That's basically the approach, so maybe there's some missing detail.
You need an index-pattern with two different geo_point fields. One geo_point would represent the origin of the request, the other geo_point would represent the destination.
And what does it show in Maps when you try to add this point-2-point layer? Does it show your index pattern in the dropdown? What do the origin and destination dropdowns show? Are they just empty?
It might have to do with the filtering based on the current extent of the map and the nature of the data.
The point2point layer gives an aggregate view of connections between sources and destinations (e.g. for worldwide traffic in&out of data-centers). It won't show every single request to the data-center (since there are possibly millions), but aggregate requests coming from nearby locations in a single line.
In the screenshot you posted, are the destinations inside the view?
I would troubleshoot with adding two separete document layers:
Calle.location (sources)
location_aire (destinations)
Then also add the point2point layer.
Zoom out so your entire data is in the view. Does it show lines going in/out of the destinations?
Yeah the windmill is one of the destinations, and the cars are the source. Certainly they don't have common values, because windmills measure air quality and the cars traffic intensity.
I would troubleshoot with adding two separete document layers:
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.