Patterns for Persistence and Streaming in Cloud Architectures
summary
Shows concrete patterns using Apache Cassandra and Apache Kafka: a live nine-node cluster spread across three data centers (AWS, GCP, Azure) replicates data for multi‑cloud availability; a reference app (Killer Video) uses gRPC services, DAOs and the DataStax drivers and runs successfully in Docker or as a managed service; common pattern is producer→Kafka→consumer→Cassandra with Kafka Connect sinks (DataStax Kafka connector) persisting streams; Cassandra offers tunable consistency, partitioned tables modeled per query and is recommended for very high‑volume tabular workloads.
AI-written summary · about
chapters (21)
- 0:00Introduction and welcome
- 2:26About the conference and speaker
- 4:58Purpose and scope of the talk
- 6:04Evolution of software architectures
- 9:16Modern microservice architectures
- 12:28Multi-cloud and hybrid architectures
- 12:49Selecting infrastructure for cloud apps
- 17:06Key architectural quality attributes
- 21:37Cost of change vs lock-in
- 22:43Abstracting database dependencies
- 23:05Choosing the right database
- 24:30Overview of Cassandra
- 27:01Killer Video reference app
- 30:39Cassandra data modeling principles
- 33:51Deployment options for Cassandra
- 35:18Polyglot persistence in microservices
- 37:46Introduction to streaming with Kafka
- 40:37Comparing Cassandra and Kafka
- 42:29Integration patterns for persistence and streaming
- 44:41Best practices summary
- 45:45Learning resources and Q&A
related talks
30mArchitecting Applications With Multiple Open-Source Big Data Technologies
59mStreaming Big Data w/Spark, Spark Streaming, Kafka, Cassandra & Akka
50mApplication development and data in the emerging world ...
46mKafka Streams at the Big Data Crossroads at Lightning Speed
14mA Cloud-native Architecture for Replicated Data Services
50mIntroduction to Apache Kafka
18mApache Cassandra and Cloud Architectures
26mLeveraging Data in Motion in a Cloud-first World