Ask Your Question

Revision history [back]

click to hide/show revision 1
initial version

heat minimum number of instances and autoscaling


I have a heat script that implements an auto scaling group. I have defined a minimum of two instances and a maximum of five. After the stack is deployed I have two instances. When I put one instance unnder load I see the number of VMs increase to five. I have a few questions about the results I got with some test that are not what I expected;

  1. After the deployment I delete (not stop, but delete) one instance. I expected the heat script to create a new Vm since I specified a minimum number of two. It does not.

  2. After putting one under load I got five VMs. The four machines not having any load are under the threshold i defined to scale down. The system starts to scale down and eliminates four VMs. Two things are strange here:

  3. I expected only three since my minimum specified was two, and more surprising
  4. the VM under load got killed, I end up with a VM that does not have any load.

Anyone to comment on this. My openstack version is juno and there is no loadbalances between the VMs.

Tnx in advance for any answers.