In my recent attempts to venture into ROS 5.4 software made me ask for some feature additions:
Netwatch tool should have interface binding feature, so it is possible to mark over what interface the tested network address should be pingable;
Netwatch tool should have TTL option;
In IP route “Check Gateway” by ping checking should be able settable to apply only to that single route and not to bring down any other route associated to that gateway;
IP DHCP-Client should have “Check Gateway” option;
Netwatch should have option to trigger host status change depending the state over numerous pings (currently it is set to 3, but it is still bit short - maybe add an option to change that default 3 times pinging?).
All these features just advance existing ones a bit and should not break any current functionality.
As Netwatch status change is being triggered by just simply one ping getting lost, it will render that tool pretty much useless for real world usage. It should be possible to monitor a host over number of ping results and then the host status will get determined by some set percent of those ping results. That will allow some dropped pings here and there without too hasty judgement as host being down.
I think a delay option would be more useful. Define a delay in seconds. When status of the test changes (ping drops) the status of the netwatch tracker stays the same until the test has stayed at the new status for the duration of the delay.
Mostly like IP SLA trackers in IOS.