OpenStack Compute (nova)

Control nova-compute on which instances start

Asked by Eric Bales

I've an experimental local installation of an OpenStack cluster. It's all working fine. The cluster has a single admin node (running everything except nova-compute) and 4 compute nodes (running only nove-compute).

What I'd like to do is be able to control which nova-compute node my instances start on. When I start them, this appears to be random. Sometimes I'll get 3 or 4 instances on the same compute node in a row, sometimes it'll use one of the other compute nodes.

Since the cluster is experimental, it's built using small (underpowered, low memory) hardware, so too many isntances on any given box results in that box running particularly slow.

I've tried using the -v and --availablity-node options, but they don't seem to have any effect. Can someone help?

Thanks.

Question information

Language:
English Edit question
Status:
Solved
For:
OpenStack Compute (nova) Edit question
Assignee:
No assignee Edit question
Solved by:
Vish Ishaya
Solved:
Last query:
Last reply:
Revision history for this message
Best Vish Ishaya (vishvananda) said : 		#1

try
--compute_scheduler_driver=nova.scheduler.simple.SimpleScheduler
or
--compute_scheduler_driver=nova.scheduler.distributed_scheduler.DistributedScheduler

Vish

On Feb 29, 2012, at 1:00 PM, Eric Bales wrote:

> New question #189261 on OpenStack Compute (nova):
> https://answers.launchpad.net/nova/+question/189261
>
> I've an experimental local installation of an OpenStack cluster. It's all working fine. The cluster has a single admin node (running everything except nova-compute) and 4 compute nodes (running only nove-compute).
>
> What I'd like to do is be able to control which nova-compute node my instances start on. When I start them, this appears to be random. Sometimes I'll get 3 or 4 instances on the same compute node in a row, sometimes it'll use one of the other compute nodes.
>
> Since the cluster is experimental, it's built using small (underpowered, low memory) hardware, so too many isntances on any given box results in that box running particularly slow.
>
> I've tried using the -v and --availablity-node options, but they don't seem to have any effect. Can someone help?
>
> Thanks.
>
>
> --
> You received this question notification because you are a member of Nova
> Core, which is an answer contact for OpenStack Compute (nova).

Revision history for this message
Eric Bales (eric-bales) said : 		#2

Thanks Vish Ishaya, that solved my question.

Revision history for this message
sanjaya dahal (sdtranquility) said : 		#3

hi i have installed openstack diablo final on one of my controller node . it works fine for controller node . i also install compute node on other server the compute node is enabled when i view it in dashboard but the up status in dashboard is false. i tried to view my compute node log file it shows ecconrefused errno111 ampq in ip x.x.x.x is unrechable. the nova.conf file on bioth node is same . i have my both server on same network. i also put multi host = T while creating nova network. my compute node have only nova-compute and nova network install along with ntp on both server. but compute node wont work . can u please help whats going on with my installetion.