Detect network failures

[Aymen Frikha]

The current implementation of masakari seems not able to detect a network failure that is happening to some important networks which are:

- Overlay Geneve network : If network interfaces are down , or a switch is down that connect a compute node to the Overlay Geneve network, the VM is stuck and does not migrate to another compute node.
- Storage network : If storage Network interfaces are down, or a switch is down that connect a compute node to the storage network, the VM is stuck and does not migrate to another compute node.

[Aymen Frikha (aym-frikha)]

+ ~field-high

[Aymen Frikha (aym-frikha)]

After some documentation, I found out that if masakari is based on consul agents rather than on pacemaker corosync , we would be able to detect those failures: https://docs.openstack.org/masakari-monitors/latest/hostmonitor.html