Hello guys,
I am running Metricbeat on Windows and observed that first 10 minutes of system events after service restart is not logged anywhere. Is it bug or a feature? Even when I switch to queue.disk, the first data in file appears after 10 minutes.
What I discovered, when changing true to false in following snippet, the issue disappears.
hi @Jaroslav , When set to true the reload.enabled just enables dynamic config reload. Could it be that you have previously set the collection period to 10 min but config was not reloaded?
Soon as you enabled this option then the new configuration took effect?
If you set it back to false do you see the same 10 min delay?
hello @MariannaD,
I'll try to extend the description. I am actually not changing configuration much, only from time to time, so I set this to true in case I forget to restart the service. Only when I restart the metricbeat service, the gap in data appears - and because 10 minutes coincides with 600-second setting, it seems like the beat configuration is not initialized properly at the the service starts, but in the next reload.
metricbeat running, reload.enabled: true
Metricbeat restarts (net stop metricbeat && net start metricbeat)
For 10 minutes, no event logged to Elastic
Getting events in Elastic => 10 minutes gap in data
metricbeat running, reload.enabled: false
Metricbeat restarts (net stop metricbeat && net start metricbeat)
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant
logo are trademarks of the
Apache Software Foundation
in the United States and/or other countries.