With 6.0 we removed the support two have 2 outputs enabled in parallel as it was more an accidential feature and caused issues. Interestingly your feature request here is more about routing events to the correct output. If you need to route your events and prefer not to run two beats binaries, I would recommend you to put Logstash in the middle for the routing.
The "accidential feature" of being able to send data to Logstash and Elasticsearch in parallel is in all Filebeat versions previous to 6.x. But I'm not sure if that fully covers your use case as all events are always sent to both outputs.
Like yodog, I need to route Project A prospectors (4) to Logstash Port 1 and Project B prospectors (2) to Logstash Port 2. It is odd to me that this common use case is not supported without running another Filebeats instance.
That said, where can I find a how-to guide to run two different Filebeats instances beyond:
Yes, you can run two filebeat instances in parallel. Make sure they use a different data directory so they have 2 different registrar files.
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.