Open Stack

On 16 Jan 2014, at 13:54, Ilya Kharin <ikharin at mirantis.com> wrote:

> On 16 Jan 2014, at 12:57, Oleg Gelbukh <ogelbukh at mirantis.com> wrote:
> 
>> 
>> 
>> On Thu, Jan 16, 2014 at 2:21 AM, Jay Pipes <jaypipes at gmail.com> wrote:
>> On Wed, 2014-01-15 at 21:46 +0000, Hugh Saunders wrote:
>> > On 15 January 2014 21:14, Ilya Kharin <ikharin at mirantis.com> wrote:
>> >
>> >         Hi, guys,
>> >
>> >         In Rally there is an entity that represents installed instance
>> >         of OpenStack.
>> >         What you think about a proper name for the entity? (a
>> >         Deployment, a Cluster, an Installation, an Instance or
>> >         something else)
>> >
>> > I vote for Deployment.
>> 
>> Doesn't it sound a bit weird to deploy a Deployment? Otherwise, it does not really matter how it is called as long as the naming is consistent.
>> 
>> I have another question. Shoud we think about separation of Deployment and Endpoint entities in API? Deployment is an object managed by deployment engine, while Endpoint can refer to existing installation which has nothing to do with deployment engine. It means that different sets of operations are applicable to those entities. What do you think?
> 
> Yep, you are right. It's very usefull because currently there is an DummyEngine which does nothing, just returns an endpoint. So, an Endpoint should be an entity that represents an installed OpenStack and contains an endpoint and credentials. In this case the deployment process should be changed. As a result of the deployment process an instance of an Endpoint will created. To start a Task an instance of an Endpoint should be passed.
Also, if try to imagine this in examples, then this can be represented like that:

1) In case of an existent cloud, only an Endpoint should be added:

   $ cat endpoint.jso
   {
       "auth_url": "http://example.com:5000/v2.0",
       "username": "admin",
       "password": "secret",
       "tenant_name": "demo"
   }
   $ rally endpoint add endpoint.json

The last command shows an endpoint_id of an added endpoint.

2.1) In case when OpenStack should be deployed:

    $ rally deployment create --filename=deployment.json

2.2) Wait when a status has a proper value ("deploy->finished").

    $ rally deployment list

In case of success deployment an endpoint is going to be automatically created for the deployment:

    $ rally endpoint list

3) To start a task an endpoint_id is required. There is one endpoint after (1) or (2.2) step, use it:

    $ rally task start --endpoint-id=<endpoint_id> --task task.json
> 
>> 
>> --
>> Best regards,
>> Oleg Gelbukh
>>  
>> 
>> ++
>> 
>> Best,
>> -jay
>> 
>> 
>> 
>> _______________________________________________
>> OpenStack-dev mailing list
>> OpenStack-dev at lists.openstack.org
>> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
>> 
>> _______________________________________________
>> OpenStack-dev mailing list
>> OpenStack-dev at lists.openstack.org
>> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack-dev/attachments/20140116/e3449dbb/attachment.html>