Currently, metrics gathered from node_exporter or prometheus scrapers must be output via remote write within the same fluentbit instance, becuase they are part of a 'different pipeline' from regular log-formatted data. This could be a limitation for some users who prefer to relay ALL data through one set of fluentbit nodes before splitting data off into different sinks.
This would help with a common use case which uses relays per zone to forward data to one centralized system which sends data to their storage back-ends.
Currently, metrics gathered from node_exporter or prometheus scrapers must be output via remote write within the same fluentbit instance, becuase they are part of a 'different pipeline' from regular log-formatted data. This could be a limitation for some users who prefer to relay ALL data through one set of fluentbit nodes before splitting data off into different sinks.
This would help with a common use case which uses relays per zone to forward data to one centralized system which sends data to their storage back-ends.