16:00:58 <akasurde> #startmeeting Ansible VMware Working Group Meeting 16:00:58 <zodbot> Meeting started Mon Jul 31 16:00:58 2017 UTC. The chair is akasurde. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:00:58 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic. 16:00:58 <zodbot> The meeting name has been set to 'ansible_vmware_working_group_meeting' 16:01:58 <DMarby> akasurde: Hi :) 16:02:06 <akasurde> #chair DMarby 16:02:06 <zodbot> Current chairs: DMarby akasurde 16:02:11 <akasurde> DMarby, Hi :) 16:02:21 <akasurde> #chair jtanner 16:02:21 <zodbot> Current chairs: DMarby akasurde jtanner 16:03:41 <dericcrago> hi akasurde 16:03:46 <akasurde> Let us wait for other people to join in 16:03:51 <akasurde> dericcrago, hi 16:03:55 <akasurde> #chair dericcrago 16:03:55 <zodbot> Current chairs: DMarby akasurde dericcrago jtanner 16:05:14 <akasurde> whatz up everyone ? 16:05:48 <DMarby> akasurde: Want to start by going over the PRs? 16:05:59 <akasurde> DMarby, sure. 16:06:13 <akasurde> #topic Module needs review 16:06:20 <akasurde> #link https://github.com/ansible/community/issues/206#issuecomment-318256386 16:07:44 <DMarby> #24841 seems fine since the reverted the import changes I think? 16:08:41 <akasurde> jtanner, wants to have test included to ensure change doesn't break anything 16:09:42 <DMarby> Fair enough 16:10:08 <akasurde> let us move to next one then 16:10:33 <akasurde> #link https://github.com/ansible/ansible/pull/25944 16:10:49 <DMarby> LGTM 16:11:01 * jtanner is here now 16:11:40 <akasurde> jtanner, hi 16:13:36 <akasurde> DMarby, can you add your shipit there? 16:13:45 <DMarby> Sure 16:13:57 <jtanner> won't help if you don't fix the description 16:14:49 <jtanner> bad description -> needs_template -> !shipit + !automerge 16:14:55 <DMarby> Ah, I see 16:15:19 <akasurde> Ohk 16:16:08 <akasurde> #action akasurde Add note in PR for correct description 16:16:16 <akasurde> next PR 16:16:18 <jtanner> bot has already mentioned it 16:16:26 <DMarby> Alright, #25857, I tested this since I needed it yesterday, and it works great on 6.5. 16:16:57 <akasurde> #link https://github.com/ansible/ansible/pull/25857 16:17:04 <akasurde> DMarby, Cool 16:19:01 <akasurde> Shall we move on to next PR ? 16:19:17 <jtanner> well .. 16:19:31 <jtanner> the two maintainers pinged in that PR are absent, iirc 16:19:40 <jtanner> so i don't know who is going to review it 16:21:22 <akasurde> jtanner, there is very little active from these two maintainers, what can we related to this ? 16:21:55 <jtanner> i don't know what is "related", but the implication is that they are never going to review and or shipit 16:23:37 <akasurde> can't we just consider community review ? 16:24:01 <akasurde> like DMarby used it and found PR working 16:24:25 <jtanner> possibly, but there's not shipits in the PR 16:24:41 <akasurde> OK 16:24:44 <jtanner> so the bot can't begin to expand the scope 16:25:30 <akasurde> Ok 16:26:14 <akasurde> Let us wait for community shipit then 16:26:44 <akasurde> Let us move on to next PR then 16:28:08 <akasurde> #link https://github.com/ansible/ansible/pull/26323 16:33:00 <akasurde> I am getting Service Outage message from Github 16:33:08 <jtanner> yeah, it's having a rough time right now 16:35:32 <akasurde> Let us give them some time to recover 16:39:14 <jtanner> akasurde: you should move on to higher level topics 16:39:32 <akasurde> Ok 16:39:45 <jtanner> github is probably not going recover before meeting is over 16:40:10 <akasurde> jtanner, Let us discuss about vmware_guest state and powerstate then 16:40:28 <akasurde> #topic vmware_guest state and powerstate 16:41:19 <akasurde> For everyone else in meeting, vmware_guest does not differentiate between state of VM and powerstate 16:41:28 <jtanner> problem: vmware_guest tries to encapsulate and infer too many things from the "state" parameter 16:42:43 <akasurde> jtanner, exactly, so to solve this problem we need to have additional parameter to make vmware_guest understand differences 16:42:45 <jtanner> does anyone have thoughts about ways to fix that? 16:42:54 <DMarby> Right, with how it only changes configuration on "present" and not on poweredon/etc, for example? 16:43:37 <jtanner> yes, a good example 16:44:13 <akasurde> for example, if user wants to poweroff a machine and accidently uses incorrect machine name which does not exists then vmware_guests tries to deploy it 16:44:19 <jtanner> my personal belief is that there should be separate modules for each part of the state 16:44:53 <akasurde> same here 16:45:03 <jtanner> i suggested this morning that a stopgap could be to add a parameter to vmware_guest to clarify the intent 16:45:07 <jtanner> clone: true/false 16:45:29 <dericcrago> I'd like to see that too, but in the short term I think just adding a powerstate parameter would help 16:45:34 <jtanner> action: clone|power|etc 16:45:49 <DMarby> How does the AWS module handle it? 16:46:16 <jtanner> theoretically, you destroy and build new in a real cloud 16:46:29 <akasurde> OK 16:46:36 <jtanner> vmware however is usually pets instead of cattle 16:47:11 <DMarby> I guess I have two thoughts: I agree that things should be turned more into separate modules, say for example configuration of nic's, disks, etc 16:47:24 <DMarby> But I don't neccesarily think that there should be another action parameter 16:48:38 <jtanner> not seeing exact examples for reconfig in ec2.py docs 16:49:08 <DMarby> I don't have access to an aws environment anymore, and I don't recall exactly what we used 16:50:06 <DMarby> But I think you're right jtanner, usually you'd just create a new machine 16:50:37 <DMarby> Hm, and there are indeed too many edgecases where powerstate can become weird unless you can tell your intent 16:50:38 <akasurde> DMarby, but what about idempotency of existing machine ? 16:51:42 <DMarby> Never mind, I thought that state would be enough if you changed a few behaviours of existing ones, but you are right, there are too many cornercases 16:52:11 <DMarby> So something like action: clone|power|etc seems reasonable 16:52:19 <jtanner> solution is to delete all vmware modules and refer to shell: govc ... >=] 16:52:49 <akasurde> thats the last resort 16:52:52 <jtanner> heh 16:53:01 <jtanner> so that's +2 on action: ... ? 16:53:17 <akasurde> +1 for action 16:53:18 <jtanner> it doesn't have to be that name, was only a sugggestion 16:53:57 <akasurde> how about clone as argument name ? 16:54:03 <akasurde> clone: yes|no 16:54:11 <jtanner> was kinda leaning towards being more generic 16:54:13 <akasurde> like you said above 16:54:22 <jtanner> action: clone|power|reconfig 16:54:50 <akasurde> Looks good to me 16:54:55 <jtanner> clone: true|false could be enough for now, but i see that kicking the can further down the road 16:55:20 <jtanner> people will have issues where clone: false wasn't supposed to reconfigure 16:55:45 <akasurde> yes, correct 16:57:50 <akasurde> Ok I will raise PR and let us have broad discussion there 16:58:00 <jtanner> cool 16:58:27 <akasurde> #action akasurde Raise PR for action in vmware_guest 16:58:38 <akasurde> #topic open floor 16:59:02 <akasurde> Anything else we would like to discuss 16:59:14 <jtanner> nope, not me 16:59:19 <jtanner> time's up anyway =) 16:59:33 <akasurde> Ok 16:59:46 <akasurde> Will wait for 5 min and then close the meeting 17:00:31 <akasurde> jtanner, I have two backports, Can you please take look at them ? https://github.com/ansible/community/issues/206#issuecomment-318875917 17:00:54 <jtanner> sigh, angry unicorns 17:02:55 <akasurde> OK. See you people in next meeting 17:02:58 <akasurde> #endmeeting