[12:59] <smoser> harlowja, hm.. i was suggesting you could render at setup.py (install) time.
[13:48] <rharper> smoser: that hang we saw ... I think it may be related to system-resolvd;  I don't think that's getting started prior to cloud-init-local/net which means if we bring up networking but we don't have resolvd running, I believe it times out resolving (in my case, the GCE datasource)
[13:49] <smoser> ah.
[13:49] <rharper> I added After=systemd-resolved.service; Requires=systemd-resolved.service to cloud-init.service (which is when netmode runs)
[13:49] <smoser> that could do it.
[13:49] <rharper> we should chat with pitti maybe and see where that ought to come up
[13:49] <smoser> http://pad.lv/1629868
[13:49] <smoser> yeah.
[13:49] <smoser> i treid to reproduce with the webhook reporter
[13:49] <smoser> but it didn't work
[13:49] <smoser> because i was using by ip address
[13:50] <rharper> ah
[13:50] <smoser> you dont have to add the 'Requires'
[13:50] <smoser> and probably shouldnt really...
[13:50] <rharper> hrm
[13:50] <rharper> well
[13:50] <smoser> After is sufficent
[13:50] <rharper> it's sufficient, the question is if dns doesn't come up
[13:50] <rharper> is it worth running ?
[13:51] <smoser> yeah, i dont know whow i feel
[13:51] <rharper> if we requires=networking , why not dns too ?
[13:52] <smoser> networking doesnt require dns :)
[13:52] <rharper> fair
[13:52] <smoser> the internet worked without dns for quite some time
[13:52] <smoser> heres my /etc/hosts file
[13:52] <smoser> :)
[13:52] <rharper> I need to play with the unit deps; systemd said I had a loop
[13:52] <smoser> hm..
[13:52] <smoser> lets just ask pitti
[13:52] <smoser> i think he's in
[13:53] <rharper> ok
[14:46] <rharper> smoser: the digitalocean merge is causing unittest failures due to older mock (I'm on xenial, no mock 2.0 here), http://paste.ubuntu.com/23275308/
[14:51] <smoser> rharper, fiddle.
[14:51] <smoser> thank you for catching.
[14:51] <smoser> proposed that merge ?
[14:55] <rharper> y
[15:00] <rharper> smoser: https://code.launchpad.net/~raharper/cloud-init/+git/cloud-init/+merge/307583
[17:17] <harlowja> smoser well its either render at install, or render at runtime, i'm ok with either :-P
[17:18] <smoser> it kinda seems overkill to render at runtime to me.
[17:19] <harlowja> k
[20:16] <smoser> rharper, where idid you get a mock that did not have that ?
[20:16] <smoser> the assert_called ?
[20:17] <smoser> bah.
[20:17] <smoser> never mind.
[20:18] <smoser> i was running your branch and trying to see it fail
[20:18] <smoser> it doesnt fail there ! :)
[22:31] <harlowja> alright, smoser https://git.launchpad.net/~harlowja/cloud-init/commit/?id=8a445f0811e8deca935f5781a84e8d08067b1b1e much smaller now
[22:31] <harlowja> a new makefile step will render that
[22:42] <harlowja> don't see how i can easily do it in setup.py time
[22:42] <harlowja> so makefile step it is :-P
[22:45] <harlowja> smoser rharper mock had some busted version a while ago that let u do self.assert_anything_u_want and it'd say true
[22:46] <rharper> harlowja: yeah; I had seen that discussion
[22:46] <harlowja> because it was thinking u were mocking a new method 'assert_anything_u_want'
[22:46] <harlowja> vs checking prior asserts, lol
[22:47] <rharper> heh