Ask Your Question

Revision history [back]

It could be that it occasionally reaches those peaks but if it is a constant behavior I think there is a problem. Is your cloud infrastructure 'very busy' (continuous creation and deletion of VMs)? As nova-conductor is in charge of making nova related RPC it could be possible that behavior only on heavy loads on your infrastructure.

Are your nova log files showing any kind of ERROR? What are the characteristics of your controller node (CPU & RAM)?