Fix NetconfNodeHandler interactions
There are a couple of failures in interacting with RemoteDeviceHandler
API (which is completely undocumented).
The API design (and interactions in implementations) assume that:
- onDeviceConnected()/onDeviceDisconnected() manage a reconnecting
process -- i.e. when 'disconnected' it is assumed the device might
come back again
- onDeviceFailed() is the hard alternative, indicating the node is node
coming back
This is then picked up and routed through NetconfTopologyDeviceSalFacade
to NetconfDeviceTopologyAdapter, which makes assumptions and propagates
them to the datastore.
JIRA: NETCONF-1097
Change-Id: Ib148a617b52e775261212aa3a809d1ab450d7d13
Signed-off-by: Ivan Hrasko <ivan.hrasko@pantheon.tech>
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
(cherry picked from commit
9db827875364d13611ad3b00f24e42ec259417f4)