Force filebeat index to use specific/correct type

Im pulling in some stdout/stderr from kubernetes via filebeat. When the indexes are created both the numeric and string values are showing up in kibana with '?' for type.

Looking at the index via GET /filebeat-<date> I see that the string fields are marked as 'text' but so is the numeric field.

Why aren't the fields being recognized as the correct type and how do I fix this?

Edit - I get that I need to setup a template but given that it's using filebeat I didn't want to mess up any defaults.

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