I am working on Resource allocation problem in IaaS cloud. The SLA parameters such as VM start up time, there are no models available to calculate in simulation. Any ideas or suggestion?? thanks in advance !!!!
I suggest that you measure VM startup time and other parameters on a real public cloud and/or on a private cloud that you set up yourself. You could set up OpenStack and make measurements yourself, and you could use the free tier of Amazon Web Services (link below) as long as you stayed beneath the usage specified. By themselves, those performance numbers for a specific OpenStack version and for AWS would be a useful dataset to share.
VM startup time depends on the instance type (spot, on-demand), operation system, size of the OS image, cloud provider, etc. There are studies of the startup of VMs in clouds.