Polling failures to output chain

[axrayn]

Rather than simply logging errors to the logstash log file, would it be possible to pass an event back to the output chain with the error message for things like snmp timeout?

Currently you can dashboard data against a host object, but this would also allow to mark a device as unresponsive on a dashboard rather than just not appearing because there's no data in the last time period.

[colinsurprenant]

great idea - I would make that configurable with an option like timeout_as_events or something similar.