How do I group per N minutes and remove duplicates within those?

[rune_hellem]

The inital search is this:

index=myindex myapplication UID=* IDX=* IDOK=*  | dedup IDX |  table _time,UID,IDX,IDOK 

I have been asked to create a report that shows the same for lets say the last 24 hours — but not removing all duplicates, only duplicates within each 5 minutes time slot. I guess time slot is created using span, but not sure how to ensure that I do not get duplicate IDX'es.

[kmorris_splunk]

Give this a try:

index=myindex myapplication UID=* IDX=* IDOK=*  | bin _time span=5m | dedup _time IDX |  table _time,UID,IDX,IDOK

The bin will group in 5 minute chunks. Doing the dedup on both _time (a five minute chunk) and IDX will dedup on IDX within the five minute block.