Cloud computing supposedly lets you instantiate on-demand processing power through additional VMs. The allocation process is frequently advertised as fast without further details. My personal experience is that VM instantiations on various cloud providers takes anywhere from 10s to 2h.
Since cloud computing providers are charge VMs per hour, I would be very interested in a comparison of VM start latency (from provisioning request, to client app being executed):
- Amazon EC2: ?
- Windows Azure: ?
- Rackspace Cloud: ?
- add yours here
What is the typically observed latency on favorite cloud provider?