| *** sshnaidm is now known as sshnaidm|afk | 00:22 | |
| *** zeestrat has quit IRC | 00:46 | |
| *** zeestrat has joined #openstack-ansible-sig | 00:47 | |
| *** openstackgerrit has quit IRC | 01:00 | |
| *** spatel has joined #openstack-ansible-sig | 02:32 | |
| *** spatel has quit IRC | 02:40 | |
| *** spatel has joined #openstack-ansible-sig | 03:09 | |
| *** spatel has quit IRC | 04:53 | |
| *** evrardjp has quit IRC | 05:33 | |
| *** evrardjp has joined #openstack-ansible-sig | 05:33 | |
| *** cyberpear has quit IRC | 07:05 | |
| *** frenzy_friday has joined #openstack-ansible-sig | 08:40 | |
| *** tremble has quit IRC | 09:11 | |
| *** tremble has joined #openstack-ansible-sig | 09:13 | |
| JqckB | sshnaidm|afk: hi, is it possible for you to review this ? https://review.opendev.org/c/openstack/ansible-collections-openstack/+/767719 | 09:34 |
|---|---|---|
| *** gtema has joined #openstack-ansible-sig | 09:44 | |
| *** tremble has quit IRC | 10:02 | |
| *** tremble has joined #openstack-ansible-sig | 10:06 | |
| frenzy_friday | Hey, does anyone know why this is failing https://zuul.opendev.org/t/openstack/build/a9df2b7076da45649ff2c4f639af8b04 ? What am I missing here? | 10:33 |
| *** gtema has quit IRC | 10:47 | |
| *** sshnaidm|afk is now known as sshnaidm | 11:17 | |
| sshnaidm | frenzy_friday, it's a glance failure in devstack install, not related to the patch | 11:20 |
| sshnaidm | frenzy_friday, you can run "recheck" and see if it happens again | 11:20 |
| sshnaidm | frenzy_friday, just post "recheck" comment and jobs will run again | 11:21 |
| sshnaidm | it happens sometimes :) | 11:21 |
| sshnaidm | JqckB, yeah, I looked at it now | 11:22 |
| sshnaidm | JqckB, and it's tricky | 11:22 |
| sshnaidm | JqckB, running ansible task you expect to get desired state | 11:22 |
| sshnaidm | JqckB, so if you configured in task flavor A it's what you should get in the end. If you don't want it to be recreated/updated - just don't configure it. We can't configure the flavor and then to get flavor B instead. | 11:24 |
| sshnaidm | Then when you configure flavor A, don't set "update..." option, running task, it finishes ok, and you're totally sure you've got your flavor A in place. And with this patch it won't happen. | 11:25 |
| sshnaidm | JqckB, maybe you need dry-run instead? | 11:27 |
| frenzy_friday | sshnaidm, like replying to the commit? | 11:36 |
| sshnaidm | frenzy_friday, yes | 11:36 |
| *** tremble has quit IRC | 12:29 | |
| *** tremble has joined #openstack-ansible-sig | 12:34 | |
| *** tremble has quit IRC | 13:00 | |
| *** tremble has joined #openstack-ansible-sig | 13:05 | |
| *** iurygregory has joined #openstack-ansible-sig | 13:16 | |
| *** tremble has quit IRC | 13:31 | |
| *** tremble has joined #openstack-ansible-sig | 13:33 | |
| *** tremble has quit IRC | 13:59 | |
| *** tremble has joined #openstack-ansible-sig | 14:04 | |
| *** tremble has quit IRC | 14:55 | |
| *** tremble has joined #openstack-ansible-sig | 14:55 | |
| *** openstackgerrit has joined #openstack-ansible-sig | 15:05 | |
| openstackgerrit | Emilien Macchi proposed openstack/ansible-collections-openstack master: ports: add tags support https://review.opendev.org/c/openstack/ansible-collections-openstack/+/767219 | 15:05 |
| JqckB | sshnaidm : thanks for the review, my take on this is that ok you should get the desired state, but only if it is possible. The API doesn't allow update, and I would be OK for this comportement if the API allowed update, but that is not the case, and delete/create is not something we should do in production. | 15:17 |
| JqckB | sshnaidm: for you, it is normal that the Ansible module do an "update", by deleting a ressource without telling the user that it will be an delete/create ? | 15:18 |
| JqckB | I can propose that the parameter I proposed should not be default, and keep the current behavior but with a possibility to fail if wanted by the user (that everyone with a production should have actually, this action shouldn't be made). | 15:20 |
| sshnaidm | JqckB, well, ansible doesn't care much *how* the state should be achieved, but it should be. You can't finish with "ok" when configured state if different from the real one. If flavor can't be updated - this should be solved in SDK, not here. If you suspect it may be deleted - you can run with --dry-run and to check this. | 15:24 |
| sshnaidm | JqckB, I'd suggest to try to solve it in SDK anyway and enable flavor to be updated if it's possible | 15:25 |
| JqckB | the API (the Nova API) doesn't handle it: https://docs.openstack.org/api-ref/compute/#list-flavors | 15:27 |
| JqckB | that is why I'm quite suprised that a module should not allow this | 15:27 |
| sshnaidm | JqckB, ansible module is not the place to solve it, but... | 15:29 |
| sshnaidm | JqckB, I can think about a different "solution" | 15:29 |
| sshnaidm | It's not so good though and I'm not sure need to go with it | 15:29 |
| sshnaidm | But you can fail the module if update_flavor is false and need actually to remove the flavor | 15:30 |
| sshnaidm | if flavor update requires removing flavor and update_flavor=false - fail, tell something "Task requires flavor deletion, and recreate_flavor is false" | 15:30 |
| sshnaidm | maybe worth to fail only if there are resources with this flavor | 15:31 |
| sshnaidm | it's kind of hackish and maybe requires a discussion | 15:31 |
| JqckB | You propose to reverse my proposal, keep the current behavior and propose the "recreate_flavor" param that if false, will fail ? | 15:34 |
| JqckB | Am I understood right ? | 15:34 |
| sshnaidm | maybe other option is to have "recreate_flavor" as true, when you can set it as "false" in your task, but then task will fail in case it need to recreate flavor | 15:40 |
| JqckB | If I'm taking the os_server module for example, and I'm looking the code right : If the server already exist, state is present, and image is changed on the image module parameter, you don't recreate the instance :) | 15:41 |
| JqckB | sshnaidm: yes, perfect I understand your proposition. I will propose a change and submit it. | 15:41 |
| sshnaidm | JqckB, good point about server, need to change it.. | 15:44 |
| sshnaidm | I'm going to prepare a release of openstack modules before a new year, please be aware | 16:18 |
| *** cyberpear has joined #openstack-ansible-sig | 17:47 | |
| openstackgerrit | Polina Gubina proposed openstack/ansible-collections-openstack master: Enable update for recordset and add tests for dns and recordset module https://review.opendev.org/c/openstack/ansible-collections-openstack/+/766528 | 19:26 |
| openstackgerrit | Polina Gubina proposed openstack/ansible-collections-openstack master: Enable update for recordset and add tests for dns and recordset module https://review.opendev.org/c/openstack/ansible-collections-openstack/+/766528 | 21:05 |
| *** redrobot has quit IRC | 23:43 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!