Hi,
I would like to implement similar pricing model as Amazon AWS has - that is pricing per running vm-hour, from the time an vm is launched until it is stopped. So basically I would lie to be able to count the hours an vm is powered on in an interval (usually per month), regardless if it is on full load or just idle.
For example, if I power on a vm at 0:01 on Jan 1st and stop it 11:50 on Jan 2nd and then start it agan at 0:01 on Jan 5th and stop it at 11:50 on Jan 5th, and then don't use it anymore, it should give me a number of 36+12 = 48 running hours for January - and I would bill only those 48 hours at a fixed rate per hour.
Is this possible, has anybody done it yet?