Push logstash aggregate map as fields in final event

[Joe_Martin]

I am trying to use an aggregate filter to pull fields from several different but related logs into one "summary" log. There is a clear end_of_task event and I would like to "push"/add all of the fields in the map into the final end_of_task event.

Do I have to use a "event.set" command for each of the possible fields in the map?
event.set('field1', map['field1'])
I would prefer NOT to use this route as the number and names of the fields in the map can vary.
OR
Can I iterate through all the fields in the map in some way?
OR
Can I just push the whole map into the event as an array with event.set?
Something like this event.set('session_summary', map)?

[Badger]

@Joe_Martin

Can I iterate through all the fields in the map in some way?

Yes, you could do

    aggregate {
        task_id => "%{task}"
        map_action => "update"
        end_of_task => true
        code => 'map.each { |k, v| event.set(k, v) }'
    }

You might also be able to do it using the push_previous_map_as_event option, but that would depend on your data.

[Joe_Martin]

Thanks @Badger!

Your comment about "push_previous_map_as event" produced another question, as this is my first time using the aggregate filter.
Due to the nature of my logs, it is very likely that I will have multiple maps being handled at the same time. My understanding of the "push_previous_map_as_event" is that it would push the previous map (and all associated data/fields) out of memory as soon as it saw a log with a different "task_id".
Is this correct?
And/or am I wrong in my understanding that this filter can handle multiple simultaneous maps?

[Badger]

@Joe_Martin

My understanding of the "push_previous_map_as_event" is that it would push the previous map (and all associated data/fields) out of memory as soon as it saw a log with a different "task_id".

That is my understanding of it too.