Stop a Cloud GPU VM

stop a cloud gpu vm in dcd to deallocate resources, release ip addresses, and suspend billing os level shutdown is not sufficient stopping a cloud gpu vm from the dcd deallocates its resources and suspends billing the vm remains in your infrastructure and can be restarted at any time from the dcd maintains dedicated resources for each user; your physical cpus are not shared with other users therefore, stopping a vm at the operating system level alone does not deallocate resources or suspend billing always use the dcd to stop cloud gpu vms important a reset forces the cloud gpu vm to shut down and restart but may result in data loss shutting down a vm at the operating system level does not deallocate resources or suspend billing regardless of how the vm was shut down, it can only be restarted from the dcd stopping a vm will disconnect power to your vm deallocate any dynamically assigned ip address to stop a cloud gpu vm, follow these steps in the dcd , go to menu > virtual data centers select the data center that has the respective cloud gpu vm select the respective cloud gpu vm in the workspace on the inspector pane, select power > stop in the power state dialog, select the checkbox to confirm the power cycle, and select apply stop select provision changes in the provision data center dialog, select confirm in the confirm starting and stopping of servers dialog, enter your password , and select ok in the provision data center dialog, select provision now result the cloud gpu vm shuts down the inspector pane displays the status as shut off , and an x marks the vm in the workspace the dcd deallocates the ip address and suspends billing