Grok fiels are removed by aggregate section

Elastic Stack Logstash
1

I have follwoing file structure:

id, iduser,datetimInit, dateTimeends
0001 0001 2023-02-03 04:45:16.78 2023-02-03 04:46:16.78
0002 0001 2023-02-03 08:45:16.78 2023-02-03 08:46:16.78
0003 0002 2023-02-04 04:45:16.78 2023-02-04 04:46:16.78

Where: I need to add duration time (datetimeEnds - dateTimeInit) , and full duration (sum duration by idUser) time by iduser. i have the following configuration file:

filter {	
	grok {
		match => { 
			"message" => "%{NUMBER:id} %{NUMBER:idUser} %{TIMESTAMP_ISO8601:dateTimeInit} %{TIMESTAMP_ISO8601:dateTimeEnds}" 
		}
	}
	date { match => ["dateTimeInit", "ISO8601"] target => "dateTimeInit" }
	date { match => ["dateTimeEnds", "ISO8601"] target => "dateTimeEnds" }
	
	#add duration by row
	ruby { 
		code => "event.set('duration', event.get('dateTimeEnds').to_i - event.get('dateTimeInit').to_i); "
	}
	
	#add fullduration depending on same iduser
	aggregate {
       task_id => "%{idUser}"
       code => "
			map['full_duration'] ||= 0 ; 
			map['full_duration'] += event.get('duration')
			event.cancel()
		"
	   push_previous_map_as_event => true
       timeout => 3
    }
}

This is the output:

{
    "full_duration" => 2,
       "@timestamp" => 2023-02-06T13:08:18.923849100Z,
         "@version" => "1"
}

Where Dates and idUser fields are missing.

How could I add all the fields to the output, not just full_duration?

Thank you in advance

2

That pushes the previous map as an event. The only field you added to the map is map['full_duration'], so that is the only non-standard field that your event has. You need to copy the other fields that you want to preserve to fields in the map hash.