NimbostratusIs there a counter (or another logging option) available which keeps track of monitor failures even if they do not result in a downed pool member? In essence, this would provide some insight in the health of the network between the node and the LTM. Packet loss would become visible this way, even if the loss itself isn't large enough to trigger a timeout.
AltostratusThis is a great question. There doesn't seem to be any counter for state transitions, let alone for monitor failures. I've looked in the output of 'tmsh show ltm pool detail', and also through the MIBS.
For up/down transition stats, the best I can come up with is to parse /var/log/ltm for node up and node down message.
If you want stats on monitor failures, you could do it with an external monitor but I'm not sure it would be worth the effort or overhead.