The DSL has some unique features that aren't in the Processor API, such as:
- KStream and KTable abstractions. - Support for time windows (tumbling windows, hopping windows) and session windows. The Processor API only has stream-time based `punctuate()`. - Record caching, which is slightly better than state store caching for the Processor API ( http://docs.confluent.io/current/streams/developer-guide.html#memory-management ) You can re-implement most of these features in the processor API too (after all, the DSL itself is based on the Processor API), but that's manual DIY work. That being said, you can also combine the DSL and the Processor API via the DSL's `process()` and `transform` / `transformValues` operations. See [1] for an example. Best, Michael [1] https://github.com/confluentinc/examples/blob/3.2.x/kafka-streams/src/test/java/io/confluent/examples/streams/MixAndMatchLambdaIntegrationTest.java On Sun, Mar 5, 2017 at 8:01 PM, Shimi Kiviti <shim...@gmail.com> wrote: > Hi Eveyone, > > I have seen a few times (here in the mailing list )were someone wanted to > use kafka streams DSL for something that wasn't possible to do with the DSL > and the suggestion was to use the processor API. > > I was wondering, except the fluent functional code of the DSL, are there > any other benefits of using the DSL over the processor API? > > thanks, > Shimi >
