The billing system will bill user for each vnode respectively. Whenever the running time of a vnode reach an hour or user delete the vnode, system will bill user for the resources the vnode has used from last billing time to now. And user's beans will decrease to pay for it.
The status page dispaly how many beans each node has cost, shown as follows:
Besides, the amount of beans to decrease this hour will be figured out according to the following formula:
B = [ Tcpu/a + Mave/b + D/c ]
B represents the amount of beans to decrease this hour. Tcpu represents the cpu time the vnode has used from last billing time to now. And Mave represents the average Memory the vnode has used per second multiply the cpu time. And D represents the disk quota instead of the used amount of disk. "a", "b" and "c" are constant value and "[ ]" means the final value will be top integral, which means the beans cost this hour will at least be 1.
At present, a = 500s, b = 1000000MB, c = 4GB
When your beans are equal to or less than 0, all your vclusters(workspaces) will be stopped. And you can not create, start and scale out(add node) any workspace.