Streaming Microservices with Akka Streams and Kafka Streams

Kafka Streams is purpose built for reading data from Kafka topics, processing it, and writing the results to new topics. With powerful stream and table abstractions, and an exactly once capability, it supports a variety of common scenarios. Akka Streams emerged as a dataflow-centric abstraction for the Akka Actor model, designed for general-purpose microservices and very low-latency event processing. It supports a wider class of application problems and third-party integrations, so it’s less Kafka-centric.

In this webinar, Dean will:

- Discuss the strengths and weaknesses of Kafka Stream and Akka Streams for particular design needs in data-centric microservices, so you’ll feel better informed when making choices.

- Contrast them with Spark Streaming and Flink, which provide richer analytics over potentially huge data sets where scalability through “automatic” partitioning is required, so you can determine when to choose them instead.

[scarface]

I just seem kafka streams to be conceptually simple. Akka streams has bunch of conceptual overhead like backpressure, akka persistence ect which is not even needed in kafka. Akka code also tries to be too smart by using some crazy patterns like magnet pattern in their akka http.

Also kafka streams has exactly once which you lose with akka streams. Speaker downplayed this fact by saying they are more clever and conceptual to take shortcuts.

I think kafka is going to akkas lunch. Price you pay for overcomplicating stuff.

[Pifagorass]

'If Kafka acknowledge your are sure it's persisted to disk' - I'm not expert ok Kafka, but by default only manager of partition needs to ack message and again by default flush on manager will not be invoked. Even person who's thinking to start using Kafka knows it's loose marketing nonsense