[01:20] <smoser> blackboxsw: i'im goin gto upload with the IBMCloud to ubuntu
[01:20] <smoser> just the ubuntu/devel as it is
[01:37] <blackboxsw> +1 smoser
[09:17] <apollo13> hi, I am testing with the NoCloud provider and noticed that cloud-init will try to reconfigure if I change the instanceid; is there a way to only allow cloud-init to run once
[09:18] <apollo13> this is currently on a fedora atomic image
=== shardy is now known as shardy_afk
=== shardy_afk is now known as shardy
[15:46] <blackboxsw> apollo13: I think you could touch /etc/cloud/cloud-init.disabled after first boot and cloud-init won't do anything anymore
[15:49] <blackboxsw> per http://cloudinit.readthedocs.io/en/latest/topics/boot.html#generator
[15:59] <rharper> apollo13: also manual_cache_clean: true  would work as well;  https://git.launchpad.net/cloud-init/tree/doc/examples/cloud-config.txt#n450
[17:06] <blackboxsw> smoser: comments on https://code.launchpad.net/~smoser/cloud-init/+git/cloud-init/+merge/342948
[17:10] <blackboxsw> rharper: I'll pull together a quick patch adding ntp:config key validation and post it
[17:13] <rharper> thx
[19:18] <smoser> blackboxsw: https://code.launchpad.net/~smoser/cloud-init/+git/cloud-init/+merge/342948 updated.
[19:18] <dpb1> ntp?
[19:19] <smoser> on that now.
[20:14] <blackboxsw> rharper: smoser my final patch suggestion ntp  validation http://paste.ubuntu.com/p/dcKcYrM5ty/
[20:15] <smoser> i had just posted some coments there.
=== natorious_ is now known as natorious
=== powersj_ is now known as powersj
=== voja_ is now known as voja
=== holser___ is now known as holser_
=== aimeeu__ is now known as aimeeu
[20:32] <smoser> rharper: i will check back in later. i have to go now.
[20:32] <smoser> but your issues with util.read_file_or_url
[20:32] <smoser> is that it was just busted.
[20:32] <smoser> it is supposed to return bytes in .contents
[20:33] <smoser> and str() should give you text on that thing.
[20:33] <blackboxsw> https://code.launchpad.net/~smoser/cloud-init/+git/cloud-init/+merge/342948 merged
[20:36] <smoser> you could just as well use load_file sinc eyou'rre loading a file in all cases
[20:36] <smoser> http://paste.ubuntu.com/p/fHSjws2hR9/
[20:37] <smoser> that set of changes works correclty, and uses the response of read_file_or_url "correctly" such that it works witih a url or a file.
[20:51] <blackboxsw> addressed review comments on https://code.launchpad.net/~chad.smith/cloud-init/+git/cloud-init/+merge/342761
[20:51] <blackboxsw> changed description
[21:14] <rharper> smoser: my issue was that in tox -e python2.7 without my code changes, jinja blows up, not the unittests;
[21:15] <rharper> jinja2 's render_template expects decoded str (ie, not utf-8); but it does load_file() on the template (chrony) and gets back utf-8 chars; those have to be decoded before you can call jinja.render_template()
[21:16] <rharper> once we decode the contents before rendering, the next issue was python2.7 calling str.decode() fails because the default encoding for python2.7 wasn't UTF-8;
[21:16] <rharper> that's my analysis;
[21:17] <rharper> I'm working on switching to attribute; I like that better than the calling in init; let's see if that makes it easier
[21:51] <smoser> something is fishy there still
[22:24] <rharper> smoser: yes, the simple recreate is the test using the "real' templates in test_handle_ntp*.py;  if you run that one under py2.7 and remove the util.decode_text() calls, you'll see the jinja error
[22:24] <rharper> blackboxsw: thanks for the schema update; pulling that in
[22:25] <rharper> smoser:  I'm just about ready to push an update switch to property for the preferred_ntp_clients; that works quite well and avoids mocking in non-ntp related test cases that use distro objects
[22:25] <blackboxsw> np rharper
[22:26] <rharper> ok, i've pushed, I've not squashed yet but will do so before we land
[22:46] <blackboxsw> lander squashes so you don't have to
[22:54] <rharper> blackboxsw: interesting
[22:54]  * rharper likes to squash