Multiple Term Join in Elastic Maps

I have a question regarding usage of multiple term join within a layer of Elastic Maps.

I have an index that stores the documents with geo_point location and some static values and other index with data that updates every 5 minutes. That second index gathers parameters of a network (cpu load, traffic outbound, traffic inbound...) but each parameter is stored as a unique document, I use 2 fields (parameter and value) to differentiate them.

My issue comes when using multiple term joins with filter for each parameter within a layer of the map, I'm capable of retrieving each value correctly if I use different aggregations for each parameter (average of value for parameter 1 and maximum value for parameter 2) but I can't use the same aggregation (average, for example) for both or I get the same value even though each layer is filtered and the values are completely different.

I can provide any additional information if it helps.

Thank you in advance.

I can provide any additional information if it helps.

Could you please provide some additional information. It would be great to know

  1. index mappings for each index
  2. sample document for each index
  3. terms join configurations

I forgot to mention that I'm using version 7.6.2.

Here's everything:

Mappings
-Mapping Index 1

-Mapping Index 2

Sample documents:
-Sample Index 1:

-Sample Index 2:

Term Join config:

Thank you!

Thanks for all of the details, this is a really neat use case.

Unfortunately, there is a bug where terms joins with the same metric, field, and index pattern have a naming collision. This is why the values are the same, both metrics are writing to the same feature property.

I have opened https://github.com/elastic/kibana/issues/63287 to track.

1 Like

Thank you so much, I'll keep up with the github issue for updates !

Just saw that the issue got fixed but just to be clear, which versions will this fix affect ? I believe the merge mentioned [7.x] ?

Thanks anyways !

The fix will be in 7.8 release.

1 Like

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.