I am using Kafka Streams to build realtime matching engine. One of my use case is to iterate records that share couple of common values in the value object. I copied over the common values and made it as prefix for the key along with its primary key which will make they key look like R#1 ( 1 is PK and R is the common field that will be used for retrieval). Proof of concept using Prefix Scan against the state store provided us the desired results. Prefix scan feature was implemented using KIP-614 and was wondering if anyone could answer my question related to ordering guarantee, if I push record in the state store with keys in following order - R#1, RK#2, RB#3, R#4 and use prefix scan to retrieve items with prefix R# will the prefix scan iteration order be deterministic based on order of insertion ? i.e R#1, R#4