Hitting an issue where cloud provider doesn't allow 100% CPU usage

Question

I am currently using Azure PaaS worker/web roles and I am hitting an annoying issue where if my database role is working really hard (100% CPU) it gets automatically restarted by the Azure App Fabric controller.

This is getting really frustrating.

My options are to over provision the role and pay more to stop the restarts. OR seek a alternative service provider which doesn't do auto / heartbeat / restarts.

Are the IaaS VM instances also subject to a Heartbeat health check? What about Amazon's EC2 - Would they also have the same issues?

Before everyone tells me that it shouldn't be using 100% CPU etc. It only really happens once a day for about 20 minutes.

Answer 1

Edited Per David M's post here Windows Azure VM (Iaas) unexpected restarts IaaS VM's can restart if unreachable by the AppFabric controller.