more frequent ping check

We worked out a method for this in Nagios via a plugin uses fping to calculate number of successful pings within successive 5 minute intervals (ping once per second).  It relies on the idea that you can have (in Nagios) a plugin launch itself as a daemon and then spit data back into the system via passive results.  In the LM world AFAIK, seems like that would be either very hard or impossible, but perhaps it could be arranged to run outside LM and feed results back in via an eventsource.  It would be nice to integrate the plugin launcher into LM itself, but it should still be possible to do.  I would have to get the host list from LM via the API rather than via a hostgroup, as we do in Nagios.  The part that would be missing here I think would be that we were also able to graph packet loss and average RTT for each target within the continuous ping interval.  There would be no way to do that with an eventsource, but it could be done with a datasource that reads the data out of a file or REST API.  Would work, just a bit clunky.

Regards,

Mark

I would like to second this request.  We have had to rely on secondary monitoring tools to accomplish this task.  60 second ping intervals does not always catch dropped packets.  

you "could" do this via a script based datasource if you wanted to. it's not elegant and bypasses the normal ping monitoring, but this could be accomplished very easily with a script that runs for, let's say 50 seconds, and polls every 5 seconds, and then stops if ping loss is detected and reports back, or success if it didn't.