For example, if you copy the text below and place it into the file /etc/xdg/telium/haast.conf.d/network_cable.sensors.conf then you can test the sensors graph on the GUI. This sensors checks if the network connection on ‘eth1’ is down. If the connection is down the health score increases by 5 points. If the connection is up the health score remains unchanged for that cycle. (i.e. 0 health points).
If HAAst can’t read the NIC, then an error score of 20 is used. If HAAst can talk to the NIC but not extract a state then the warning score of 10 is used. If the sensors reports 0, then the cumulative score for this sensors is reset to 0. This sensor runs every 5 seconds.
So restart HAAst with this sensor in place and open up the graph on the HAAst GUI. Unlpug the network cable and watch your score, then replug the cable and see it reset to 0. If you leave the cable unplugged long enough and the health score reaches the critical level, then your cluster will fail-over to the peer.
; Test to ensure a network cable is plugged into the NIC, and that the cable is live (i.e. other end is plugged in)
network-cable/description=Cable plugged in
network-cable/type=nic
network-cable/input=state
network-cable/parameters=state:ethernet | interface:eth1
network-cable/scoring= =up:0 | :5
network-cable/warningscore=10
network-cable/errorscore=20
network-cable/resetcumulativescore=0
network-cable/interval=5
network-cable/debug=false
