Im using Elastic stack 6.2.2, Centos 7.
LS-Collector have 8GB RAM, 8vPCU.
Number event proccess : 4k e/s
My data flow :
Filebeat -> LS-Collector -> RabbitMQ -> LS-Proccessor -> ES
In LS-Collector, i use file output filter to debug. Enabled filebeat debug
In client, i try to make some log example, all log pushed to LS-Collector via filebeat (1)
In LS-Collector, i check file output and i saw: sometime log was double, and logstash queue increase ..
In RabbitMQ, i checked in RabbitMQ mamangement, no data in queue. Checked on Pipeline monitor (in Kiabana), number event proccessed < all event sent by filebeat . Have same result when i check log in kibana. Not enought event/document from client (sent by filebeat) ...
Why did queue in LS-Collector increase? Why are some documents not pushed forward to RabbitMQ and LS-Processor from LS-Forwarder.
P/s: I'm using multiple pipeline in LS and make sure all event will be proccess by Logtash.
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.