How to plan an openstack environment so if the controller node is to fail the cloud get be recovered in a reasonable time frame?
What should be done before-hand in case the controller node fails? What steps should be taken to make sure we have mysql database copies, or any other information off of the controller node?
Any other recommendations would be appreciated!
Hi you need to have backup of all the database like keystone,glance,nova,neutron
install a new controller and restore the database
you need to have the same ip details and same hostname on the new controller
Copy the configuration files with cron to a shared folder between servers, make a backup of all the databases with mysqldump and proceed to recover your controller node. Also you can use Puppet or other automation tool to recover your nodes when they fail.
Asked: 2015-02-26 14:37:08 -0600
Seen: 588 times
Last updated: Feb 26 '15
