
the theory that updating nodepool in the middle of the night to not be disruptive runs afoul of the fact that if they do, there isn't anyone around to fix them (due to when infra-root working hours are). The net being that half a day of development is lost, asia and europe lose most of their day, until the first infra root member is up and can trigger a roll back. given that the infra team is largely centered on west coast if we update this at 14 UTC instead of 2 UTC that means any issues will surface while the entire infra team is on. Which means the ability to catch these early. Change-Id: I515bcd401ad09dcec405cd13dcb1f5652e10b2dc
These are a set of puppet manifests and modules that are currently being used to manage the OpenStack Project infrastructure.
The main entry point is in manifests/site.pp.
In general, most of the modules here are designed to be able to be run either in agent or apply mode.
These puppet modules require puppet 2.7 or greater. Additionally, the site.pp manifest assumes the existence of hiera.
See http://ci.openstack.org for more information.
Description
Languages
Python
37.2%
Jinja
36.6%
Shell
13.6%
Dockerfile
3.8%
JavaScript
3%
Other
5.8%