I have launched an instance (bootable from volume) using Heat template. The VM and the volume reside in separate nodes. The instance becomes read-only from time to time (don't know why) and becomes all right after reboot. I suspect that there occurs a temporary disconnectivity of the network between the corresponding compute and cinder nodes of the instance.

What I plan to do is this :

• Get the Compute node IP/hostname of the instance
• Get the Cinder node IP/hostname of the instance
• Set an alarm which triggers when the network gets disconnected between the compute and the cinder nodes.
• Wait till the network reconnects
• On reconnecting, check if the instance OS has become read-only.
• If yes, reboot the instance to fix the read-only issue

What should be the heat template achieving the above objectives?