Who needs batch these days?

on YouTube ↗01h 29m

" summary

Compares real demos: Flink reads MySQL and writes CSV/HDFS but lacks built-in batch restartability; Spring Batch runs parallel flows importing posts/comments/users/votes with chunk size 1000, supports transactional commits, retries and restart from the exact failure point; Spark Streaming shows 1s micro‑batches ingesting tweets (firehose) to HDFS; an ETL workflow checks 24 hourly files and emails missing ones; a port scanner partitions 65,000 ports (demo scans low 500, finds ports 53/80/443). Uses hybrid streaming+batch rather than replacing batch.

AI-written summary · about

" chapters (17)

1. 0:00Introduction and welcome
2. 2:36Shiny object syndrome in tech
3. 5:46Batch vs streaming definitions
4. 14:05Why move from batch to streaming
5. 15:31Apache Flink demo overview
6. 27:59Running the Flink ETL demo
7. 36:57Spring Batch demo introduction
8. 43:21Running the Spring Batch ETL demo
9. 50:11Data science use cases
10. 53:00Workflow orchestration with batch
11. 1:01:44Running the Spring Batch workflow demo
12. 1:03:49Non-interactive processing
13. 1:09:27Port scanner batch demo
14. 1:15:13When to choose batch or streaming
15. 1:18:05Integrating batch with streaming
16. 1:24:59Spring ecosystem advantages
17. 1:27:48Batch's enduring relevance