Hi Ajay,

Let me explain how Rally works in this case:


1) run all context setup() methods. (Create tenants, users, setup quotas,
roles, upload images, ....)

2) Run load (100 times stuff with concurrency 10 )

3) run all context cleanup() methods. (Run in vise versa order all chosen
context including generic cleanup)


So, the answer on your questions is that you shouldn't change anything in
Rally to have 100 active instance before cleanup.

You should just use proper benchmark scenario.
Like this one:
https://github.com/stackforge/rally/blob/master/doc/samples/tasks/scenarios/nova/boot.json
That boots VMs but doesn't delete them. So they will be deleted only in
generic cleanup context.


Best regards,
Boris Pavlovic


On Thu, Nov 20, 2014 at 4:28 AM, Ajay Kalambur (akalambu) <
akala...@cisco.com> wrote:

>  Hi
> In rally I have 2 questions. When you use a context like example
> users: 2, tenant:2 I want the users and tenants to not be cleaned up but
> accumulated and cleaned up at end.
> Example create 1 user 1 tenant run tests create second user tenant run
> test and finally clean up user and tenant
>
>  Similarly when you specify
> times: 100
> concurrency: 10
> I want it to launch 10 at a time but not clean them up immediately and
> only clean up at end so 100 instances are active at end before clean up
>
>  What changes would be needed to Rally to support this. Is there also a
> temporary way to do this if not present
>
>  Ajay
>
>
> _______________________________________________
> OpenStack-dev mailing list
> OpenStack-dev@lists.openstack.org
> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
>
>
_______________________________________________
OpenStack-dev mailing list
OpenStack-dev@lists.openstack.org
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev

Reply via email to