← Back to team overview

openstack team mailing list archive

Re: Ansible playbooks for OpenStack

 

On 07/10/2013 08:09 AM, Daniel P. Berrange wrote:
On Mon, Jul 08, 2013 at 10:02:59AM -0500, Timothy Gerla wrote:
Hi everyone,

I wanted to share this with the group in the hopes that we could get
some feedback. We've built a set of Ansible playbooks to install &
configure a "full" OpenStack deployment based on the Red Hat packaging
on CentOS 6.

(Ansible is a configuration management and software deployment
automation tool: http://www.ansibleworks.com/)

Here are the playbooks and some basic documentation:

https://github.com/ansible/ansible-redhat-openstack

I'm interested to know what folks think and if anyone finds this
useful. I think it might be useful for quick deployments (you can go
from a group of minimally installed CentOS boxes to OpenStack in 20-30
minutes), and I think this could serve as the basis for a more
sophisticated production deployment mechanism.

FWIW, there is a tool called PackStack which uses Puppet to fully
automate deployments of OpenStack in a matter of minutes. It is
the tool that Red Hat currently recommend for deployment on Fedora,
RHEL & CentOS, etc

   https://wiki.openstack.org/wiki/Packstack

And for completeness, for folks using Chef, there is an OpenStack + Chef mailing list [1], an example Chef repo, and a set of OpenStack-specific cookbooks housed on Stackforge [2]. Currently everything is under very active development, with a focus on getting turnkey installation like PackStack done by the Havana release.

[1] https://groups.google.com/forum/?fromgroups=#!forum/opscode-chef-openstack
[2] https://github.com/stackforge/openstack-chef-repo
and https://github.com/stackforge/cookbook-openstack-XXX

where XXX in (common, compute, block-storage, image, identity, network, object-storage, metering, orchestration, ops-messaging, ops-database)

Best,
-jay



References