | 15:55 |
| Marbug | sure, thanks, I'll be back later in case I get really stuck :) | 15:55 |
| dmsimard | ++ | 16:00 |
| *** zbr|ruck is now known as zbr | 17:18 | |
| *** altlogbot_2 has quit IRC | 17:55 | |
| *** altlogbot_2 has joined #ara | 17:55 | |
| *** altlogbot_2 has quit IRC | 17:57 | |
| *** altlogbot_3 has joined #ara | 17:58 | |
| *** altlogbot_3 has quit IRC | 18:01 | |
| *** altlogbot_1 has joined #ara | 18:02 | |
| *** altlogbot_1 has quit IRC | 19:45 | |
| *** altlogbot_3 has joined #ara | 19:46 | |
| *** altlogbot_3 has quit IRC | 20:15 | |
| *** altlogbot_2 has joined #ara | 20:18 | |
| *** dbpiv has joined #ara | 20:18 | |
| openstackgerrit | David Moreau Simard proposed recordsansible/ara master: Add distributed_sqlite db backend and WSGI app https://review.opendev.org/665323 | 20:34 |
| openstackgerrit | David Moreau Simard proposed recordsansible/ara master: ara_api role: Update selinux for gunicorn from home directory https://review.opendev.org/667724 | 20:34 |
| openstackgerrit | David Moreau Simard proposed recordsansible/ara master: ara_api role: Stop managing the gunicorn pidfile https://review.opendev.org/667725 | 20:34 |
| *** altlogbot_2 has quit IRC | 20:43 | |
| *** altlogbot_2 has joined #ara | 20:44 | |
| *** altlogbot_2 has quit IRC | 21:00 | |
| *** altlogbot_3 has joined #ara | 21:04 | |
| openstackgerrit | David Moreau Simard proposed recordsansible/ara master: Add distributed_sqlite db backend and WSGI app https://review.opendev.org/665323 | 21:25 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!