=== rangerpb is now known as rangerpbzzzz
=== cpaelzer_ is now known as cpaelzer
=== sambetts|afk is now known as sambetts
=== shardy is now known as shardy_lunch
[11:23] <fekle> hi! does anyone know a way to install a newer version than cloud-init 0.7.5 on CentOS 7 ? any repos/rpms?
[11:24] <fekle> this old version in centos7 unfortunately doesn't work with openstack and chef, as discussed here yesterday.
[11:25] <fekle> i tried installing the fedora 25 package, however it doesn't install due to python abi differences (centos is 3.4, cloud init wants 3.5)
=== shardy_lunch is now known as shardy
=== rangerpbzzzz is now known as rangerpb
[15:37] <powersj> https://code.launchpad.net/~powersj/cloud-init/+git/cloud-init/+merge/323246
[15:37] <powersj> Enables artful
=== smatzek_ is now known as smatzek
=== sambetts is now known as sambetts|afk
[17:41] <rharper> powersj: I'm trying to build the centos7 rpm in lxd container with smoser's gist, and there looks to be some unittest failures for non-ubuntu; known issue? http://paste.ubuntu.com/24461774/
[17:42] <powersj> rharper: I saw those as well
[17:42] <powersj> haven't put in a fix yet or looked into what to do
[17:42] <rharper> ok, I can push a MR with fixes
[17:42] <rharper> it's just not mocking out distro properly
[17:43] <smoser> that isnt specific to centos
[17:44] <rharper> non debian will fail
[17:44] <rharper> if that's what you mean
[17:44] <smoser> it should not be looking at that path, ever. (/etc/cloud/)
[17:44] <smoser> how would that pass on my desktop ?
[17:44] <rharper> I dunno, it works today, right ? nosetests ?
[17:45] <rharper> pretty strange
[17:45] <rharper> I don't have those files but trunk passes locally
[17:49] <smoser> yeah, you're right. its not mocking distro properly i think
[17:50] <smoser> those tests *have* passed before in centos and even python 2.6
[17:50] <smoser> see e8730078df8c99696b1b684e09c803eef7c4926c
[17:56] <rharper> hrm
[17:57] <rharper> smoser: I think it's fallout from us not running apt unless there is config
[18:01] <rharper> yeah
[18:01] <rharper> we pass an empty config 'apt: {}'  which then prevents the default apt templating from firing
[18:02] <rharper> if you pass a non-empty apt config, then it passes;  what I can't understand yet is why it passes on ubuntu but fails under centos
[18:03] <rharper> e80dbb80987ba44be2899e34fbbbf7d48389b6b5
[18:03] <rharper> so, on ubuntu we have the apt command
[18:03] <rharper> but on centos we dont
[18:05] <smoser> right
[18:06] <rharper> shall I send a PR or do you want to fix ?
[18:10] <smoser> if you have it, please do
[18:11] <smoser> realisitcly we shoudl be mockking the which calls
[18:11] <smoser> rharper, http://paste.ubuntu.com/24461986/ <-- is another piece of it
[18:13] <rharper> smoser: only a simple workaround (adding non-zero config to the dict passed to the apt_configure module)
[18:13] <smoser> actually, iguess ew dont need the importlib portions . those fell out from trying to run tox on the tests/vmtests/
[18:13] <rharper> vmtest?
[18:13] <smoser> er... integration tests
[18:14] <smoser> the default 'tox -e centos6' ran on tests/
[18:14] <rharper> gotcha
[18:14] <smoser> which failed, and i saw it needed importlib
[18:14] <smoser> but didnt' realize that was from the tests/cloudtests
[18:21] <smoser> rharper, i'll look
[18:21] <smoser> i think i have areasonable way
[18:23] <rharper> smoser: ok
[18:23]  * rharper built a centos7 rpm anyhow to test passthrough network
[18:41] <blackboxsw> ok smoser powersj rharper i have an cut at github travis integration for unit tests ++ a project README https://github.com/blackboxsw/cloud-init/pull/1
[18:42] <blackboxsw> had to sort out hithub's default clone depth of 50 not being enough to "see" our latest official tag so tools/read-version wasn't working in the travis env.
[18:45] <blackboxsw> you'll notice the above PR has travis PR checks voting on the merge request. I am still working on jenkins integration which would also give us integration test votes on the project.
[19:00] <powersj> blackboxsw: this is awesome
[19:44] <smoser> rharper, https://code.launchpad.net/~smoser/cloud-init/+git/cloud-init/+merge/323265
[19:54] <rharper> smoser: reading
[20:19] <rharper> smoser: http://paste.ubuntu.com/24462666/  ;  where would we get python-oauthlib for centos 7 ?
[20:21] <smoser> its there on cent6
[20:22] <rharper> I'm on 7
[20:23] <rharper> maybe bad mirror ?
[20:23] <rharper> I see it in my cent7 lxd container
[20:27] <smoser> rharper, i dont love that set of tests as it is
[20:27] <smoser> the and i dont love one more level of indentation for every patch
[20:27] <smoser> actually, the docroartor on that fuction would have owrked i think.
[20:27] <rharper> I don't either
[20:28] <smoser> oh. but patch.object doesnt decorate
[20:28] <rharper> you don't need to patch object
[20:28] <smoser> or does it
[20:28] <smoser> your'e right, but thats what it does now
[20:28] <rharper> you can even just patch it outright before calling the function via an add_patch()
[20:28] <rharper> or decorator
[20:30] <smoser> add_patch () ?
[20:35] <rharper> one of the base unittest classes added an add_patch method
[20:35] <rharper> which does an auto start and stop on remove
[20:35] <rharper> hrm
[20:35] <rharper> guess not, we have that one in curtin
[20:35] <rharper> it's handy
[20:37] <rharper> http://paste.ubuntu.com/24462767/
[20:37] <rharper> looks like the maas centos image doesn't have epel-release package installed by default
[20:37] <rharper> so that's why it didn't find python-oauthlib
[20:38] <rharper> rather annoying that the lxd image and the centos image in maas aren't the same despite being the same release (7.3.116  or something)
[22:09] <rharper> smoser: harlowja: just updated https://bugs.launchpad.net/cloud-init/+bug/1603533/ ;  I built cent7 rpm , installed it but it complains at runtime;
[22:09] <ubot5> Ubuntu bug 1603533 in cloud-init "Can't build correctly using brpm on cent7" [Medium,Fix released]
[22:09] <rharper> thanks ubot5
=== rangerpb is now known as rangerpbzzzz
[23:55] <harlowja> rharper kk, need to build that more often to expose those things