Bootstrapping Microservices using Akka, Kafka ans Spark by Alex Silva55:16 12582 views 100% Published 4 years ago
This talk was recorded at Scala Days Chicago, 2017. Follow along on Twitter @scaladays and on the website for more information http://scaladays.org/.
Compared to a traditional analysis-centric data hub, today's data platforms need to fulfill many different use cases. The need for real-time, transport-agnostic data protocols have become a crucial feature shared across many different use cases.
During this talk, we will discuss our approach to bootstrapping and bounded context subscription, leveraging a mix of open source technologies and home-grown services aimed at providing a full end-to-end solution.
We will demonstrate and discuss our use of Kafka, Spark Streaming and Akka to orchestrate a unified data transfer protocol that frees developers from having to listen to and process events within their bounded contexts. More specifically:
- Leveraging Kafka as the source of truth
- Topic formats, retention rules and derived topics
- Using Kafka's distributed commit logs to produce durable datasets
- Log compaction and its role in service bootstrapping
- Message delivery guarantees using Akka actors
- Ingestion at scale: consuming data in different formats across different teams at different latencies
- Using Spark Streaming and Akka to perform real-time analysis and implement transfer protocols