[one-users] default scheduler obsesses over failing hosts

[Carlos A.]

Hello,

I have a vmware-based ONE deployment, and I am facing this problem:

host1 is running 1 VM
host2 is running 2 VM

When I create a new VM (without any requirements), the scheduler tries 
to deploy it in host2.

The deployment of the VM fails due to a VMWare issue (it falls into a 
internal error state, waiting for a query to be answered). Any other VM 
deployed in a host in this state will fail.

The problem is that any subsequent vm creation is tried to be deployed 
in host2 while host1 is properly working!

If I stated any requirement for the VM to be deployed into host1 (ex. 
-RUNNING_VM), the scheduler will properly deploy in host1.

Do I have any way to avoid obsessing on host2? (appart from creating my 
own scheduler)

On the other side... Do I have any possibility to force the default 
scheduler to follow a specific deployment policy? i.e. force to use an 
expression such as RANK=-RUNNING_VM in the templates, or an alternative way.

Regards,
Carlos A.