Event Messaging and Streaming with Apache Pulsar
summary
Apache Pulsar positions itself as a cloud-native event streaming platform that separates compute and storage by delegating durable logs to Apache BookKeeper, supports built-in geo-replication and four subscription modes (exclusive, failover, shared, key_shared), and claims production-scale throughput (examples of “million messages/sec” are cited). It bundles serverless Pulsar functions, tiered storage offloads to S3/HDFS, auto-rebalancing for added nodes, and has growing adoption (GitHub stars and contributors rising; used at Yahoo, Splunk, Verizon). Managed Astra and on-prem Luna options are available with a free $25/month starter tier.
AI-written summary · about
chapters (9)
- 0:00Introduction to event computing
- 3:55Why event-driven systems
- 7:30Event streaming basics
- 13:09Event-driven vs batch processing
- 19:11Apache Pulsar overview
- 26:50Pulsar features and benefits
- 34:14Why choose Apache Pulsar
- 43:25DataStax streaming demo
- 52:36Resources and wrap-up
related talks
34mEvent Streaming in the Cloud Native World with Apache Pulsar
59mEvent Streaming and Processing Using Apache Pulsar
48mApache Pulsar: Event Streaming in The Cloud Native World
48mEvent Streaming and Processing using Apache Pulsar and Apache Pinot.
35mStreaming Data Made Easy
45mApache Pulsar: Finally an Alternative to Kafka?
41mDeep Dive into Building Streaming Applications with Apache Pulsar
51mApache Pulsar 101: architecture, concepts & benchmarks