Thank you for the response. That did work for that simple example. Unfortunately my data has a lot of other fields underneath the array so I cannot specify the specific field to mvexpand. In this slightly more complicated example I am not able to unroll by just specifying one field like in the solution you provided. (My dataset has many keys underneath the jvm_memory field so naming them manually won't work) | makeresults | eval _raw="
{
\"misc_field\": 0,
\"content\":{
\"jvm_memory\": [{\"num\":1.0, \"data\": {\"test\":2.4, \"test2\": 2.3}},{\"num\":2.0, \"data\":{\"test\":3, \"test2\": 2.3}}],
\"field_test2\": 3
}
}"
| spath | spath path=content.jvm_memory{} output=event_data | mvexpand event_data | eval _raw=event_data | kv This above yields what I want (The Json array elements have been split into separate rows and the keys have become columns)
... View more