We need to add the description of the recovering procedure for the following case:
1. Customer have HA environment with MOS 8.x-9.x
2. One partition on one controller become full (not enough free disk space error)
3. Pacemaker automatically shut down all services on this controller
4. Operator should login to the controller node, move/remove extra files from the disks and then execute the following command to recover pacemaker:
crm node status-attr `hostname -f` delete "#health_disk"
Other possible workarounds:
1. Restart pacemaker service:
service pacemaker restart
2. Reboot controller node
We need to describe in the documentation for OpenStack operators and support team the right workflow of recovering for this situation.
We need to add the description of the recovering procedure for the following case:
1. Customer have HA environment with MOS 8.x-9.x
2. One partition on one controller become full (not enough free disk space error)
3. Pacemaker automatically shut down all services on this controller
4. Operator should login to the controller node, move/remove extra files from the disks and then execute the following command to recover pacemaker:
crm node status-attr `hostname -f` delete "#health_disk"
Other possible workarounds:
1. Restart pacemaker service:
service pacemaker restart
2. Reboot controller node
We need to describe in the documentation for OpenStack operators and support team the right workflow of recovering for this situation.
Please see comments from Vladimir Kuklin here for more detailed information: /bugs.launchpad .net/fuel/ +bug/1595100
https:/