On 03/06/2014 12:51 PM, Matthew Miller wrote:
On Thu, Mar 06, 2014 at 10:58:04AM -0600, Joe Brockmeier wrote:
I am half-minded to make the proposal that we drop cloud-init. We haven't
built up a big library of cloud-init examples for Fedora, and once
<https://github.com/cgwalters/min-metadata-service/issues/2> is implemented,
we can focus on doing so as simple shell scripts. Most cloud-init examples
I've seen using the cloud-config syntax end up not being very cross-platform
anyway.
How would this go over for people coming from $other_images that
have cloud-init and expect to use it? If I understand correctly,
you're saying their existing stuff isn't going to work well w/Fedora
even if we have cloud-init, but I wonder if not having it is going
to be seen as a negative?
That's why I'm only half-minded. But yes, I am saying that. The more I think
about it, having a tailored image with cloud-init might be the way to go --
then we can measure and see how much it is used in comparison.
Alternately, here's a crazy idea: the min-metadata service could detect if
userdata is in a syntax it doesn't understand and fire off a helper script.
That helper script could then install and exec cloud-init.
That sounds like it could be potentially breaky/complicated. (I could be
wrong.)
My gut would be to go with the tailored image and see if it gets used,
as the messaging issue concerns me. We could phase the image out in a
release or two if it isn't picked up.
--
Joe Brockmeier | Principal Cloud & Storage Analyst
jzb@xxxxxxxxxx | http://community.redhat.com/
Twitter: @jzb | http://dissociatedpress.net/
_______________________________________________
cloud mailing list
cloud@xxxxxxxxxxxxxxxxxxxxxxx
https://admin.fedoraproject.org/mailman/listinfo/cloud
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
