16:20:08 <akasurde> #startmeeting Ansible VMware Working Group Meeting
16:20:08 <zodbot> Meeting started Mon Apr 16 16:20:08 2018 UTC.  The chair is akasurde. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:20:08 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
16:20:08 <zodbot> The meeting name has been set to 'ansible_vmware_working_group_meeting'
16:20:14 <akasurde> Hi everyone
16:20:17 <dericcrago> no worries
16:20:35 <dericcrago> warthog9 - we're meeting now
16:20:51 <warthog9> I'm back
16:21:10 <akasurde> #chair dericcrago
16:21:10 <zodbot> Current chairs: akasurde dericcrago
16:21:15 <akasurde> #chair warthog9
16:21:15 <zodbot> Current chairs: akasurde dericcrago warthog9
16:21:32 <akasurde> Sorry was in another meeting
16:24:46 <akasurde> #link https://github.com/ansible/ansible/pulls?utf8=%E2%9C%93&q=is%3Aopen+is%3Apr+label%3Avmware+label%3Anew_module+
16:25:11 <akasurde> I would like bring to your notice that there are lot of modules without any reviews
16:26:32 <dericcrago> akasurde - a couple of weeks ago, we talked about kind of a review outline / overview, I believe you said you were going to take that on, has there been any progress?
16:26:56 <dericcrago> a "here's how to do a review that we'd like to see" guide
16:27:49 <akasurde> nope
16:27:58 <dericcrago> :)
16:27:58 <akasurde> I will added that to my action item
16:28:33 <dericcrago> even just some bullet points to get us started would be great
16:28:40 <dericcrago> we can all help fill it in
16:28:53 <akasurde> #action akasurde Add guideline to review code
16:29:43 <dericcrago> yeah, last time we talked about checking the other SIG for their review procedures
16:31:40 <akasurde> Yes. I remember :)
16:33:03 <dericcrago> ok
16:33:38 <jtanner> sorry, forgot to open irccloud this morning
16:34:51 <dericcrago> hi jtanner
16:36:09 <akasurde> #chair jtanner
16:36:09 <zodbot> Current chairs: akasurde dericcrago jtanner warthog9
16:40:26 <akasurde> warthog9, I would like to know if there is any way to relate Resource Pool and its member ESXi host ?
16:41:00 <warthog9> off hand I'm not sure, I'll go poking today if it's helpful
16:41:34 <warthog9> the answer is 'likely' just not 100% sure on the exact how
16:41:39 <jtanner> i thought RPs were children of the datacenter or of a cluster, and it the RP had children of each esxi host?
16:41:56 <jtanner> datacenter|cluster -> resourcepool -> esxi
16:42:44 <warthog9> akasurde: are you trying to figure out where a specific resource in a pool is at?
16:42:46 <akasurde> Agreed
16:42:59 <akasurde> warthog9, exactly
16:43:00 <akasurde> https://github.com/ansible/ansible/blob/devel/lib/ansible/modules/cloud/vmware/vmware_guest.py#L1648-L1667
16:43:42 <warthog9> akasurde: let me go poking around
16:43:52 <akasurde> I feel there is chace of  improvement  here, once I have the reason how and I can fix it
16:43:56 * jtanner relocates
16:44:47 <akasurde> I felt this while I was debugging this issue - https://github.com/ansible/ansible/issues/38043
16:48:33 <dericcrago> akasurde - could you change the starting point from self.content to the host.parent or something? ``resource_pools = self.cache.get_all_objs(self.content, [vim.ResourcePool])``
16:49:39 <dericcrago> I'm just trying to think if there's a way to localize it enough that you wouldn't run into a name conflict
16:52:23 <akasurde> dericcrago, get_all_objs expects content as first params, I will try to find something around
16:53:18 <dericcrago> oh, right, sorry, I was thinking the folder argument
16:53:30 <dericcrago> ``def get_all_objs(content, vimtype, folder=None, recurse=True)``
16:54:07 <dericcrago> which ends up defaulting to ``content.rootFolder`` if folder isn't provided
16:54:37 <akasurde> ok
16:54:59 <dericcrago> just a thought
16:56:31 <akasurde> will try that as well
16:58:04 <akasurde> Does someone like to discuss anything else ?
16:58:06 <dericcrago> yes
16:58:21 <akasurde> dericcrago, go ahead
16:58:28 <dericcrago> in an effort to improve the testing with vcsim, I got a couple of existing issues with the current version of vcsim fixed and submitted a PR to update the vcsim that is used when testing the modules https://github.com/ansible/ansible/pull/38793 has anyone had a chance to take a look? I know akasurde requested a review from jtanner and mattclay
17:00:02 <jtanner> curious if the golang container is going to be more or less compatible over time with doug's work on vcsim
17:00:03 <akasurde> dericcrago, I have question - is there any reason to pull golang:1 rather than fedora:26
17:00:25 <warthog9> akasurde: it's flipping to a debian based container it looks like
17:00:35 <jtanner> what's doug's dev machine running?
17:00:42 <dericcrago> yeah, I switched as it made the dockerfile simpler
17:01:20 <dericcrago> https://github.com/dericcrago/ansible/blob/cf5d8abb0be6eb11352e1b5b602e8b39fe452f46/test/utils/docker/vcenter-simulator/Dockerfile
17:01:58 <jtanner> i'm okay with switching, i just hesitate to deviate too much from what we know has remained stable
17:02:15 <dericcrago> if we wanted to peg the go version we could change the ``FROM golang:1`` to ``FROM golang:1.10`` or whatever
17:02:17 <akasurde> same here
17:02:29 <warthog9> I'd agree with jtanner though maybe a little stronger on the not changing
17:02:56 <jtanner> dericcrago: can you check with doug on that and see what he thinks?
17:03:15 <dericcrago> sure
17:04:08 <jtanner> otherwise, lgtm
17:06:13 <dericcrago> just so everyone is clear, the container is more or less pegged to a working version by the hash here: https://github.com/ansible/ansible/blob/devel/test/runner/lib/cloud/vcenter.py#L46
17:06:59 <dericcrago> so even if this PR was merged today, it wouldn't flip over to using it until a 2nd PR was merged to update that hash
17:07:11 <jtanner> yep
17:07:50 <akasurde> yes
17:07:59 <jtanner> merge PR, mattclay or gundalow builds + publish, someone tests the new hash locally across all tests, someone files PR to update hash in the cloud provider
17:08:07 <jtanner> in that order
17:09:07 <mattclay> I'm setting up a new repo just for the simulator so manual builds won't be needed any more.
17:09:24 <dericcrago> yep, I've already started on the followup PR using a local build
17:09:25 <akasurde> historically, vcsim is known to be unstable with latest changes
17:09:32 <akasurde> #chair mattclay
17:09:32 <zodbot> Current chairs: akasurde dericcrago jtanner mattclay warthog9
17:09:36 <dericcrago> to do the testing
17:09:57 <akasurde> mattclay, cool
17:11:33 <mattclay> The code will be here shortly: https://github.com/ansible/vcenter-test-container
17:12:38 <akasurde> #link https://github.com/ansible/vcenter-test-container
17:12:52 <dericcrago> in running all of the 'v{center,mware}_*' module tests, the only thing I ran into with the newer version was this block: https://github.com/ansible/ansible/blob/devel/lib/ansible/module_utils/vmware.py#L175-L180
17:13:53 <dericcrago> it needed to use the except block, but it wasn't hitting the exception
17:19:18 <dericcrago> alright, just shot doug a message, I'll let everyone know what he says
17:19:37 <jtanner> mattclay: since we have at least 2 "simulators" now, should they have a common prefix in their reponame?
17:20:08 <mattclay> jtanner: The two repos are common suffixes "-test-container"
17:20:14 <mattclay> s/are/have/
17:20:21 <jtanner> ok
17:21:09 <mattclay> I've imported the existing vcenter-simulator into the repo and am testing a build on quay.io. If that works I'll tag it as 1.0.0.
17:21:37 <akasurde> ok
17:21:48 <dericcrago> akasurde - the specific vcsim commit hash and python package versions are all called out in that PR so it will be easier to use look at that dockerfile and know exactly which versions of things are being used
17:21:53 <dericcrago> or at least, that's the idea
17:22:28 <akasurde> ok
17:24:43 <akasurde> I think, we are little over time. We can close this meeting if we don't have anything else to discuss
17:24:53 <jtanner> i have nothing
17:25:07 <jtanner> dericcrago: thanks for poking at that,  should help out in the long run
17:25:37 <dericcrago> yep, no problem
17:25:54 <jtanner> adding features to the flask app and not worrying about getting a new broken vcsim will be awesome
17:26:26 <dericcrago> :)
17:26:58 <akasurde> Thanks for everyone
17:27:01 <akasurde> #endmeeting