[09:50] <Odd_Bloke> smoser: Yeah, I can't really understand it.
[09:50] <Odd_Bloke> I'm not sure why the failure would only appear once, either.
[09:51] <Odd_Bloke> It's not the only test that uses custom data.
[16:08] <smoser> larsks, i think that fedora 24 (and i'm guessing 25) has a newer httpretty than is released on pypi (0.8.14).
[16:09] <smoser> is that plausible ?
[16:10] <larsks> smoser: I don't think so. I was running from a virtualenv (there is no httpretty package installed on my host).
[16:10] <smoser> well that is odd.
[16:10] <larsks> smoser: I'm going to work on reproducing et al this afternoon.
[16:10] <larsks> Maybe my computer is insane.
[16:10] <smoser> well, i see code in fedora24 that shows this
[16:10] <smoser> and it definitely acts differently than pypi version
[16:11] <larsks> If I were to install the python-httpretty package, I get 0.8.14-1.20161011git70af1f8, so that may be newer than pypi.  But I don't have that installed locally.
[16:11] <larsks> (This is in F25)
[16:23] <smoser> larsks, i updated bug. and opened an httpretty issue
[19:16] <aixtools> quick question: as I do not know enough about python.
[19:17] <aixtools> what actually gets started by python by:
[19:17] <aixtools> #!/opt/bin/python
[19:17] <aixtools> # EASY-INSTALL-SCRIPT: 'cloud-init==0.7.5','cloud-init'
[19:17] <aixtools> __requires__ = 'cloud-init==0.7.5'
[19:17] <aixtools> __import__('pkg_resources').run_script('cloud-init==0.7.5', 'cloud-init')
[19:17] <aixtools> sorry for the multi-line. thought I had cut it down to the last line
[19:19] <larsks> aixtools: that runs cloudinit.cmd.main:main (possibly /usr/lib/python2.7/site-packages/cloudinit/cmd/main.py)
[19:19] <aixtools> so, in the sources - main.py
[19:20] <larsks> Right.
[19:20] <larsks> Specifically, cloudinit/cmd/main.py
[19:20] <aixtools> thx very much.
[19:20] <aixtools> single-step I go...
[19:45] <ssalevan> hey there!  i was wondering if i could ask a quick cloud-init question; if I have an ec2 datasource, is it possible to template out some of that information into a file using the write_files module?
[19:51] <magicalChicken> ssalevan: write_files just writes out specified content, it doesn't have a template system afaik
[19:51] <magicalChicken> ssalevan: there may be other ways to achieve that though, which parts of the config do you need written out?
[19:52] <ssalevan> magicalChicken: right now i basically need to stick eth0's private IP into a JSON configuration file
[19:55] <magicalChicken> ssalevan: I don't know of a good way to handle that that's currently built in, althought someone else may
[19:55] <magicalChicken> One option would be to write a script that handles that and runs after the eni has been written
[19:56] <larsks> ssalevan: for the *private* ip, seems like you could as easily parse the output of 'ip addr'.
[19:56] <larsks> And do that in a runcmd script or something.
[19:56] <ssalevan> ah, good point, could just rig up a bash session
[19:57] <ssalevan> well hey thanks much for your time y'all, much appreciated
[21:46] <ssalevan> i was wondering if i could trouble you for one last question: i've packed up my user-data into a base64-encoded YANL, but cloud-init is returning the following error:  __init__.py[WARNING]: Unhandled non-multipart (text/x-not-multipart) userdata: '---\npreserve_hostname: f...'
[21:46] <ssalevan> i suspect this may be related to a MIME typing issue
[21:48] <larsks> ssalevan: why are you providing a base64-encoded yaml?  Why not just pass in the unencoded yaml?
[21:49] <larsks> Also, does your userdata start with a "#cloud-config" marker? It doesn't look like it from that output...
[21:49] <ssalevan> larsks: so when i try and send up the raw YAML user-data as part of a call to Amazon's RunInstances call, it returns an error that it was unable to encode into base64.  t
[21:49] <ssalevan> ahhh that's probably it
[21:49] <larsks> ssalevan: interesting.  I suspect my second comment is more relevant, yeah.
[21:53] <ssalevan> larsks: that was totally it.  thank you so much!
[21:53] <larsks> Glad it worked!
[21:54] <smoser> larsks, thanks for helping ssalevan
[21:54] <larsks> smoser: sure!