Open Stack

On Fri, Apr 4, 2014 at 5:19 AM, Vladimir Kozhukalov <
vkozhukalov at mirantis.com> wrote:

> On the other hand, it is easy to imagine a situation when you want to run
> agent on every node of your cluster after installing OS. It could be useful
> to keep hardware info consistent (for example, many hardware configurations
> allow one to add hard drives in run time). It also could be useful for "on
> the fly" firmware updates. It could be useful for "on the fly"
> manipulations with lvm groups/volumes and so on.
>
>
There are lots of configuration management agents already out there (chef?
puppet? salt? ansible? ... the list is pretty long these days...) which you
can bake into the images that you deploy with Ironic, but I'd like to be
clear that, in my opinion, Ironic's responsibility ends where the host OS
begins. Ironic is a bare metal provisioning service, not a configuration
management service.

What you're suggesting is similar to saying, "we want to run an agent in
every KVM VM in our cloud," except most customers would clearly object to
this. The only difference here is that you (and tripleo) are the deployer
*and* the user of Ironic; that's a special case, but not the only use case
which Ironic is servicing.

Regards,
Devananda
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack-dev/attachments/20140404/55c3bbf0/attachment.html>