medium10 pts · first answer countskafkadelivery-semantics

Duplicates that do not matter

A pipeline writes the latest status of each shipment into a table keyed by shipment_id using upserts. Delivery is at-least-once. Why are duplicates mostly harmless here?

A.Upserts are faster than plain insertsB.Writing the same key and value again gives the same rowC.Kafka never sends duplicates to sink connectorsD.The table rejects writes that arrive twice

Discussion 0

Join free to post in the discussion.

Duplicates that do not matter

Discussion 0

Related questions