I have a requirement to push a subset of universal and heavy forwarders originating data to a third party, for which I enabled a set of HFs for data forwarding alone.
This is working fine, as data arrives uncooked to a target syslog-ng.
The troublesome part was being asked to ensure the HF resends the data in case the target undergoes maintenance, or has an outage lasting up to 2 days.
Considering Persistent Queues don't work over splunktcp streams, is it even an option for me to push uncooked data to the HFs, enabling a standard TCP input (not splunktcp) with Persistent Queue enabled, say, to 200GB?
Never heard of anyone using this approach. Would this work?