Understanding the filebeat loglines

[cyb3rjithu (Jithin)]

I was just going through the loglines created by filebeat.
Can someone please explain to me what exactly the following logline indicate,

2017-02-14T10:28:57+05:30 INFO Total non-zero values: registrar.states.cleanup=292 libbeat.publisher.published_events=86371133 registar.states.current=3 registrar.writes=42763 filebeat.harvester.closed=507 filebeat.prospector.log.files.truncated=217 filebeat.harvester.started=507 publish.events=86372439 registrar.states.update=86372439

what does registrar.satetes.cleanup=292 means? what about filebeat.harvester.started=507?

and in some loglines i can see some negetive values like in the following logline,
2017-02-09T14:06:36+05:30 INFO Non-zero metrics in the last 30s: registar.states.current=-8 publish.events=67580 filebeat.harvester.closed=2 filebeat.harvester.open_files=-1 registrar.writes=34 registrar.states.update=67580 libbeat.publisher.published_events=67562 filebeat.harvester.started=1 registrar.states.cleanup=9 filebeat.harvester.running=-1

I'm really new to filebeat and someone please explain to me what is going here.
Thanks

[ruflin]

This are some internal statistics that are logged every 30s (default). Only the diff of the values is logged. So if less harvesters are running then 30s ago, you get a negative value. We introduced these logs to have some insights into what is actually happening in filebeat without having to turn on debug log level. There is currently no documentation for the entries but most of them should be self explanatory:

507 in your case means 507 new harvesters were started ...