We used to store filebeat data from different sources in a different index due to storage size and document category using different ILM policies.
Now we can't do that because if we use ILM, we can't change the index name, so now all documents store on the same index and with the same ILM policy. That's a problem because there are policies with 1 week and others with a year that mage huge difference when it comes to storage.
Is there a way to change de data stream name for filebeat?
Could be an option to set filebeat output to send events to logstash?
It seems possible to handle index name and ilm in a more flexible way form logstash than filebeat.
A solution might be to add a source or app id into the filebeat configuration for the input and the use logstash to split the sources into separate indices or streams, using IF / ELSE statements in Logstash.Output.
Here is a filebeat example that creates and writes to datastream filebeat-8.x.x-elastic-internal, index names are like .ds-filebeat-8.1.2-elastic-internal-2022.04.04-000001
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.