Watch the logs at first filebeat startup, does it load a template and create the first index? Always check the logs of the first (in the stack for this index) filebeat startup. I think the index with yyyy.mm.dd is the default with very little filebeat customization. This acts like there is no alias. Does the first index may not be ILM'ed, even with the name. Then the second index is created because the filebeat target is "filebeat-{version}", but no alias exists, so it creates that index.
Sorry @rugenl , the first index is filebeat-7.9.2-2020.43-000001 actually. I writed wrongly, my fault.
When I delete a index, do I have to restart the filebear for proper index creation?
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.