16:06:19 #startmeeting Ansible Network Working Group 16:06:19 Meeting started Wed Jul 3 16:06:19 2019 UTC. 16:06:19 This meeting is logged and archived in a public location. 16:06:19 The chair is pabelanger. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:06:19 Useful Commands: #action #agreed #halp #info #idea #link #topic. 16:06:19 The meeting name has been set to 'ansible_network_working_group' 16:06:30 #chair Qalthos 16:06:30 Current chairs: Qalthos pabelanger 16:06:37 #topic Core Updates 16:06:58 I believe dot releases are still on track for this week 16:07:09 which means 2.8.2 should be landing any time 16:08:24 It is also US holiday tomorrow, July 4th, and was Canadian holiday on Monday (July 1st), so expect a lighter then usually week for activities 16:08:49 But expect things to settle back into regular routine next week 16:09:35 Agenda for today can be found at: 16:09:36 #link https://github.com/ansible/community/labels/network 16:09:46 Lets get started 16:10:34 #topic onyx 16:10:38 #link https://github.com/ansible/ansible/pull/57066 16:11:13 This looks to be in good order, I can see test coverage which is always good 16:11:30 It is just a matter of getting some more eyes on for reviews 16:11:46 I don't think samerd is around 16:12:12 Oh, I also am looking at the wrong section 16:12:31 #topic Meraki modules 16:12:49 We talked last week about the httpapi questions I had. Were you able to discuss with the team at all? 16:12:50 #link https://github.com/ansible/ansible/pull/52889 16:12:54 #link https://github.com/ansible/ansible/pull/57289 16:13:00 #link https://github.com/ansible/ansible/pull/57855 16:13:07 #link https://github.com/ansible/ansible/pull/55485 16:13:22 So, I owe mrproper a round of reviews, I wasn't able to do that last week 16:13:32 I've made a note to schedule some time tomorrow 16:13:47 No problem. The ones I listed here were the same as last week; nothing new. 16:13:49 as for httpapi, yes raised the topic with team 16:14:17 mrproper: now that Qalthos is back from PTO, I know he expressed interest in digging deeper into your feedback 16:14:25 and know in the past you have worked together 16:14:31 We had, yes. 16:15:04 What do you recommend is the best way to get in touch with him to discuss? 16:15:20 mrproper: I think it is just a matter of time before Qalthos gets to speed on issue, and work around upcoming holiday 16:15:39 mrproper: regarding authentication, it isn't the greatest fit, but you could have the auth key as a httpapi plugin option and just ignore username/password 16:16:10 Qalthos: How does the user specify the httpapi plugin option? Is that inventory? 16:16:51 However you define it in the plugin. Could be inventory (or any other var), could be env var or config too 16:17:39 Let me get you some examples 16:17:41 Please. 16:18:49 the eos httpapi plugin defines an option to use sessions or not, https://github.com/ansible/ansible/blob/devel/lib/ansible/plugins/httpapi/eos.py#L17 16:19:37 I think that would be a good way to do it. 16:19:46 the `env` key is what env vars it looks for, and the `vars` key are the vars it will examine 16:20:47 There's other options, too https://docs.ansible.com/ansible/latest/dev_guide/developing_plugins.html#plugin-configuration-documentation-standards 16:20:53 Given the option, I'd rather not have any sort of inventory file requirement within Meraki. The hostname is almost always static and the username/pass is a simple API key. 16:22:56 You probably won't /need/ one, but it would probably be helpful to allow for it for people more used to the Ansible way of specifying things 16:23:31 this should give you the flexibility for both, I think 16:23:53 Fair. I'm happy to provide both options, but don't want inventory files to be a requirement. 16:24:57 yah, both would be great I think 16:25:15 I'll try to implement it this way and see how it works. Qlathos, if I have problems, mind if I contact you? If so, here or or Slack? 16:26:02 You're more likely to get a response here 16:26:13 Yah, I tend to follow IRC more 16:26:20 so, +1 for here 16:26:22 IRC it is. 16:26:45 That's all for me on httpapi. pabelanger, you said you'll take a look at the PRs on the side? 16:26:59 Yup, I'll make time this week to review 16:27:09 Excellent, I think that's all for me then. 16:27:19 great 16:27:24 lets move on to next topic 16:27:36 #topic avi modules 16:27:43 #link https://github.com/ansible/ansible/pull/57116 16:27:49 #link https://github.com/ansible/ansible/pull/58382 16:27:56 #link https://github.com/ansible/ansible/pull/58667 16:28:12 chaitanyaavi: around? 16:28:16 Yes 16:28:20 great 16:28:57 for 57116, it looks like you've address all the comments, so we likely need another round of reviews there 16:29:19 ok 16:29:24 also great to see there are tests 16:29:50 58382 looks straightforward too, I'll add to list of PRs to review 16:29:52 yes, these tests would also clarify the role of the common apis for all avi modules 16:30:53 chaitanyaavi: for 58667, how did you test the new code? Is there any existing avi integration tests for ansible-test? 16:33:19 We do it in our internal jenkins as there is not much code here we can add mocs if that helps 16:34:31 Yah, not sure how it works for lookup plugins to be honest, would need to look myself. By adding some sort of test, even mock in this case, is helpful to test code path 16:34:49 but, I'll add it to review and make leave a comment about that 16:35:15 ok sure 16:35:28 great 16:35:32 moving on 16:35:38 #topic Open discussion 16:36:00 we didn't have anything else on the agenda, so if anybody would like to raise a topic, we can do so now. 16:36:06 Or look to wrap up early 16:39:47 thanks everybody, see you next week 16:39:50 #endmeeting