
Backup strategy
The following are considerations you should understand and discuss before performing a Management Host
backup:
1.
Perform a full backup of HP Cloud OS for Moonshot Management host as soon as the cloud and region
are active.
2.
Perform a nightly back of the Controller Node and Baremetal Host at the same time.
3.
Trigger a backup before any scheduled maintenance. Ensure you follow a power ON/OFF sequence. See
the Powering on/off the Management Host.
4.
Trigger a backup before applying hot-fixes and patches through the Cloud OS Delivery Network (CODN).
5.
Perform a backup before you plan an upgrade.
Restore strategy
The following are considerations you should take into account before performing a Management Host restore.
1.
Before performing any restoration, power off the HP Cloud OS for Moonshot Management host (Admin
and Controller nodes and Baremetal host).
2.
Use latest backup files and maintain the existing naming convention (Virtual Machine names).
3.
If there is a node failure, remove the datastore files of the corrupt node and then perform a restore.
4.
If your services are corrupted and the Admin node fails, restore only the Admin node.
5.
If both your Baremetal host or Controller node fail, restore them both.
Powering on/off the Management host
It is highly recommended you follow the steps for powering off the HP Cloud OS for Moonshot Management
host after a successful cloud deployment.
1.
Power off the Baremetal Host.
2.
Once the Baremetal host is off, power off the Controller Node.
3.
Once Controller node is off, power off the Admin Node.
4.
Then, power off the hypervisor power (if required).
Once you have successfully completed the powering off steps, power on the HP Cloud OS for Moonshot
Management host in this order:
1.
Power on the hypervisor.
2.
Power on the Admin node.
© Copyright 2014 Hewlett-Packard Development Company Page 4
Kommentare zu diesen Handbüchern