fuel-dev team mailing list archive

[Vladimir Kuklin <vkuklin@xxxxxxxxxxxx>]

Ok, guys, let's start a blueprint for this feature. There should be some
information on the current state of the art. AFAIK there is no testing
result on parallel controller deployment.

Andrew, did you try to implement this feature? Did you test parallel
deployment of the controllers? Did it fail or did it succeed?

As for me speaking for parallel deployment reminds us of our pluggable
architecture that should allow us to run some roles in parallels. I am not
quite sure that we have an opportunity to spend resources on such tasks
right now.


On Wed, Mar 12, 2014 at 2:42 AM, Sergey Vasilenko
<svasilenko@xxxxxxxxxxxx>wrote:

> For decrease count of possible HA-issues in the parallel deployment mode
> we should:
> * create any Pacemaker resources only at primary controller deployment
> process
> * move Galera deployment to the its ows stage (for prevent parallel galera
> deployment)
> * OR rewrite galera ocf script, for syncing slave nodes with master as
> node-by-node (not all together)
>
>
> /sv
>
>


-- 
Yours Faithfully,
Vladimir Kuklin,
Fuel Library Tech Lead,
Mirantis, Inc.
+7 (495) 640-49-04
+7 (926) 702-39-68
Skype kuklinvv
45bk3, Vorontsovskaya Str.
Moscow, Russia,
www.mirantis.com <http://www.mirantis.ru/>
www.mirantis.ru
vkuklin@xxxxxxxxxxxx