Hard reboot

General concept

If an instance is hung (does not reply to commands), a hard reboot of the instance is the same as “pulling the plug” on a physical computer. Where a soft reboot sends the signal to the OS, a hard reboot simply… reboots. This is useful if the signal from the soft reboot will not get picked up resulting in no action.

Tip

Aside from actually rebooting the instance, a hard reboot will also re-define the instance on the hypervisor. This means that the instance will get the latest hardware version in the platform which is good for keeping compatible with the evolving eco system of OpenStack. Doing an occasional hard reboot with the instance in powered off mode is therefore recommended (when otherwise rebooting).

Executing a hard reboot

The hard reboot option is only available via OpenStack Horizon or OpenStack terminal client. We recommend checking out our guides on working with these platforms if you normally only work with the Cloud management portal.

Important

Please note that a hard reboot will not care about the state of the operating system. Any current writes to the disk in the middle of a hard reboot will get interrupted, potentially causing data loss. We recommend this action as a last resort, or to be executed after the operating system has halted (is shut down).

Hard reboot using the OpenStack Horizon portal

To hard reboot an instance using the Cloud management portal, follow these steps:

  • Press “Compute” and then “Instances” in the sidebar menu.

  • Press the dropdown button “actions”.

  • Press “Hard reboot instance”.

  • Press the confirmation button (“Hard reboot instance”).

  • Wait until the progress bar has finished.

The instance will start again (even if hard rebooted from a powered off state).

Hard reboot using the OpenStack terminal client

To hard reboot an instance using the OpenStack terminal client, follow these steps:

  • Run this command: openstack server list. Save the name of the instance you want to hard reboot.

  • Run this command: openstack server reboot --hard [SERVER NAME], replacing the values in angle brackets by the information from the previous steps.

The instance will start again (even if hard rebooted from a powered off state).