I would like to use netwatch for monitoring/recovering WANs, including 3G/LTE networks; however, the current implementation of netwatch lacks necessary flexibility and would needlessly trigger actions during intermittent or inconsequential outages.
Could you please add the following enhancements:
1. Ability to specify a (comma separated) list of (up to 8?) IP addresses to ping (for better outage detection):
Justification: A single IP address being unreachable is not a good indicator that a network in unreachable: (1) There may be redundancy that the single IP address cannot account for; (2) The single IP address pinged may simply be one of a multitude of hosts pingable hosts. Therefore, the one host being unreachable – possibly because it is down for maintenance – need not trigger an action; (3) If the host being pinged is someone else’s public host (possibly also being pinged by others), being able to round-robin ping a group of addresses can help alleviate the bombardment.
My view is that the list would always be looped in a round-robin fashion, so that a different address is pinged during each subsequent interval.
2. Ability to specify the number of consecutive failed replies before triggering:
Justification: On a network that is susceptible to intermittent packet loss (e.g. WANs, including 3G, 4G/LTE, and WiFi) a single timeout is rarely an indication of a problem requiring action. In fact, depending on the network quality, even a small number of consecutive timeouts can be expected/tolerated. For this reason it is helpful to be able to set thresholds that best accommodates the quality of the monitored network.
Thanks